AustLII Home | Databases | WorldLII | Search | Feedback

Federal Court of Australia

You are here: 
AustLII >> Databases >> Federal Court of Australia >> 2019 >> [2019] FCA 849

Database Search | Name Search | Recent Decisions | Noteup | LawCite | Download | Context | No Context | Help

Australian Securities and Investment Commission v  Whitebox Trading  Pty Ltd (No 7) [2019] FCA 849 (7 June 2019)

Last Updated: 7 June 2019

FEDERAL COURT OF AUSTRALIA

Australian Securities and Investment Commission v  Whitebox Trading  Pty Ltd (No 7) [2019] FCA 849

File number:


Judge:


Date of judgment:
7 June 2019


Catchwords:
CORPORATIONS – securities index arbitrage trading – alleged market manipulation – whether the placement of certain orders for XJO Securities, including certain order cancellations and reductions, in the Pre-Open Phase were made for the purpose of affecting Opening Prices in the Opening Single Price Auction of the market conducted by the Australian Securities Exchange – whether defendants were trading with the intention of creating or causing the creation of a false or misleading appearance with respect to the market for and/or price for trading in XJO Securities under s 1041B(1)(b) of the Corporations Act 2001 (Cth) – whether defendants were trading with the intention of creating an artificial price for trading in XJO Securities under s 1041A(1)(c) of the Corporations Act 2001 (Cth)


Legislation:
Australian Securities and Investments Commission Act 2001 (Cth), s 33
Competition and Consumer Act 2010 (Cth)
Corporations Act 2001 (Cth), ss 180(1), 760A, 1041A, 1041B, 1041 E
Corporations Act 1989 (Cth), s 998(1)
Criminal Code Act 1995 (Cth), Sch 2
Criminal Procedure Act 2009 (Vic), s 302
Evidence Act 1995 (Cth), ss 79(1), 140(2)
Securities Exchange Act 1934 (US), s 9
Securities Industry Act 1970 (NSW), s 70
Trade Practices Act 1974 (Cth)


Cases cited:
Adams v Director of the Fair Work Industry Inspectorate [2017] FCAFC 228
Australian Competition and Consumer Commission v Metcash [2011] FCAFC 151; (2011) 198 FCR 297
Australian Competition and Consumer Commission v Olex Australia Pty Ltd [2017] FCA 222
Australian Securities and Investments Commission v Nomura International plc [1998] FCA 1570; (1998) 89 FCR 301
Australian Securities and Investments Commission v Westpac Banking Corporation (No 2) [2018] FCA 751; (2018) 357 ALR 240
Australian Securities and Investments Commission v  Whitebox Trading  Pty Ltd & Anor [2017] FCAFC; [2017] FCAFC 100; (2017) 251 FCR 448
Australian Securities and Investments Commission, in the matter of  Whitebox Trading  Pty Ltd v  Whitebox Trading  Pty Ltd (No 5) [2018] FCA 1059
Australian Securities and Investments Commission, in the matter of  Whitebox Trading  Pty Ltd v  Whitebox Trading  Pty Ltd (No 6) [2018] FCA 1077
Briginshaw v Briginshaw [1938] HCA 34; (1938) 60 CLR 336
Casaclang v WealthSure Pty Ltd [2015] FCA 761; (2015) 238 FCR 55
Communications Electrical, Energy, Information, Postal, Plumbing and Allied Services Union of Australia v Australian Competition and Consumer Commission [2007] FCAFC 132; (2007) 162 FCR 466
Director of Public Prosecutions (Cth) v JM [2012] VSCA 21; (2012) 37 VR 1
Director of Public Prosecutions for the Commonwealth of Australia v JM [2013] HCA 30; (2013) 250 CLR 135
Fame Decorator Industries Pty Ltd v Jeffries Industries Ltd [1998] NSWSC 157; (1988) 28 ACSR 58
In the matter of Kidder Peabody & Co 18 SEC 559
North v Marra Developments Limited [1981] HCA 68; (1981) 148 CLR 42
R v Manasseh and Austin [2002] NSWCCA 27
Salvation Army (Vic) Property Trust v Ferntree Gully Corporation [1952] HCA 4; (1952) 85 CLR 159
SEC v LEK Securities Corporation 276 F Supp 3d 49 (2017)
Tillmanns Butcheries Pty Ltd v Australasian Meat Industry Employees’ Union [1979] FCA 85; (1979) 42 FLR 331


Date of hearing:
2 to 17 July 2018, 19 July 2018, 23 July 2018,
25 to 27 July 2018


Registry:
New South Wales


Division:
General Division


National Practice Area:
Commercial and Corporations


Sub-area:
Corporations and Corporate Insolvency


Category:
Catchwords


Number of paragraphs:
568


Counsel for the Plaintiff:
Mr J A Halley SC with Mr I J M Ahmed and
Mr P J Holmes


Solicitor for the Plaintiff:
Johnson Winter & Slattery


Counsel for the Defendants:
Mr R G McHugh SC with Mr M J Steele SC and
Mr L T Livingston


Solicitor for the Defendants:
Thompson Eslick Solicitors



ORDERS


NSD 383 of 2016

BETWEEN:
AUSTRALIAN SECURITIES AND INVESTMENT COMMISSION
Plaintiff
AND:
 WHITEBOX TRADING  PTY LTD
First Defendant

JOHANNES HENDRIK BOSHOFF
Second Defendant

JUDGE:
YATES J
DATE OF ORDER:
7 JUNE 2019



THE COURT ORDERS THAT:

  1. The Originating Process be dismissed.
  2. Unless the question of costs can be agreed:
(a) the defendants file and serve written submissions on that question (limited to three pages) by 4.00 pm on 28 June 2019;
(b) the plaintiff file and serve written submissions in response (limited to three pages) by 4.00 pm on 10 July 2019; and
(c) the defendants file and serve any submissions in reply (limited to two pages) by 4.00 pm on 17 July 2019.
  1. The question of costs be determined on the papers unless a party wishes to be heard on that question orally.
Note: Entry of orders is dealt with in Rule 39.32 of the Federal Court Rules 2011.


REASONS FOR JUDGMENT

[1]
[7]
[7]
[20]
[20]
[30]
[45]
[47]
[55]
[81]
[90]
[105]
[131]
[131]
[143]
[146]
[149]
[152]
[155]
[159]
[163]
[163]
[166]
[182]
[196]
[216]
[238]
[257]
[269]
[287]
[288]
[292]
[310]
[313]
[313]
[314]
[336]
[336]
[344]
[366]
[370]
[374]
[386]
[386]
[400]
[419]
[428]
[443]
[453]
[453]
[463]
[472]
[472]
[486]
[492]
[493]
[503]
[511]
[511]
[516]
[521]
[541]
[549]
[549]
[554]
[558]
[558]
[562]
[567]
[568]


YATES J:

INTRODUCTION

  1. This proceeding concerns alleged market manipulation in the context of securities index arbitrage trading. The plaintiff, the Australian Securities and Investments Commission (ASIC), alleges that, over five successive Serial Expiry Days in 2012 (being 19 April 2012, 17 May 2012, 19 July 2012, 16 August 2012 and 18 October 2012), the first defendant,  Whitebox Trading  Pty Ltd (Whitebox), and the second defendant, Mr Boshoff, engaged in a deliberate strategy to manipulate the prices of all 200 individual securities comprising the XJO Index in the OSPA (Opening Single Price Auction), for the sole or dominant purpose of achieving a more favourable Basis (the mispricing between the traded price of SPI Futures and their Fair Price), and thereby to make the index arbitrage positions that Whitebox traded on each of those days in SPI Futures and XJO Securities profitable, or more profitable.
  2. ASIC contends that the heart of this strategy was the placement of very large orders or order amendments for XJO Securities late in the Pre-Open Phase, which the defendants could not have had any reasonable expectation that they could trade in the OSPA. These orders were either cancelled or reduced in volume shortly prior to the Rotation in the OSPA in which the securities were due to trade. ASIC contends that this was done intentionally to manipulate the Opening Prices of the securities in the OSPA and, thereby, Basis. ASIC further contends that none of these orders or order amendments represented genuine supply or demand for XJO Securities.
  3. By reason of this conduct, ASIC alleges that the defendants contravened certain market manipulation provisions of the Corporations Act 2001 (Cth) (the Act)—s 1041A and s 1041B. In addition, ASIC alleges that, by causing or otherwise permitting Whitebox to contravene the market manipulation provisions, Mr Boshoff also contravened s 180(1) of the Act.
  4. In Australian Securities and Investments Commission v  Whitebox Trading  Pty Ltd & Anor [2017] FCAFC; [2017] FCAFC 100; (2017) 251 FCR 448, the Full Court held, in the context of hearing a separate question in this proceeding, that Ch 2 of the Criminal Code (being the Schedule to the Criminal Code Act 1995 (Cth)) was not engaged in proceedings brought for the imposition of a civil penalty for a contravention of s 1041A and s 1041B of the Act, such as the present. The Full Court reasoned that, if otherwise, ASIC would have to prove the criminal fault elements derived from Ch 2 of the Criminal Code for each of the proscribed physical elements, rather than any state of mind that may be drawn from the terms of the two provisions themselves. The Full Court noted (at [9]) that it was common ground that it should not, ahead of the present hearing, venture into any consideration of the state of mind that might be required to be established in order to prove a contravention of the two provisions in civil penalty proceedings. This was because, at the hearing of the separate question, that issue had not been the subject of full and proper consideration in the context of the pleadings, evidence and appropriate submissions. As events have transpired, that issue has not even fallen for consideration and determination now. ASIC framed its case against the defendants as one of specific intentional conduct, and the parties have joined issue on that question. The present hearing has been conducted accordingly. Therefore, ASIC’s case against the defendants turns on whether it has established, on the balance of probabilities, that the defendants acted with the intention it alleges.
  5. Although ASIC’s case is brought against Whitebox and Mr Boshoff, for ease of exposition I will simply refer to Whitebox when summarising and discussing the activities of placing, amending, cancelling and/or reducing orders for XJO Securities and purchasing or selling SPI Futures that were undertaken on the five Serial Expiry Days in question, and on the other trading days discussed in the evidence.
  6. Capitalised terms and expressions used in these reasons have the meanings given to them in the Primer and Glossary reproduced in the Schedule to these reasons: see pages 144 to 169 below.

LEGAL BACKGROUND

Relevant provisions

  1. Section 1041B(1)(b) of the Act provides:
A person must not do, or omit to do, an act (whether in this jurisdiction or elsewhere) if that act or omission has or is likely to have the effect of creating, or causing the creation of, a false or misleading appearance:
...
(b) with respect to the market for, or the price for trading in, financial products on a financial market operated in this jurisdiction.
  1. ASIC alleges that the defendants contravened this provision in two ways.
  2. First, the defendants engaged in acts that had, or were likely to have, the effect of creating, or causing the creation of, a false or misleading appearance as to the market for XJO Securities in the Pre-Open Phase of the market conducted by the ASX. ASIC referred to this as the “market for” case.
  3. Secondly, the defendants engaged in acts that were likely to have the effect of creating, or causing the creation of, a false or misleading appearance as to the price for trading in XJO Securities on the ASX. ASIC referred to this as the “price for trading” case.
  4. In respect of Whitebox, the alleged contraventions, in each case, relate to each of the Serial Expiry Days in question in this proceeding. In respect of Mr Boshoff, the alleged contraventions relate to each Serial Expiry Day other than 16 August 2012.
  5. Section 1041A of the Act relevantly provides:
A person must not take part in, or carry out (whether directly or indirectly and whether in this jurisdiction or elsewhere):
(a) a transaction that has or is likely to have; or
(b) 2 or more transactions that have or are likely to have;
the effect of:
(c) creating an artificial price for trading in financial products on a financial market operated in this jurisdiction; ...
  1. ASIC alleges that the defendants took part in a series of transactions that were likely to have the effect of creating an artificial price for trading in XJO Securities. I will refer to this as the “artificial price for trading” case. In respect of Whitebox, the alleged contraventions relate to each of the Serial Expiry Days in question in this proceeding. In respect of Mr Boshoff, the alleged contraventions relate to each Serial Expiry Day other than 16 August 2012.
  2. These market manipulation provisions are found in Ch 7 of the Act, whose main object is set out in s 760A:
The main object of this Chapter is to promote:
(a) confident and informed decision making by consumers of financial products and services while facilitating efficiency, flexibility and innovation in the provision of those products and services; and
(b) fairness, honesty and professionalism by those who provide financial services; and
(c) fair, orderly and transparent markets for financial products; and
(d) the reduction of systemic risk and the provision of fair and effective services by clearing and settlement facilities.
  1. ASIC submitted that consistently with s 760A, numerous authorities have emphasised that ss 1041A and 1041B (or their precursors) are designed to protect the integrity of the market and to ensure that it operates in a transparent and efficient manner.
  2. In R v Lloyd (1996) 19 ACSR 528 (Lloyd), Murray J said (at 548):
... an essential feature of a system of trading in securities in listed corporations by bidding for stock at centralised Stock Exchanges [is] that the system does reflect the interplay of the market forces of supply and demand... Investor confidence is the quality which the market must be able to maintain. If investors become suspicious that the values of securities which the market reflects do not provide a measure of the worth of the stock then it seems obvious that they will be reluctant to invest.
  1. In Fame Decorator Industries Pty Ltd v Jeffries Industries Ltd [1998] NSWSC 157; (1988) 28 ACSR 58 (Fame Decorator), Gleeson CJ said (at 62-63):
Section 998 [the precursor to s 1041B] aims to preserve the integrity of the share market. Markets, in reflecting the interaction of forces of supply and demand, may suffer from a variety of imperfections, including mismatches of information, without such imperfections destroying their integrity.
  1. ASIC submitted that these statements provided the lens through which this case must be approached.
  2. To these statements can be added the following statement by Mason J concerning the object of s 70 of the Securities Industry Act 1970 (NSW) (a precursor of s 1041B(1)(b) of the Act) in North v Marra Developments Limited [1981] HCA 68; (1981) 148 CLR 42 (Marra Developments) (at 59):
It seems to me that the object of the section is to protect the market for securities against activities which will result in artificial or managed manipulation. The section seeks to ensure that the market reflects the forces of genuine supply and demand. By “genuine supply and demand” I exclude buyers and sellers whose transactions are undertaken for the sole or primary purpose of setting or maintaining the market price. It is in the interests of the community that the market for securities should be real and genuine, free from manipulation. The section is a legislative measure designed to ensure such a market and it should be interpreted accordingly.

THE EVIDENCE

ASIC’s evidence

  1. ASIC read the following affidavits:
  2. The only deponent from this group who was cross-examined was Mr Feeney.
  3. ASIC also called expert evidence from William Morgan and Hugo Graves.
  4. Mr Morgan has a Bachelor of Commerce degree awarded by the University of Queensland. His early career was spent in a variety of roles in the IT industry, including computer programming, business analysis, project management and system architecture. From 2005 to 2012, he worked in Hong Kong and Sydney for Credit Suisse in the Equity Derivatives group. From late 2007 until the end of 2012, Mr Morgan was employed by Credit Suisse as an Equity Derivatives Trader in the Sydney office. In this role, he implemented index arbitrage trading between SPI Futures and XJO Securities. He was also involved in trading SPI Futures in 2010 when based in the Hong Kong office.
  5. Mr Morgan prepared three reports which responded to a series of questions dealing with index arbitrage trading, and specifically the trading carried out by Whitebox on the Serial Expiry Days in question in this proceeding.
  6. There were substantial objections to parts of Mr Morgan’s first report (the other two reports were amending reports). One set of objections was directed to paragraphs 101 to 103 of the report and certain sections of Mr Morgan’s Joint Report with Mr de Kantzow and Professor Frino. I ruled on those objections during the course of the hearing: Australian Securities and Investments Commission, in the matter of  Whitebox Trading  Pty Ltd v  Whitebox Trading  Pty Ltd (No 5) [2018] FCA 1059 (Whitebox No 5). Another set of objections was directed to paragraph 129 of Mr Morgan’s first report and related paragraphs, which the parties, by agreement, accepted were dependent for their admissibility on the admissibility of paragraph 129. In the circumstances discussed in a later section of these reasons (see [297] – [302]), the parties agreed that the ruling on the admissibility of these paragraphs should await the cross-examination of Mr Morgan, and should be made in these reasons. For the reasons given at [303] – [306] below, I will admit the challenged paragraphs.
  7. The defendants maintained a challenge to the weight I should give to Mr Morgan’s evidence in light of the extent and nature of his trading activities. The basis for this challenge is adequately summarised at [10] in Whitebox No 5 and is not repeated here. The defendants made clear, however, that, beyond the matters noted above, they did not challenge Mr Morgan’s ability to discuss principles of index arbitrage trading or index arbitrage pricing.
  8. Mr Graves is a stockbroker currently working as a retail investment adviser. He has worked in the field of stockbroking and funds management for over thirty years. He has managed orders for equities and derivatives on behalf of retail and institutional clients on a full-time basis. As part of his duties as an adviser, Mr Graves has managed the placement of orders, and their amendment and cancellation, on the ASX , including in the Pre-Open Phase of the market. Mr Graves prepared a report which responded to a series of questions dealing with the effects of amending or cancelling orders for XJO Securities in the Pre-Open Phase, specifically in relation to the orders, amendments and cancellations placed by Whitebox on the Serial Expiry Days in question in this proceeding.
  9. The defendants criticised substantial parts of Mr Graves’ report which, they submitted, were based on “illogical and absurd reasoning”. I deal with those criticisms below. ASIC suggested in closing submissions that, at times, Mr Graves found it difficult to follow the thrust of certain questions put to him in cross-examination. This suggestion seems to have been made to support Mr Graves’ evidence, as if to contend that, where his evidence was unfavourable to ASIC’s case, Mr Graves might not have fully understood the questions he was asked. If that is the suggestion, I reject it. I accept that Mr Graves was a cautious witness. But I have no doubt that he understood the questions put to him and that he answered those questions honestly and to the best of his ability.
  10. ASIC also relied on a report prepared by Associate Professor Artem Prokhorov, who is an econometrician. Associate Professor Prokhorov performed testing to validate work undertaken by Mr Clifford (who is referred to above) in analysing data and producing a series of tables on trading undertaken by Whitebox on the Serial Expiry Days in question in this proceeding. He was not required for cross-examination.

The defendants’ evidence

  1. The defendants called expert evidence from Michael de Kantzow, Professor Alex Frino, and Professor Michael James Aitken AM.
  2. Mr de Kantzow has worked in financial markets since 1987. From 1992 until 1996, he was employed by A.B.S. White and Co, where he had responsibility for advising retail and institutional clients on equity option strategies and risk management, executing orders on the ASX derivatives trading floor, and supporting the registered trader operation in option market making. From 1996 to 2001, he was employed by HSBC Securities (Australia) Limited, initially in institutional derivative broking and execution, and subsequently as Head of Equity Option Trading. From 2002 to late 2008, he was employed by ABN-Amro Australia as Head of Derivatives Trading, where he was responsible for listed and over-the-counter equity option market making and SPI Futures options trading including, amongst other things, index arbitrage. In 2009, he established Cross Capital Pty Limited, which has conducted derivative trading activity in the equity, interest rate and SPI Futures markets.
  3. Mr de Kantzow prepared a report in which he commented on specific aspects of Mr Morgan’s and Mr Graves’ respective reports, commented generally on how securities index arbitrage trading was undertaken in the Australian market in 2012, and reviewed and commented on auction time lines, graphs and tables on which the defendants relied in their case.
  4. In cross-examination, ASIC mounted a substantial attack on Mr de Kantzow’s credit. On the fourteenth day of the hearing, in the context of an application by ASIC to re-open its case, the defendants agreed to the following acknowledgement:
...the parties have agreed that the defendants accept that Mr de Kantzow’s evidence should be given no credit and that his opinions should be given no weight except insofar as his evidence may support the plaintiff’s case, and, accordingly, the plaintiff is at liberty to rely on his evidence to the extent that it supports the plaintiff’s case...
  1. Professor Frino is Professor of Economics and Deputy Vice Chancellor at Wollongong University. He is also an Executive of Capital Markets Cooperative Research Centre Limited (CMCRC). He has previously worked as a Senior and Visiting Economist at the Sydney Futures Exchange. He was instrumental in developing the S&P/ASX 2000 Futures Contract and selecting the index (the XJO) on which those contracts are written. Professor Frino prepared two reports in which he commented on, and responded to questions directed to, Mr Morgan’s reports.
  2. ASIC criticised Professor Frino’s evidence on the basis that, on the matters in question, his evidence was “entirely academic and theoretical”. This criticism was made in the context of comparing Professor Frino’s evidence with Mr Morgan’s evidence which, ASIC submitted, was based on practical experience in securities index arbitrage trading. Despite its criticism, ASIC did not object to the admissibility of Professor Frino’s opinions. I do not accept that the evidence Professor Frino gave was in any way undermined by the fact that, for example, he had never personally traded SPI Futures or had never been asked before to consider securities index arbitrage activities undertaken in the Pre-Open Phase.
  3. ASIC submitted that “the greatest difficulty” with Professor Frino’s evidence was that he had not analysed the data relating to Whitebox’s order and trading activity on the Serial Expiry Days in question and was not aware of Whitebox’s stated position on its trading strategy. ASIC submitted that these omissions “radically diminish(ed)” the utility of Professor Frino’s opinions in drawing inferences about the defendants’ “intentions and purposes”.
  4. I do not accept that the utility of the opinions expressed by Professor Frino were diminished by these matters. As the defendants correctly pointed out, Professor Frino was not asked to address the defendants’ states of mind. The questions he was asked were directed to specific parts of Mr Morgan’s report. On these matters, Professor Frino had the requisite expertise to express opinions.
  5. Professor Aitken is Professor in ICT Strategy at Macquarie University and the Chief Executive Officer and Chief Scientist of CMCRC. Professor Aitken prepared two reports in which he commented on, and responded to questions directed to, Mr Graves’ report.
  6. ASIC submitted that Professor Aitken was “an unsatisfactory witness in several respects”. First, ASIC submitted that he was argumentative, defensive and resistant to acknowledging errors in his own report. Secondly, ASIC submitted that he showed a lack of familiarity with the contents of his reports and an inattention to the detail in them. Thirdly, ASIC submitted that Professor Aitken displayed a tendency to behave as an advocate, particularly on the need for “surveillance protocol evidence”, rather than focussing on the evidence he had been asked to consider and address.
  7. I accept that, at times, Professor Aitken was argumentative beyond merely seeking to explain the basis for the opinions he had expressed. This was particularly so in his advocacy (because that is what it was) of the need for “surveillance protocol evidence”. I think, however, that, at other times, Professor Aitken’s evidence and demeanour were the product of the frustration he genuinely felt at the unsoundness of opining about price effects in the market for XJO Securities on the basis of rules of thumb (as Mr Graves had done), rather than on the basis of empirical evidence. In light of my findings in relation to Mr Graves’ evidence, some of Professor Aitken’s frustration is understandable.
  8. It is also true that, on occasion, Professor Aitken was corrected on matters of detail in his reports. I do not think, however, that he is to be criticised for that, given the nature and the amount of the data covered in his reports.
  9. Overall, and contrary to ASIC’s submission, I do not accept that Professor Aitken was an unsatisfactory witness. On relevant matters, I do not think that his evidence was seriously undermined.
  10. The most trenchant criticism of Professor Aitken’s evidence by ASIC was that it was “tainted” because of his view that it was not possible to express an opinion on whether the placement of Whitebox’s impugned orders would be likely to have any effect on the Opening Prices of XJO Securities in the OSPA. This was because, in Professor Aitken’s view, there was no scientific or reliable method to recreate the market place to identify what the Opening Prices would have been absent Whitebox’s orders. This is a legitimate opinion. I do not accept that Professor Aitken’s evidence was “tainted” simply because he steadfastly maintained this view—a view that was directly critical of the reliability of the evidence that ASIC adduced on this question.
  11. The defendants also read an affidavit by Michael Rudolf Samerski affirmed 15 November 2017. Mr Samerski is a trading system software engineer who reviewed and verified a series of computer programs used to generate output reports and graphs deployed by the defendants, in particular a number of auction time lines to which I will make further reference. Mr Samerski was not required for cross-examination.

The joint reports

  1. Pursuant to orders made by the Court, joint experts’ reports were prepared by Associate Professor Prokhorov and Professor Aitken; Mr Morgan and Mr de Kantzow; Mr Morgan, Mr de Kantzow, and Professor Frino; and Mr Graves, Mr de Kantzow and Professor Aitken. These reports were tendered in ASIC’s case. An updated summary of responses, prepared by Associate Professor Prokhorov, was also tendered in ASIC’s case.
  2. Evidence was given concurrently by Mr Morgan, Mr de Kantzow and Professor Frino. Evidence was also given concurrently by Mr Graves, Mr de Kantzow and Professor Aitken. The witnesses were cross-examined in the course of giving evidence in these sessions.

Other matters

  1. It is convenient at this juncture to briefly note other matters concerning the basis on which ASIC’s case falls to be considered.
  2. First, the parties accepted that ASIC bears the onus of proof of its allegations and that the civil standard of proof applies, informed by s 140(2) of the Evidence Act 1995 (Cth) which provides:
Without limiting the matters that the court may take into account in deciding whether it is so satisfied, it is to take into account:
(a) the nature of the cause of action or defence; and
(b) the nature of the subjectmatter of the proceeding; and
(c) the gravity of the matters alleged.
  1. It is to be noted in this connection that, in the present case, ASIC seeks, amongst other relief, the imposition of pecuniary penalties and banning orders against each defendant. I bear in mind the admonition expressed by Dixon J in Briginshaw v Briginshaw [1938] HCA 34; (1938) 60 CLR 336 that in reaching reasonable satisfaction, the Court should not proceed on “inexact proofs, indefinite testimony, or indirect inferences”.
  2. Secondly, as the case brought by ASIC is one of intentional conduct, and as there is little direct evidence that elucidates the defendants’ intentions, it will be necessary for the Court to draw inferences from established facts. Thus, consistently with the propositions stated above, ASIC bears the onus of establishing the inferences for which it contends on the balance of probabilities.
  3. In Australian Competition and Consumer Commission v Olex Australia Pty Ltd [2017] FCA 222 (Olex) at [479], Beach J observed:
A finding may be made in the absence of direct evidence. All that is necessary is that the more probable inference from the circumstances that sufficiently appear by evidence, left unexplained, justifies the conclusion. “More probable” means no more than that upon the balance of probabilities, such an inference has a greater degree of likelihood. A party who relies on circumstantial evidence must show that the circumstances raise the more probable inference in favour of what is alleged. It is not sufficient that the circumstances give rise to conflicting inferences of an equal degree of probability or plausibility or that the choice between them can only be made by conjecture. I accept though that the process of inference may involve an intuitive element that is not susceptible to detailed support or explanation.
  1. Thirdly and relatedly, the Court may infer the effect that particular conduct has or is likely to have from its purpose as established on the evidence. Thus, in Director of Public Prosecutions for the Commonwealth of Australia v JM [2013] HCA 30; (2013) 250 CLR 135, (JM), the High Court said at [73]:
Because s 1041A prohibits transactions which are likely to have that effect, it is not necessary to demonstrate, whether by some counterfactual analysis or otherwise, that the impugned transactions did create or maintain an artificial price. It is sufficient to show that the buyer or seller set the price with the sole or dominant purpose described.
  1. Similarly, the Court may infer the purpose of particular conduct from its effect: Olex at [494]; Australian Securities and Investments Commission v Westpac Banking Corporation (No 2) [2018] FCA 751; (2018) 357 ALR 240 (Westpac) at [1938].
  2. Fourthly, where there is an unexplained failure to call evidence, the Court can infer that such evidence would not have assisted the defendants, even in a civil penalty case: Adams v Director of the Fair Work Industry Inspectorate [2017] FCAFC 228 at [147]; Communications Electrical, Energy, Information, Postal, Plumbing and Allied Services Union of Australia v Australian Competition and Consumer Commission [2007] FCAFC 132; (2007) 162 FCR 466 at [28] and [76].

SECURITIES INDEX ARBITRAGE TRADING: AN OVERVIEW

  1. On 9 April 2018, I made orders to facilitate the preparation by the parties of an agreed primer, including a glossary of terms, which would provide a basic description of securities index arbitrage trading. The Primer and Glossary, prepared by the parties, are reproduced in the Schedule to these reasons. The statements made in them can be treated as agreed facts for the purpose of this proceeding.
  2. In these reasons, I have also treated the statements as assumed knowledge on the part of the reader. Proceeding on this basis, I have endeavoured to avoid unnecessary repetition of this background material in the body of these reasons. Nonetheless, it will assist in understanding the context in which the present proceeding arises to provide, at this stage, a brief overview of securities index arbitrage trading which draws on various strands of the Primer and certain aspects of the evidence.
  3. Securities index arbitrage trading involves trading stock index futures against the constituent securities that make up the stock index.
  4. In the case of the ASX, the primary stock futures are SPI Futures. These are futures contracts written over the S&P/ASX 200 Index. The index is conveniently referred to as the XJO. The securities that make up the XJO are conveniently referred to as XJO Securities.
  5. The terms of the futures contracts are standardised and provide for the purchase or sale of the financial instrument on a given future date (the Expiry Date) at a price fixed at the time the contract is made (the Contract Value).
  6. There are two types of SPI Futures—Quarterly SPI Futures (SPI Futures that expire on the third Thursday of March, June, September or December) and Serial SPI Futures (SPI Futures that expire on the third Thursday of each of the other, intervening months). This proceeding is concerned with trading that involves Quarterly SPI Futures. However, the impugned trading took place on Serial Expiry Days.
  7. SPI Futures are cash-settled contracts. They do not require the physical securities making up the index to be transferred on the Expiry Date. Rather, the buyer or seller is required to pay to the counterparty the difference between the Contract Value and the value of the instrument at the Expiry Date. The direction of the movement in value will determine whether it is the buyer or the seller who pays.
  8. SPI Futures derive their value from the index. The XJO has a price level which changes constantly. It is calculated from the prices at which XJO Securities trade during the hours that the ASX is open. A person who holds an amount of constituent securities in the exact proportion in which the securities make up the XJO is able to replicate the overall value of the index.
  9. The strategy involved in securities index arbitrage trading is to make a profit from any real-time price discrepancy between the Fair Price of given SPI Futures and their market value. The measure of this mispricing is called Basis. Arbitrageurs who identify mispricing seek to capture the available potential profit (represented by Basis) by buying an amount of one of the financial products (SPI Futures or XJO Securities) while selling an amount of equivalent value of the other financial product (XJO Securities or SPI Futures), thereby executing trades on the two markets.
  10. These mispricing opportunities do not only arise when the two markets are open. They can also exist at times when the market for XJO Securities is in the Pre-Open Phase (i.e., not open for Normal Trading) and when the ASX24 has commenced to trade SPI Futures. This is because the arbitrageur is able to make calculations based on the indicative prices for each security that are generated and published in the Pre-Open Phase (when buy or sell orders for securities and baskets of securities can be placed) leading up to the OSPA.
  11. The Pre-Open Phase commences at 07:00 to 10:00 (+/- 15 seconds) running through to 10:09 (+/- 15 seconds). The OSPA is a series of opening single price auctions for securities on the ASX which take place after the Pre-Open Phase but before market opening, according to groupings of securities based on the alphabetical precedence of their Ticker Codes. There are five groupings, referred to as Rotation Groups. The auctions are staggered, but each auction (or Rotation) occurs at a precise time that falls within a 30-second window randomly generated by ASX Trade (the platform on which the XJO Securities are traded). The window for A-B securities (Group 1) is 09:59:45 to 10:00:15; for C-F securities (Group 2), 10:02:00 to 10:02:30; for G-M securities (Group 3), 10:04:15 to 10:04:45; for N-R securities (Group 4), 10:06:30 to 10:07:00; and for S-Z securities (Group 5), 10:08:45 to 10:09:15.
  12. As more orders come into the Order Book in the Pre-Open Phase, calculated and published indicative prices might change. Correspondingly, arbitrageurs’ assessments of mispricing might also change. Arbitrageurs can monitor, calculate and re-calculate Basis throughout this phase. Fast Basis is a calculation of Basis undertaken at the rate of five times per second. It includes the effect of the trader’s own orders as well as of the others traders’ orders.
  13. In this phase, some arbitrageurs can also calculate Underlying Basis, which is a hypothetical calculation of Basis that ignores the arbitrageur’s own orders in relation to XJO Securities. Therefore, comparing Underlying Basis with Basis will indicate the impact of the arbitrageur’s trading on Basis. If the arbitrageur is not amending or changing orders, but Underlying Basis is moving, then Underlying Basis will provide some insight to the arbitrageur as to the direction of the market and the flow of orders placed by others. For this reason, Underlying Basis is a useful information cue which enables the arbitrageur to understand what is happening in the order book for XJO Securities. Over time, Underlying Basis will show the direction and persistence of mispricing, particularly where Basis swings between substantially positive and substantially negative values. There was significant volatility in Basis on the five Serial Expiry Days in question in this proceeding.
  14. It is appropriate at this stage to introduce the notion of Basis Decay. This is the phenomenon where, in the Pre-Open Phase, mispricing declines towards zero as the OSPA approaches. Because of this phenomenon, an experienced and competent arbitrageur would not expect any large available Basis to persist until the Opening Phase.
  15. Ideally, an arbitrageur might aim to establish an arbitrage with matched positions, in value, of SPI Futures and XJO Securities, with the XJO Securities held in their proportionate weight in the XJO. In such a position, the arbitrageur would not be exposed, in the main, to market movements, and the arbitrageur would be able to establish and carry large arbitrage transactions without taking significant market risk. In practice, however, there will be discrepancies between the value of XJO Securities held in an arbitrage position and the value of the SPI Futures held against that position.
  16. Where the SPI Futures are not matched with XJO Securities in equivalent value or in their proportionate weight in the XJO, the arbitrageur will have market exposure to the extent of the mismatch. Put simply, if the two legs of the arbitrage transaction do not create a perfect hedge, the arbitrageur will be exposed to a corresponding market risk. But the appetite for market risk (which will include the perceived size of an arbitrage profit) will vary among arbitrageurs. This may be reflected in the various strategies that different arbitrageurs might choose to pursue with a view to profiting from this form of trading, including holding an unhedged (unmatched) position in the course of actual market trading in either XJO Securities or, indeed, in SPI Futures. Index arbitrageurs can and do behave differently.
  17. A securities index arbitrage trading position in which futures are sold and securities are bought is called a Positive Index Arbitrage (or, simply, a positive arbitrage). In a positive arbitrage, the transaction will be profitable only if the calculated figure for Basis is positive. All things being equal, the more positive this figure, the more profitable the transaction should be.
  18. Conversely, a trading position in which futures are bought and securities are sold is called a Negative Index Arbitrage (or, simply, a negative arbitrage). In a negative arbitrage, the transaction will be profitable only if the calculated figure for Basis is negative. All things being equal, the more negative this figure, the more profitable the transaction should be.
  19. Other variables that can affect the profitability of a particular trade include the size of the position taken and the price at which SPI Futures can be bought or sold. The price at which SPI futures can be bought or sold is a function of the liquidity of SPI Futures in the market. This liquidity can increase or decrease substantially in a short period of time (in a matter of seconds).
  20. Generally speaking, a larger “cash” order (i.e., an order to buy or sell XJO Securities) which is hedged will be more profitable than a smaller order which is hedged, for a given level of mispricing.
  21. In the evidence, two measures have been adopted to indicate the size of orders (whether buy or sell) that were being placed by Whitebox for XJO Securities on the Serial Expiry Days in question.
  22. The first is called a SPI Equivalent. This is essentially a measure of value. It represents the volume of SPI Futures that is equivalent in value to the dollar value of the orders for XJO Securities (whether buy or sell) that Whitebox was seeking to trade. This measure was adopted by ASIC in its evidence.
  23. The second is called Shares per Hedge Equivalent or, simply, SPHE. This is essentially a measure of volume. It represents the number of futures equivalent shares that an arbitrageur is seeking to trade (whether buy or sell) by the order that is placed. For example, an order for XJO Securities equivalent to 500 SPHE means that the order is for 500 times each security’s specific shares per hedge. This measure was adopted by the defendants in their evidence.
  24. In these reasons, I refer to both measures. It has been necessary to do so because of the way in which the evidence has been presented and because of the way in which the parties, correspondingly, advanced their respective cases. In undertaking comparisons it should be borne in mind that the numerical values of each measure do not necessarily correlate.
  25. Additionally, in respect of SPHE figures utilised in the defendants’ evidence, I have approximated the precise numbers for ease of expression. These have generally been rounded to the closest 100 SPHE. In respect of the July, August and October Reductions (which are defined below), these have been rounded to the closest 10 SPHE in order to delineate each incremental reduction in SPHE occurring at each order amendment.
  26. Similarly, times in the auction timelines, which are expressed to microseconds, have been rounded to the closest second.

ASX GUIDANCE NOTE NO 32

  1. On 18 January 2008 ASX issued GN 32 – Bulk authorisation of Index Arbitrage Orders in ITS (GN 32). GN 32 is expressed to apply to orders entered to exit or unwind an index arbitrage hedge position during the Pre-Open Phase on an expiry day and any other time when opening or altering an index arbitrage hedge position with orders to be executed at market open. It applies, therefore, to securities index arbitrage orders placed in the Pre-Open Phase on any trading day, not just on expiry.
  2. GN 32 draws particular attention to a participant’s underlying responsibilities to avoid conduct that may result in either an unfair or disorderly market (Rule 14.2) or market manipulation (Rule 13.4). It makes the following general statements:
ASX recognises the importance and legitimacy of index arbitrage orders in the Australian financial markets.
ASX also recognises that the close out of an index option or index futures position on a derivatives expiry day necessitates the entire underlying basket relating to that index position to be liquidated, irrespective of price, at a time when the price of the basket equates as closely as possible to the settlement price of the related index derivative. That is, timing and execution of volume is more important to the holder of the position than price. This is achieved by unwinding the underlying basket during the opening option on expiry day.
The same strategy applies to any re-weighting of an index arbitrage position prior to expiry when the underlying is bought or sold, again irrespective of price.
However, a Participant’s rights and obligations under the Rules apply equally in respect of trading index arbitrage orders as they do when trading other types of orders. The need to convert products into cash (or vice versa) over a short period does not confer on the index arbitrageur the right to ignore the consequences or effect that its trading might have on the market. Consequently an arbitrageur may need to give special consideration to the effect of its trading in large size where the trading is also condensed into a short period.
A Participant that takes account of, and has care as to the likely effect of, its index arbitrage orders is less likely to contravene the Rules as a result of trading such orders. Where a Participant can demonstrate that the index arbitrage orders:
(a) reflect genuine index arbitrage activity; and
(b) do not appear to have been placed with an intention to manipulate or cause a disorderly market having regard to, amongst other things:
(i) the time the order was placed;
(ii) the size of the order; and
(iii) the effect of the composition of the order,
then ASX will take such matters into consideration when determining whether the Rules appeared to have been contravened.
  1. With respect to the size of index arbitrage orders, GN 32 states:
The size of an index arbitrage order can cause imbalance to the existing market price levels. Prior to effecting an index arbitrage order the Participant should assess how much volume has been traded on previous expiries, and together with more recent volumes, assess whether the size of the order can be reasonably absorbed by the market given current trading conditions.
Where the volume to be executed in the index arbitrage order is in excess of previous levels, such that the price might be expected to be materially affected, the Participant should take special precautions such as early entry of the index arbitrage orders to minimise market impact. Precautionary activity of this kind may demonstrate the Participant has exercised sufficient care to minimise the market impact of a high-volume index arbitrage order.
Conversely, materially increasing the size of the order once the order has been initially entered may indicate the Participant has not exercised sufficient care in assessing the likely impact of the order amendment. In this regard it is noted that where a Participant materially amends the volume of an existing index arbitrage order, ASX will regard the amendment as the entry of a new index arbitrage order.
  1. GN 32 sets out four explicit guidelines.
  2. Guideline 1 is:
ASX expects a Participant to be able to demonstrate that it has taken reasonable steps to verify that there is an approximately equal and opposite derivatives transaction corresponding with the index arbitrage order.
Appropriate verification may include documentary evidence such as contemporaneous records of discussions with clients prior to the placement of the orders, transaction confirmations, correspondence or directions. Such evidence must be available to allow ASX Compliance and Surveillance to readily conduct a review of bulk authorisation transactions.
  1. Guideline 2 includes the following:
The basic guideline that Participants should apply to the entry of index arbitrage orders is to enter them as early as possible. This allows other Participants and their clients to react to the orders and take advantage of the increased volumes to be traded at the open. Early entry of arbitrage orders gives others the opportunity to trade against them which may minimise price volatility at the open.
  1. Guideline 2 also sets out the time by which an order should be placed, based on the total basket size. For baskets valued at up to $100 million, the guideline stipulates that orders should be placed no later than one minute before each Group opening. For larger baskets, longer time periods are stipulated.
  2. Guideline 3 is:
A Participant which has entered an index arbitrage order in accordance with this Guidance Note should be able to amend the prices of the stocks included in the order so as to ensure that the required volume is transacted. If for example, a Participant enters an order in accordance with this Guidance Note, and, by virtue of that time at which the order was entered the order is immediately overbid or offered, the Participant is able to amend the price of its orders with a view to trading the order. Any price amendment should be made within the principles of the operation of a fair, orderly and transparent market, as waiting until immediately before the group opening may cause unreasonable market disruption.
  1. Guideline 4 is:
As noted above ASX will consider a Participant materially amending the volume of an existing index arbitrage order to have entered a new index arbitrage order. As a “new” index arbitrage order, that order would be subject to the verification, time and price guidelines above.

THE RELATIONSHIP BETWEEN WHITEBOX AND NAB

  1. In late 2009, Whitebox entered into negotiations with National Australia Bank Limited (NAB) in respect of the provision by Whitebox of securities index arbitrage trading services to NAB. In 2010, the negotiations culminated in a series of agreements by which NAB effectively outsourced its index arbitrage trading activities to Whitebox. At all times relevant to this proceeding, Mr Boshoff was Whitebox’s sole director.
  2. The contractual arrangements included an agreement styled Software Supply and Commercial and Personnel Services Agreement. This agreement was entered into on 9 March 2010 between NAB, Whitebox, Mr Boshoff, Warren Green, Simon Archer and Alexander Gamper. It was amended on 9 July 2010 and 24 September 2010. Under these various agreements, Whitebox agreed to provide, in broad terms, the services of Mr Boshoff and Mr Green in relation to trading financial products; Mr Gamper in relation to maintaining, developing and servicing the software to be employed in respect of the trading; and Mr Archer in relation business management and co-ordination. The agreements also granted licences to NAB to install and use certain identified software. These arrangements were implemented in about mid-2010. In the ordinary course, Mr Boshoff and Mr Green placed orders for trading. If Mr Boshoff or Mr Green were absent, Mr Archer would place and trade orders.
  3. By Deed dated 19 January 2011, NAB appointed Whitebox, Mr Boshoff and Mr Green as its authorised representatives under its Australian Financial Services Licence. Later (it seems sometime in 2011), NAB also appointed Mr Archer as one of its authorised representatives.
  4. During the time that securities index arbitrage trading was conducted by Whitebox and its personnel pursuant to these arrangements, UBS Securities Australia Pty Ltd (UBS), as a “trading participant” of ASX, provided direct market access services to NAB to enable Whitebox and its personnel to send orders and trading messages to ASX Trade regarding XJO Securities. JP Morgan Securities Limited (JP Morgan) provided services to NAB that enabled Whitebox and its personnel to trade and clear SPI Futures on ASX24 (the futures trading platform) for the purpose of conducting securities index arbitrage trading.
  5. NAB placed market risk limits on Whitebox in respect of its trading. One of these limits was a Gross Delta Limit. This was the absolute value limit of Whitebox’s positions in XJO Securities and SPI Futures.
  6. Other limits included a Stop Loss Daily Limit (essentially, the amount that could be lost in a single day) and a Stop Loss Rolling Month to Date (Rolling MTD) Limit (essentially, the sum of one-day losses over the preceding 20 business days). ASIC submitted that these limits were of particular relevance on the Serial Expiry Days in question because the greater the extent to which Whitebox’s trades for NAB were unhedged at any given time, the greater the risk that it could lose money for NAB and breach those limits. ASIC submitted that, in practical terms, this put a considerable constraint on Whitebox’s ability to carry an unhedged position in either XJO Securities or SPI Futures from the Pre-Open Phase into Normal Trading, even if such a strategy were considered to be rational.
  7. Under its arrangements with NAB, Whitebox derived Personnel Fees, Software Fees and Performance Payments.
  8. The Personnel Fees comprised stipulated monthly payments in respect of the provision of services rendered by Mr Boshoff, Mr Gamper, Mr Green and Mr Archer.
  9. The Software Fees were a stipulated percentage of “Annual Net Revenue” as defined in the Software Supply and Commercial and Personnel Services Agreement and its amending agreements. These fees were invoiced annually in arrears. In essence, “Annual Net Revenue” was revenue earned by NAB from Whitebox’s securities index arbitrage trading using the software licensed by Whitebox to NAB for this purpose, less Whitebox’s costs in providing the services (other than salaries, contractor fees and other remuneration or benefits paid by NAB in respect of Whitebox’s personnel).
  10. The Performance Payments were required to be made under a “Performance Plan” based on the performance of Whitebox’s Traders, essentially Mr Boshoff and Mr Green. The payments were calculated by reference to an “Incentive Multiple” which was, initially, 12.5% for Mr Boshoff and 2.5% for Mr Green, although the multiple could be varied as between Mr Boshoff and Mr Green, provided the total Incentive Multiple was 15%. The payment of part of each year’s Performance Payment (25%) was deferred and paid (subject to certain conditions) in the following financial year.
  11. ASIC submitted that, by reasons of these various financial arrangements, Whitebox had a direct financial interest in maximising the revenue it generated for NAB, as well as an indirect interest in maintaining NAB’s commercial relationship with Whitebox and its personnel. ASIC submitted that Mr Boshoff had an indirect financial interest in Whitebox maximising the revenue it generated for NAB because of his 100% shareholding in a company called H B Capital Aust. Pty. Limited, which in turn held dividend shares and 100% of the ordinary shares in Whitebox. In the period 9 March 2010 to 31 December 2012, Whitebox earned, and was paid, a total of $4,639,952.63 from NAB in respect of Personnel Fees, Software Fees and Performance Payments.
  12. On 21 December 2012, NAB terminated its contractual arrangements with Whitebox. The reason for that termination can be stated shortly. On 18 October 2012 (one of the trading days in question in this proceeding), Whitebox placed orders which resulted in a spike in the price for Group 1 securities. This trading is described in more detail below: see [238] – [256]. NAB considered Whitebox’s trading to be a breach of its express warranty in the Software Supply and Commercial and Personnel Services Agreement (as amended) that it would perform its contractual obligations with due care and skill and in a proper and professional manner. NAB also considered this to be a breach by Whitebox of an implied term of the agreement that it would perform its obligations with reasonable care.
  13. Following that price spike, NAB (on 19 October 2012) directed Whitebox to cease all trading on its account. On the same day, UBS sent a letter to NAB confirming that, by mutual agreement, NAB would no longer trade through UBS’s direct market access services, although this was revised temporarily on 23 October 2012 to allow NAB to reduce and close out the positions it then held.
  14. In early December 2012, NAB closed its Index Arbitrage Desk. By 20 December 2012, the positions held by that desk had been reduced to zero. On 21 December 2012, at NAB’s request, JP Morgan disabled the accounts through which Whitebox personnel had traded SPI Futures on ASX24.
  15. For the period 1 October 2011 to 30 September 2012, Whitebox earned Performance Payments of $521,330.00 (excluding GST) (i.e., in addition to those amounts referenced above). NAB did not pay this sum to Whitebox as it considered the sum to have been forfeited under the Performance Plan by reason of Whitebox’s alleged breaches of contract/duty.

WHITEBOX’S STATED POSITION ON ITS TRADING STRATEGY

  1. In a document prepared in 2009 by Whitebox’s index arbitrage team at the time that negotiations were underway between Whitebox and NAB to provide securities index arbitrage trading (and, I infer, used for the purpose of those negotiations), Whitebox referred to “our (t)rading strategy” as “at all times perfectly hedged (Futures and ETF’s to Cash)” with “no delta to generate PNL”. The document emphasised that Whitebox had traded in Australia “with a very low risk profile”. When dealing with market risk, the document stated:
... In short WB strategy is one of NO delta, thus making us immune of [sic] Market movements both intraday and those from offshore and overnight.
  1. Following the price spike caused by Whitebox’s trading on 18 October 2012, ASIC was provided with a document that sought to explain Whitebox’s trading strategy on that day. This document was prepared with Mr Boshoff’s and Mr Archer’s assistance, and confirmed by them as being correct. This document was referred to in the evidence as the Schedule A document.
  2. In a section headed “Overview of ASX SPI 200 Futures /ASX SPI 200 financial product arbitrage trading”, the Schedule A document states (omitting footnotes):
1.1 National Australia Bank Limited (“NAB”) engages in proprietary trading in ASX SPI 200 Futures (“Futures”) and the financial products that comprise the ASX SPI 200 Index (“Cash”) so as to profit from mispricings between the relative market prices of Futures and Cash, taking into account the financing costs of carrying Cash until the expiry of the relevant Future and the likely effect of any future dividend payments.
1.2 Known as “program trading”, this is a form of index arbitrage trading and involves NAB’s index arbitrage team using proprietary technology to identify mispricings between Futures and Cash. Specifically, the team runs algorithms to determine the “fair value” of a weighted (yet imperfect) basket of the underlying financial products that comprise Cash against the strike price of the relevant future contracts at that time. These algorithms take into account the funding costs of holding Cash to expiry, and model the impact of dividends that will be received or required to be passed though by NAB during the period. Where the strike price of the futures contract is assessed as being equal to the “fair value” of the underlying weighted Cash basket, no mispricing exists and NAB will not trade. The index arbitrage team continually calculates this “fair value” based on live price data which is marked to market.
1.3 Where this process identifies that Futures are more expensive than the fair value of the underlying weighted Cash basket, the index arbitrage team will sell Futures and purchase Cash. Where Futures are cheaper than the fair value of the underlying weighted Cash basket, Futures are purchased and Cash is sold (by short selling if required). The number of Futures and Cash bought and sold is guided by reference to an algorithm which estimates, given the level of mispricing, how much Cash can be bought or sold before the mispricing is spent and there is no more opportunity to profit on the arbitrage.
  1. The reference in these passages to NAB’s index arbitrage team is to the Whitebox traders.
  2. The document goes on to describe Whitebox’s trading strategy, including by reference to the document prepared in 2009:
1.4 The index arbitrage team’s strategy is to capitalise on mispricing without exposing themselves to either market (delta) or specific risk (risk associated with specific stocks). The reference to delta in this context is the misbalance between the nominal value of the Cash and Futures position stated in Futures contracts. To address the delta risk, the index arbitrage team hedges its Cash position through the buying or selling of Futures. Tab 1 contains a more detailed description of the risk mitigations built into the index arbitrage team’s approach. This description was prepared in 2009 by the index arbitrage team at the time they were negotiating with NAB to provide arbitrage trading services.
  1. The document then describes how this strategy is, in general terms, implemented:
1.5 The index arbitrage team begin reviewing the developing indicative auction price for Cash from approximately 8am every trading day as against the last closing price of Futures on Sycom, (available at 7am on trading days other than Monday) and analysis of overseas markets and other external factors that may impact the opening price of Futures. Based on that analysis, the arbitrage trading team places and amends orders in the Pre-Opening Phase of the Cash market in anticipation of potential mispricing at Market Open and orders start to be placed from 8am. The index arbitrage team’s strategy with the early orders is to submit Cash orders or bids based on the perception of mispricing being seen at that time as if such mispricing was to persist to unwrap. As the perception of mispricing changes throughout the morning (with other activity in the auction) the orders are amended (and in some cases from Sell to Buy orders and vice versa) as more information becomes available. From 9.40am, indicative auction prices for Futures become available through pre-open on ASX24, and at that time, the central order book for the Cash market starts reflecting the majority of orders that is likely to be present in the opening auction of the Cash market. Therefore, following that time the accuracy of information relating to mispricing improves, and the ability of the index arbitrage team to predict mispricing at Market Open increases, and following the opening of the Futures Market on ASX24 at 9.50am, it becomes possible for the index arbitrage team to establish a strategy for execution. However, due to the transient nature of mispricing, there is a risk that the index arbitrage team may buy or sell Futures too early and find that the mispricings disappear before the Opening rotation, leaving them with a delta and a possible loss on those Futures trades. Accordingly, the purchase or sale of Futures to secure the “hedge” is executed as close possible to when the Cash can actually be bought or sold.
  1. In later paragraphs of the document, dealing specifically with the trading on 18 October 2012, Whitebox referred (in para 4.9) to “the index arbitrage team’s usual practice” of transacting, as far as possible, futures:
... as closely as possible to the time at which it transacts in Cash to minimise risks of market movement before positions can be offset.
  1. In the same vein, another passage in the document (para 4.10, dealing with the disparity between Whitebox’s cash and futures positions close to the commencement of the OSPA on 18 October 2012) referred to the need for Whitebox to have acquired 471 SPI Futures:
... to have its hedge on the then Cash offer, in place before the opening of Group 1 on the ASX at 10.00.00AM +/- 15 seconds.
  1. The Schedule A document then explains:
4.12 Given these developments, if the index arbitrage team had maintained the Cash offer at its then size, it would have resulted in NAB obtaining a Cash position following the Market Open that was unhedged and carrying severe delta risk. This would have been contrary to the index arbitrage team’s trading strategy.
  1. I accept that the “This” in the last sentence of the quote refers not simply to carrying an unhedged position into Normal Trading but an unhedged position carrying “severe” delta risk.
  2. In closing submissions, the defendants sought to explain or qualify various passages in the Schedule A document, including by reference to the 2009 document. In light of various email chains in evidence, the defendants submitted that the Schedule A document had been prepared rapidly and in circumstances of urgency to answer a notice served under s 33 of the Australian Securities and Investments Commission Act 2001 (Cth). They argued that Mr Boshoff had either two or five minutes to review the document before providing his confirmation of its accuracy. They pointed to various omissions and misdescriptions of a relatively minor nature to suggest that Mr Boshoff did not read a draft of the Schedule A document carefully before providing his confirmation.
  3. The defendants also submitted that a number of passages in the Schedule A document—in particular, para 1.4 quoted above—should be understood as referring to intraday trading, when both the ASX and ASX24 were open; not to arbitrage activity in the Pre-Open Phase. This was because para 1.4 refers to a strategy of capitalising on mispricing without exposure to market (delta) risk or specific risk (risk associated with specific stocks). The defendants submitted that it is impossible to avoid risk in the Pre-Open Phase because it is impossible to trade both the cash and futures legs of an index arbitrage simultaneously. This trading can only be done when the cash and futures markets are open. In the meantime, an imbalance between the cash and futures positions cannot be avoided.
  4. The defendants submitted that the statements made in paras 4.9, 4.10 and 4.12 of the Schedule A document were dealing specifically with Whitebox’s order activity on 18 October 2012, and not more generally in the course of its index arbitrage trading. I accept that this is the context in which those statements were made. The defendants submitted that the statement in para 4.9 about transacting the futures leg of the arbitrage as closely as possible to transacting the cash leg should not be understood as being directed to the time of Whitebox completing its trading on the futures leg. They submitted that completing trading on the futures leg is a function of the size of the mispricing that might have been captured, and the conditions in the futures market, on the day. They argued that if significant mispricing had been captured, the risk involved in deferring part of the futures leg until after market opening would be acceptable.
  5. The defendants submitted that the statement in para 4.10 about the “need” for Whitebox to acquire 471 SPI Futures before the opening of the Group 1 Rotation on 18 October 2012 (quoted at [112] above), should not be read as an absolute statement. The defendants submitted that it was not necessary to attempt to trade, Group by Group, a number of SPI Futures corresponding in value to the percentage of the ASX 200 represented by each Group. The defendants supported this submission by reference to data on Whitebox’s futures trading on Expiry Days in 2012. The defendants submitted, and I accept, that the data shows that Whitebox did not adopt a rigid, formulaic approach of matching futures with the proportion of the cash leg as it traded. The defendants also submitted that the data shows that Whitebox did not have a rigid approach of precisely matching the futures and cash legs even as at the time of the Group 5 Rotation. As a matter of fact, this is true, although on a number of days the difference between the futures leg and the cash leg is either non-existent or small. On other days, the difference is more marked. However, overall, the data shows that, in practice, the matching of the two legs at particular times was less rigid than the statement in para 4.10 suggests: see Appendix D to the defendants’ closing submissions.
  6. The Schedule A document was prepared for submission to ASIC on a matter of considerable importance, namely to address and explain Whitebox’s arbitrage activity on 18 October 2012. It is written in plain terms with, in my view, a plain meaning. I do not think it calls for the nuanced reading advocated by the defendants in submissions. If the document does not mean what it says, or did not accurately represent Whitebox’s position at the time it was prepared and given to ASIC, then the defendants could have called, and should have called, evidence on that matter. There is nothing before me to suggest that at any time since 2012, the defendants sought to qualify the deliberate statements made in the document, other than by submissions in this proceeding.
  7. I am particularly unmoved by the suggestion that Mr Boshoff did not carefully read or did not carefully consider the draft of the Schedule A document that was submitted for his confirmation. The document was prepared with his assistance. Given the circumstances in which, and the occasion for which, it was prepared, I am not prepared to act on the mere suggestion that, at the time, Mr Boshoff might not have given the document the attention it so clearly deserved. If Mr Boshoff acted carelessly in that regard, then that circumstance calls for evidence, not simply a suggestion made argumentatively in submissions.
  8. The passages I have quoted above stand as Whitebox’s general and usual trading strategy. They are consistent with a strategy that was directed to avoiding market risk due to holding an unhedged position. The passages indicate that the manner in which risk was to be minimised was to ensure that there was no significant time-gap between buying/selling XJO Securities and selling/buying SPI Futures. It was also to be managed by not carrying an unhedged position into Normal Trading. Read in context, the Schedule A document was seeking to explain that, as matters transpired, Whitebox’s trading on 18 October 2012 was, overall, in accordance with its general and usual trading strategy. However, the evidence shows that this general and usual trading strategy was not followed invariably on other days. As I will come to explain, Whitebox’s appetite for risk was not as adverse as the Schedule A document and, indeed, the 2009 document suggest.
  9. ASIC submitted that the fact that Whitebox sought to trade XJO Securities and SPI Futures contemporaneously and in a manner that minimised risk was important for two reasons. First, it informed the question whether there was an intention on the Serial Expiry Days in question to trade the large orders for XJO Securities which Whitebox had placed (discussed in more detail below) in circumstances where (as ASIC contended) Whitebox could not have expected to trade SPI Futures profitably in a manner consistent with its stated strategy of minimising market risk. Secondly, the rate at which, and manner in which, Whitebox traded SPI Futures in the Pre-Open Phase and the OSPA enables the Court to infer the amount of XJO Securities Whitebox actually intended to trade during the OSPA. These submissions are related to (what ASIC described as) the second and third “pillars” of its case: see below.
  10. It is convenient at this point to also refer to certain email correspondence between Mr Boshoff and Mr Stevenson from NAB in October 2012; in particular an email on 26 October 2012 when Mr Boshoff was responding to a query raised by Mr Stevenson (who was then the Head of Market Risk Oversight – Rates, Credit & Equities) about Whitebox’s trading on 18 October 2012. In that email, Mr Boshoff said:
Our strategy during the early lead up of the Auction is to submit orders that can possibly lead to a mispricing trade if mispricing persists to the unwrap (an obviously presuming we can hedge the futures). When there is as much volatility in the basis as we saw from 9:40am and so much of it we are not overly concerned to balance the market precisely with our orders (the volatility makes that impossible) but rather to make sure we respond to the big basis moves. Referring to table provided by Mallesons the amend that we did at 9:48:14 from nominal 2000 futures was a response to basis spiking up to 17bps. Our amendment to 1900 nominal futures brought basis back to 7.2 bps and our system indicated that an amendment of a further 100 lots would take it to -14bps. Our experience is that as the market opening approaches basis generally becomes less volatile and we are then able to accurately calibrate it to provide us with enough mispricing to commercially justify our basket. Please note in table 1 the ‘No Action’ entries during 9:59am.

(Errors in original; emphasis added.)

  1. As the defendants pointed out in submissions, the figures quoted in the email for Basis do not appear to be accurate. They certainly do not correlate with the figures given in the auction time line for 18 October 2012, which the parties accepted is accurate.
  2. However, the matter of present significance is Mr Boshoff’s statement that, as the market opening approaches, Basis generally becomes less volatile and “we are then able to accurately calibrate it to provide us with enough mispricing to commercially justify our basket”.
  3. The defendants submitted that the word “calibrate” in this passage should be understood as referring to the size of Whitebox’s cash orders in SPHE, not to Basis as such. The defendants submitted that while Whitebox was able to set (calibrate) its cash order size, it was not, and is not, possible for any arbitrageur to set (calibrate) Basis, given that Basis is a function of the orders in the Order Book at any one time. Thus, it would have been impossible for Whitebox to predict how other orders in the book would change, either in response to Whitebox’s orders or otherwise, so as to affect Basis. I note in this connection that the amendment to which Mr Boshoff referred was made at 9:48:14, significantly more than eleven minutes before the earliest time that the Group 1 Rotation could occur—a very lengthy period in the scheme of things, in which the order activity of other arbitrageurs and traders could radically change the size and even direction of Basis. I accept that, in these circumstances, this would be a strange use of the word “calibrate”.
  4. To support their submission, the defendants also drew attention to the words: “to provide us with enough mispricing to commercially justify our basket”. The defendants submitted that Whitebox could not “commercially justify” a loss-making basket or one that would require it to carry a disproportionate risk on the futures leg. This was something Whitebox could “calibrate” by amending orders closer to the time of the Rotation for each Group, given that Basis tends to decay as market opening approaches.
  5. As a matter of syntax, it is difficult to read the word “calibrate” in the quoted passage in Mr Boshoff’s email as referring to anything other than Basis. However, if Mr Boshoff did intend to convey that meaning, I accept the defendants’ submission that Whitebox could not “calibrate” Basis. No doubt Whitebox’s orders could affect Basis, but the extent to which this would be so depended on the order activity of other arbitrageurs and traders, a matter over which Whitebox had no foreknowledge or control.
  6. The defendants submitted that there is nothing improper in amending orders as the OSPA approaches, even very close to the OSPA, so as to avoid a loss-making position or to reduce risk or to increase the prospect of profiting from an order. They accepted as uncontroversial the fact that an amendment to an order can have an immediate effect on the IMP. But, they submitted, to do so is entirely consistent with the rationality of the strategies accepted by Mr Morgan in the witness box. The defendants emphasised that ASIC’s case is that Whitebox placed earlier orders or amendments without even a contingent intention to execute them, for the purpose of manipulating the Opening Prices of the securities concerned.
  7. I accept the defendants’ submission that there is nothing improper in amending orders as the OSPA approaches, even very close to the OSPA, so as to avoid a loss-making position or to reduce risk or to increase the prospect of profiting from an order. The conduct which ASIC alleges is of an entirely different character.

THE IMPUGNED TRADING

ASIC’s identification of the impugned trading and the defendants’ strategy

  1. In this section of my reasons, I summarise ASIC’s identification, in its written closing submissions, of the impugned trading on each of the five Serial Expiry Days in question. ASIC’s written submissions were supplemented by charts (used in opening), and tables which nominated the particular orders that were said to give rise to the allegedly contravening conduct.
  2. ASIC submitted that Whitebox’s trading on these days was characterised by a “deliberate, clear and substantially consistent strategy” which enables the Court to interpret the  Whitebox trading  and to assess the defendants’ intentions when placing (or causing Whitebox to place) the orders the subject of this proceeding.
  3. In general terms, this strategy was to place very large orders for XJO Securities in the Pre-Open Phase which were subsequently cancelled or significantly reduced in volume in the period shortly prior to the relevant rotation of the OSPA in which the orders were due to trade, without permitting the market time to react. ASIC’s case is that, by placing these particular orders, and cancelling or reducing them when it did, Whitebox and Mr Boshoff were intending to manipulate the Opening Price for the XJO Securities to create mispricing that could then be exploited for profit. In short, there was no genuine intention to trade the large orders.
  4. ASIC submitted that, consistently with this strategy, the defendants never sought to trade SPI Futures at either the rate or with the “aggression” that would have been needed to match these very large orders of XJO securities if carrying out a perfectly hedged index arbitrage. According to ASIC, the rate at which Whitebox did trade SPI Futures was consistent with trading the much smaller SPI Futures position it in fact traded on each of the Serial Expiry Days in question. In this connection, ASIC singled out for particular mention Whitebox’s trading on 19 April 2012 when Whitebox did not commence placing orders for and trading SPI Futures until 9:59:22. This allowed only 23 seconds within which to trade the approximately 306 SPI Futures necessary to match, by the earliest commencement of the Group 1 Rotation, the Buy Orders for Group 1 securities.
  5. ASIC submitted that, consistently with its stated trading strategy, on the Serial Expiry Days in question, Whitebox placed and traded index arbitrage orders in a way that sought to minimise its exposure to the risks that might arise from a mismatch between its positions in XJO Securities and SPI Futures. ASIC submitted that, with the exception of subsequent trading undertaken on 19 April 2012, Whitebox did not trade SPI Futures after 10:08:45, the earliest possible time for the commencement of the last Rotation of the OSPA. The SPI Futures that Whitebox traded in the period prior to 10:08:45 substantially matched the value of Whitebox’s position in XJO Securities in the OSPA. Exceptionally, on 19 April 2012, Whitebox in fact acquired an additional position in XJO Securities after the commencement of the OSPA. In order to hedge this position, it sold additional SPI Futures.
  6. ASIC submitted further that the defendants’ strategy was to “capture” Basis and “mask” Underlying Basis (i.e., the potential profit that would exist in the absence of Whitebox’s orders) from the rest of the market. ASIC explained this aspect of its case by reference to the phenomenon of Basis Decay. ASIC said that, because of the size of the orders that Whitebox had in the market, there was a large difference between the mispricing that would have existed absent Whitebox’s order (Underlying Basis) and the mispricing that existed with Whitebox’s orders (Basis). This difference was only visible to Whitebox and, according to ASIC, was thereby “masked” from the rest of the market. ASIC argued that it was this difference that provided the opportunity for Whitebox to profit by removing order volume. Relatedly, it explained the rationale for Whitebox placing large orders for XJO Securities and then cancelling them or reducing their volume substantially shortly before the orders were due to trade in the OSPA.
  7. ASIC submitted that the defendants were “keenly aware” of the impact that Whitebox’s placement, amendment and cancellation of orders for XJO Securities would have on the market—particularly the effect of withdrawing order volume. ASIC submitted that Whitebox adjusted its orders to affect Basis which, it argued, “necessarily bespoke an intent to affect the prices of XJO Securities in order to alter the level of the XJO”.
  8. In this connection, ASIC pointed to Mr Boshoff’s email to Mr Stevenson on 26 October 2012. ASIC submitted that this email—in particular, Mr Boshoff’s reference to “calibrating” Basis—showed that Whitebox was regularly monitoring the effect that removal of its order volume would have on Basis and the likely profitability of the index arbitrage trading it was conducting. ASIC submitted that the email makes clear that the defendants were “nakedly attempting to manipulate Basis” for the purpose of making the trading profitable, rather than “merely attempting to profit from mispricing that existed in the market”. According to ASIC, this puts in context the large orders placed in the period leading up to the OSPA. Once again, these orders had not been placed with an expectation that they would be traded. On ASIC’s case, these orders were the means by which the defendants could calibrate Basis to create the degree of mispricing necessary to “commercially justify” Whitebox’s basket.
  9. ASIC submitted that, as part of this strategy, the defendants determined the XJO position that Whitebox would trade by the time that the Rotation for Group 1 commenced. The ultimate orders for XJO Securities in the last four Rotations substantially matched the orders, in SPHE, for the Group 1 Rotation. In other words, based on their weightings in the calculation of the XJO, these orders were proportionate in volume to the Group 1 securities that were traded, so as to ensure that the overall basket of securities replicated the XJO. ASIC submitted that this strategy was “conceptually consistent” with Whitebox’s stated position on risk.
  10. Concomitantly, ASIC submitted that the defendants delayed cancelling or reducing Whitebox’s orders in the last four Rotations even though, according to ASIC, they must have known that those orders were not going to be traded either at all or for their full volume. ASIC submitted, once again, that this aspect of the strategy was to ensure that the market did not have sufficient time to react to Whitebox’s cancellations or reductions. In turn, this ensured that there was likely to be an impact on the prices of the XJO Securities set during the OSPA.
  11. ASIC contended that this overall strategy was tested and evolved over the course of the Serial Expiry Days in question. According to ASIC, this is reflected in the “almost uniform increase in the XJO position” that Whitebox traded on those days. I will refer to this pattern in more detail below. It is sufficient for present purposes to note that ASIC described this pattern as a “steady, uniform and consistent increase in the XJO position” that was “divorced from the trading conditions on any particular day”. ASIC submitted that this pattern illustrates the deliberate nature of the defendants’ strategy and that the defendants were aware of the size of the position they intended to trade prior to  Whitebox trading  on a particular day.
  12. Aspect of the strategy I have just described are reflected in what ASIC described as the eight “pillars” of its case: see below at [271] – [284].

19 April 2012

  1. As at the commencement of trading on 19 April 2012, Whitebox held a total position of approximately $1.736 billion. That position was made up of SPI Futures worth approximately $868 million, XJO Securities worth approximately $863 million and Exchange Traded Funds (ETFs) worth approximately $5 million. This position was held against Whitebox’s Gross Delta Limit of $2 billion (therefore, 87% of the limit had been utilised).
  2. On 19 April 2012, Whitebox traded a positive arbitrage. ASIC relies on the following events:
(a) In the period from 09:54:49 to 09:55:43, Whitebox placed two sets of orders to acquire XJO Securities that in aggregate were equivalent in value to 1171 June SPI Futures (Buy Orders). These were the Quarterly SPI Futures that were the next to expire after 19 April 2012.
(b) In the period from 09:59:38 to 09:59:40, Whitebox placed or caused the placement of a further set of orders to acquire XJO Securities that in aggregate were equivalent in value to 100.3 June SPI Futures (Subsequent Buy Orders).
(c) Shortly before each Rotation of the OSPA, Whitebox cancelled or caused the cancellation of the Buy Orders that were due to trade in that OSPA rotation (Cancellations).
(d) Whitebox then permitted the Subsequent Buy Orders (other than for securities with ASX Ticker Codes AIO, WPL and WTF) to trade in the OSPA.
(e) In the time between 9:59:22 and the start of the OSPA Rotation for Group 5 (10:08:45), Whitebox placed orders to sell June SPI Futures and sold 99 June SPI Futures.
  1. At the conclusion of the OSPA on 19 April 2012, Whitebox had traded an arbitrage position constituting offsetting and substantially matching values of June SPI Futures and XJO Securities bought pursuant to the Subsequent Buy Orders. The original Buy Orders had been abandoned by Whitebox shortly before the relevant Rotation of the OSPA in which those orders were due to trade.

17 May 2012

  1. As at the commencement of trading on 17 May 2012, Whitebox held a total position of approximately $2.527 billion. That position was made up of SPI Futures worth approximately $1.263 billion, XJO Securities worth approximately $1.259 billion and ETFs worth approximately $4.5 million. This position was held against Whitebox’s Gross Delta Limit of $3 billion (therefore, 84% of the limit had been utilised).
  2. On 17 May 2012, Whitebox traded a negative arbitrage. ASIC relies on the following events:
(a) At approximately 9:56:21, Whitebox placed on ASX a set of orders to sell XJO Securities (May Sell Orders).
(b) In the period 09:58:26 to 10:06:22, Whitebox amended the May Sell Orders by placing sets of order amendments to sell XJO Securities that were equivalent in aggregate value, initially, to 977.8 June SPI Futures (May Amended Sell Orders).
(c) Shortly before each Rotation of the OSPA, Whitebox placed a set of order amendments in substantially reduced quantities for each of the May Amended Sell Orders due to trade in that OSPA Rotation, which were in aggregate equivalent in value to 199.2 June SPI Futures (May Reductions).
(d) Whitebox then permitted the May Reductions for each group of XJO Securities (other than for the security with ASX Ticker Code SPN) to trade in the OSPA.
(e) In the time between 9:58:31 and the start of the OSPA Rotation for Group 5 (10:08:45), Whitebox placed orders to buy June SPI Futures and bought 200 June SPI Futures.
  1. At the conclusion of the OSPA on 17 May 2012, Whitebox had traded an arbitrage position constituting offsetting and substantially matching values of June SPI Futures and XJO Securities sold pursuant to the May Reductions. ASIC says that, like the position in April, very large volumes of orders for XJO Securities that had been the subject of the May Amended Sell Orders were abandoned shortly prior to the Rotation of the OSPA in which those orders were due to trade.

19 July 2012

  1. As at the commencement of trading on 19 July 2012, Whitebox held a total position of approximately $870 million. That position was made up of SPI Futures worth approximately $435 million and XJO Securities also worth approximately $435 million. This position was held against Whitebox’s Gross Delta Limit of $3 billion (therefore, 29% of the limit had been utilised).
  2. On 19 July 2012, Whitebox traded a negative arbitrage. ASIC relies on the following events:
(a) At approximately 7:49:26, Whitebox placed a set of orders to sell XJO Securities (July Sell Orders).
(b) In the period 9:58:13 to 10:06:23, Whitebox amended the July Sell Orders by placing sets of order amendments to sell XJO Securities that in aggregate were equivalent in value, initially, to 1514.3 September SPI Futures (July Amended Sell Orders). These were the Quarterly SPI Futures that were the next to expire after 19 July 2012.
(c) Shortly before each Rotation of the OSPA, Whitebox placed sets of order amendments in substantially reduced quantities for each of the July Amended Sell Orders due to trade in that Rotation of the OSPA, which were in aggregate equivalent in value, ultimately, to 303.5 September SPI Futures (July Reductions).
(d) Whitebox then permitted the July Reductions, or orders for substantially the same quantities of XJO Securities that were the subject of the July Reductions (other than for securities with ASX Ticker Codes AGO, BWP, MDL, OGC, RSG, STO and WPL), to trade in the OSPA.
(e) In the time between 9:57:04 and the start of the OSPA Rotation for Group 5 (10:08:45), Whitebox placed orders to buy September SPI Futures and bought 300 September SPI Futures.
  1. ASIC says that this trading presented a result substantially in conformity with that which had occurred on the previous Serial Expiry Days. Whitebox had traded an arbitrage position constituting offsetting and substantially matching values of September SPI Futures and XJO Securities. ASIC says that the very much larger volumes of orders for XJO Securities that had been the subject of the July Amended Sell Orders were abandoned shortly prior to the Rotation of the OSPA in which those orders were due to trade.

16 August 2012

  1. As at the commencement of trading on 16 August 2012, Whitebox held a total position of approximately $1.441 billion. That position was made up of SPI Futures worth approximately $722 million, XJO Securities worth approximately $695 million and ETFs worth approximately $24 million. This position was held against Whitebox’s Gross Delta Limit of $2 billion (therefore, 72% of that limit had been utilised).
  2. On 16 August 2012, Whitebox traded a negative arbitrage. ASIC relies on the following events:
(a) At approximately 8:10:03, Whitebox placed a set of orders to sell XJO Securities (August Sell Orders).
(b) In the period between 9:57:00 am and 10:06:29 am, Whitebox amended the August Sell Orders by placing sets of order amendments to sell XJO Securities that were in aggregate equivalent in value, initially, to 1514.5 September SPI Futures (August Amended Sell Orders).
(c) Shortly before each Rotation of the OSPA, Whitebox placed sets of order amendments in substantially reduced quantities for each of the August Amended Sell Orders due to trade in that Rotation of the OSPA, which were in aggregate equivalent in value to, ultimately, 404.3 September SPI Futures (August Reductions).
(d) Whitebox then permitted orders for substantially the same quantities of XJO Securities as were the subject of the last set of the August Reductions for each group of XJO Securities (other than securities with ASX Ticker Codes AMC, DXS, SWM and TIS) to trade in the OSPA.
(e) In the time between 9:53:36 and the start of the OSPA Rotation for Group 5 (10:08:45), Whitebox placed orders to buy September SPI Futures (which were, on 16 August 2012, the next Quarterly SPI Futures) and bought 400 September SPI Futures.
  1. ASIC says that, similarly to the previous Serial Expiry Days, at the conclusion of the OSPA Whitebox had traded an arbitrage position constituting offsetting and substantially matching values of September SPI Futures and XJO Securities. Again, the much larger volumes of orders for XJO Securities that had been the subject of the August Amended Sell Orders were abandoned shortly prior to the Rotation of the OSPA in which those orders were due to trade.

18 October 2012

  1. As at the commencement of trading on 18 October 2012, Whitebox held a total position of approximately $1.363 billion. That position was made up of SPI Futures worth approximately $682 million (“short” $905 million for December SPI Futures and “long” $223 for October SPI Futures), XJO Securities worth approximately $648 million and ETFs worth approximately $32 million. This position was held against Whitebox’s Gross Delta Limit of $2 billion (therefore, 68% of that limit had been utilised).
  2. On 18 October 2012, NAB’s absolute position in SPI Futures and XJO Securities was due to increase by approximately $400 million, by reason of the expiry on that date of the October SPI Futures it held and the necessary purchase of XJO Securities to close out the matching hedge against those SPI Futures.
  3. On 18 October 2012, Whitebox traded a negative arbitrage. ASIC relies on the following events:
(a) At approximately 8:03:04, Whitebox placed a set of orders to sell XJO Securities (October Sell Orders).
(b) In the period between 9:57:15 and 10:05:52, Whitebox amended the October Sell Orders by placing sets of order amendments to sell XJO Securities that were in aggregate equivalent in value, initially, to 1810.3 December SPI Futures (October Amended Sell Orders). These were the Quarterly SPI Futures that were next to expire after 18 October 2012.
(c) Shortly before each Rotation of the OSPA, Whitebox placed sets of order amendments in substantially reduced quantities for each of the October Amended Sell Orders due to trade in that Rotation of the OSPA, which were in aggregate equivalent in value, ultimately, to 503.3 December SPI Futures (October Reductions).
(d) Whitebox then permitted the October Reductions or orders for substantially the same quantities of XJO Securities as were the subject of the last set of the October Reductions for each group of XJO Securities (other than for securities with ASX Ticker Codes CQR, ENV, MGR and WPL), to trade in the OSPA.
(e) In the time between 9:55:02 and the start of the OSPA Rotation for Group 5 (10:08:45), Whitebox placed orders to buy December SPI Futures and bought 504 December SPI Futures.
  1. ASIC says that, as with the previous Serial Expiry Days, Whitebox’s trading in October was marked by the earlier placement of very large orders, the volumes of which were subsequently substantially reduced shortly prior to the Rotation of the OSPA in which they were due to trade. At the conclusion of the OSPA, Whitebox had traded an arbitrage position constituting offsetting and substantially matching values of December SPI Futures and XJO Securities. ASIC says that, again, very large volumes of the October Amended Sell Orders were abandoned on the “precipice” of the OSPA.

Some matters of definition in relation to the impugned trading

  1. In its written closing submissions, ASIC referred collectively to the May Amended Sell Orders, July Amended Sell Orders, August Amended Sell Orders and October Amended Sell orders as the Amended Sell Orders, and to the May Reductions, July Reductions, August Reductions and October Reductions as the Reductions.
  2. ASIC submitted that, for the purposes of this case, all the Reductions are relevant for the second and third categories of the alleged market-based contraventions (the s 1041B “price for trading” and the s 1041A “artificial price for trading” cases).
  3. ASIC’s position in relation to the first category of alleged market-based contraventions (the s 1041B “market for” case) is more complicated. On the Serial Expiry Days in July, August and October 2012, the Reductions were made, in large part, incrementally rather than in one go. This was graphically represented in charts that ASIC had prepared. While the orders that made up the final set of Reductions (or orders for substantially the same quantities of XJO Securities) did in fact trade in the market, none of the incremental Reductions that were made in the lead up to the final set of Reductions were traded.
  4. ASIC submitted that, for the purposes of the “market for” case, the Reductions that are relevant for the first category of alleged market-based contraventions are the Reductions for C-Z Securities (other than the last set of those Reductions) in July, the Reductions (other than the last set of those Reductions) in August, and the Reductions for C-Z Securities (other than the last set of those Reductions) in October. ASIC referred to these as the Incremental Reductions.

The auction time lines

Introduction

  1. The auction time lines adduced in evidence by the defendants provide a far more complex picture of Whitebox’s trading on the Serial Expiry Days in question. Indeed, as the defendants contended, one cannot simply isolate the impugned orders and reach conclusions about them without having regard to the pattern of trading that took place on those days. The auction time lines show the activity of other arbitrageurs, but not the activity of other traders.
  2. The following account of Whitebox’s order activity is based on my own reading of the auction time lines, assisted by the description given in the defendants’ written closing submissions. When I refer to Whitebox having entered orders, my reference is to Whitebox having entered orders across all Groups of XJO Securities, unless I state otherwise.
  3. Further, in the auction time lines, the orders for XJO Securities are expressed as SPHE, not as SPI Equivalents as in the previous section of these reasons dealing with ASIC’s identification of the impugned conduct. This needs to be borne in mind when comparing this section with the previous section because the values expressed in this section do not neatly correlate with the values expressed in the previous section.

19 April 2012

  1. On 19 April 2012, Whitebox entered its first orders for XJO Securities at 9:40:43. These were buy orders for approximately 500 SPHE. At that time, Fast Basis and Underlying Basis were 431.3 and 434.0, respectively. In other words, mispricing was greater than 4%.
  2. After entering and cancelling further buy orders, Whitebox entered new buy orders for approximately 2000 SPHE at 9:42:46. At that time, mispricing was approximately 4.29%.
  3. By 9:48:27, Fast Basis had reduced to 137.2 and Underlying Basis had reduced to 255.9. Whitebox cancelled the buy orders. However, within approximately five seconds at 9:48:32, Whitebox again entered new buy orders for approximately 2000 SPHE. At this time, Fast Basis was 267.7 and Underlying Basis was 268.00.
  4. The entry of these buy baskets was consistent with a contemplated positive arbitrage, particularly having regard to the level and direction of mispricing indicated in the period 9:40:43 to 9:48:32. But, by 9:53:18, Fast Basis had moved negatively to -4.0, indicating that a positive arbitrage for Whitebox would have been potentially unprofitable. This movement in Basis may have been in response to the placement of buy orders by other traders for certain buckets of securities.
  5. At 9:53:50, Whitebox cancelled its buy orders. But, in the period between the entry of those orders and their execution, Fast Basis moved positively (towards 1%) and, by 9:53:52, both Fast Basis and Underlying Basis were 222.6 (i.e., just above 2%), indicating the potential for a profitable positive arbitrage. At 9:54:11, Whitebox entered new buy orders for approximately 2000 SPHE.
  6. By 9:54:12, Fast Basis had deteriorated substantially to -37.2, although Underlying Basis was 187.4. Whitebox cancelled the buy orders for approximately 2000 SPHE and replaced them with new orders for approximately 1000 SPHE. This is one of the two sets of Buy Orders that ASIC seeks to impugn. At the time this set of buy orders was placed, Fast Basis was 185.3 and Underlying Basis was 185.6, indicating a potentially profitable positive arbitrage.
  7. Between 9:55:04 and 9:55:41, Whitebox placed buy orders for approximately 500 SPHE and 200 SPHE, but cancelled these orders. At 9:55:41, Whitebox placed further buy orders for approximately 200 SPHE. This is the second set of the two sets of Buy Orders that ASIC seeks to impugn.
  8. At this time, Fast Basis was 8.6 and Underlying Basis was 144.2, once again indicating a potentially profitable positive arbitrage. But, by the time that these orders were executed (9:55:43), Fast Basis moved to -18.4, with Underlying Basis remaining positive at 135.3. Fast Basis appears to have moved in response to the placement of the second set of Buy Orders.
  9. The auction time line shows that between 9:55:04 (when Whitebox placed the subsequently cancelled buy orders for approximately 500 SPHE) and 9:55:43 (when the second set of Buy Orders was executed) Fast Basis was swinging between positive and negative. After 9:55:43, Fast Basis remained negative into the commencement of the OSPA.
  10. Between 9:55:43 and 9:59:17, Underlying Basis remained positive but had deteriorated from 135.3 to 85.8. This occurred in a period when there had been no change in Whitebox’s orders (a period of 3 minutes and 34 seconds). In and after this period, other arbitrageurs were entering certain buy orders. This may have accounted for the deterioration in Underlying Basis at this time.
  11. At 9:59:22, Whitebox entered its first order to sell SPI Futures. This is consistent with Whitebox seeking to execute a positive arbitrage. At this time, Underlying Basis was 98.6 (approximately 1%) and had been substantially positive all that morning. However, other arbitrageurs were continuing to place buy orders which seem to have either caused or contributed to a deterioration in both Underlying Basis and Fast Basis.
  12. At 9:59:31, Underlying Basis was 82.0 and Fast Basis was -11.7, indicating a potentially unprofitable positive arbitrage for Whitebox. Whitebox entered a cancellation of its buy orders for 1000 SPHE in respect of Group 1 (i.e., part of the first set of impugned Buy Orders): see [171] above. This was 14 seconds before the earliest time for the Group 1 rotation in the OSPA. At 9:59:35 Whitebox entered a cancellation of its buy orders for 200 SPHE in respect of Group 1 (i.e., part of the second set of impugned Buy Orders). These cancellations are two of the Cancellations to which ASIC refers. Whitebox maintained the two sets of impugned Buy Orders (at 1200 SPHE) in respect of each of the other Groups.
  13. From 9:59:38, Whitebox placed new buy orders across all Groups for 100 SPHE. These are the Subsequent Buy Orders to which ASIC refers. Therefore, by this time Whitebox’s buy orders in respect of Group 1 securities had been effectively reduced to a level of 100 SPHE with the remaining Groups increased to the level of 1300 SPHE. Thereafter, Whitebox cancelled its buy orders for 1000 SPHE and 200 SPHE (i.e., the remainder of the two sets of impugned Buy Orders) for each Group shortly before the earliest time for the commencement of each Group’s Rotation in the OSPA—at 10:01.32 and 10:01:35 for Group 2; 10:03:37 and 10:03:41 for Group 3; 10:05:36 and 10:05:44 for Group 4 and 10:08:08 and 10:08:11 for Group 5—thereby bringing the level of buy orders for securities in each of Groups 2 to 5 into line with the buy order for Group 1 (i.e., 100 SPHE). These are the other Cancellations to which ASIC refers.
  14. By the time the last Group 5 Cancellation had been executed, Fast Basis had moved positively to 10.6. At 10:09:07, just before the Group 5 securities opened for trading, Fast Basis was 1.0, meaning that the positive arbitrage executed by Whitebox would be profitable. Inconsistently with ASIC’s case theory, Whitebox bought a further 30 SPHE of Group 5 securities.
  15. With respect to Whitebox’s activities on 19 April 2012, the defendants submitted:
Throughout the morning of 19 April 2012 from 9:40am, the Whitebox trader was acting consistently with seeking to “capture” available mispricing by watching the calculation of Underlying Basis. The trader’s conduct was consistent with seeking to profit from potentially lucrative positive arbitrage opportunities (which at times showed in excess of 4% mispricing ie over 400bps). The Whitebox trader was actively managing basket orders in a manner consistent with an intention to “capture” that mispricing (in the sense of placing orders against it in the book, so as to position the trader to execute an order at market open to profit from any mispricing at that point), which had the effect of masking this Basis from competing arbitrageurs. It appears that this last endeavour was only partly successful, because of the magnitude of other participants’ activities and their effect on Basis and Underlying Basis. During this process, the Whitebox trader was also dealing with the decay of Basis as the time of the market opening rotations approached.
  1. The defendants submitted that the Cancellations in respect of the Buy Orders for Groups 2 to 5 were a:
... staged reduction ... consistent with the trader keeping his options open in case Basis swung more favourably in the later rotations (as occurred in February 2012). The existence of that possibility is consistent with the fact that there was still positive Underlying Basis.

17 May 2012

  1. Whitebox appears to have commenced its ordering activity at around 8:19:32. At 8:20:51, Whitebox entered sell orders for approximately 100 SPHE, which increased to approximately 200 SPHE by 8:27:16. During this period, Underlying Basis ranged between -19.5 and -15.3. Fast Basis ranged between 34.2 and 77.3. This level of Fast Basis made the sell orders potentially unprofitable. At 8:43:54 and 9:05:05, sell orders were entered by other arbitrageurs, without activity by Whitebox. These orders appear to have moved Underlying Basis from negative to positive, and increased Fast Basis substantially. Even so, Whitebox maintained its sell orders. At 9:09:47, Whitebox commenced to increase its sell orders to 300 SPHE. At that time, Underlying Basis was 151.6 and Fast Basis was 215.0. This ordering activity does not appear to have been consistent with a profitable negative arbitrage strategy.
  2. However, by 9:12:36, Whitebox amended its sell orders to zero and replaced those orders with buy orders. By 9:13:22, the buy orders were at the level of 260 SPHE. At this time, Underlying Basis was 139.9 and Fast Basis was 111.8, indicating a potentially profitable positive arbitrage. Therefore, the switch from sell to buy appears to have been a rational change in strategy.
  3. Whitebox increased its buy orders to approximately 460 SPHE at 9:14:35. However, Fast Basis swung to -26.9, indicating that a positive arbitrage was potentially unprofitable. At 9:14:39, Whitebox reduced the buy orders to 210 SPHE. Underlying Basis and Fast Basis moved favourably (for a positive arbitrage) to 59.6 and 133.0, respectively.
  4. At 9:18:56, Whitebox increased its buy orders by a further 200 SPHE. However, Fast Basis deteriorated. By 9:21:19, it was -2.4. It became increasingly negative thereafter. But Underlying Basis remained relatively stable at around 125.9. At 9:28:33, Whitebox increased its buy orders to approximately 500 SPHE. Underlying Basis and Fast Basis moved negatively.
  5. At 9:32:49, Whitebox changed directions and entered sell orders. By that time, it had been reducing its buy orders. By 9:33:08, it had cancelled its buy orders.
  6. By 9:43:22, Fast Basis was -0.3 and Underlying Basis was -184.8. Fast Basis and Underlying Basis became more negative. Whitebox commenced to make incremental increases to its sell orders which, by 9:45:27, were at approximately 750 SPHE. This is consistent with Whitebox pursuing a profitable negative arbitrage. By 9:48:13, Whitebox had increased its sell orders to 1100 SPHE. However, Fast Basis and Underlying Basis moved positively, and Whitebox amended its sell orders to zero.
  7. Then, at 9:56:38, Fast Basis moved to -202.7 and Underlying Basis moved to -222.8. Whitebox entered sell orders for 450 SPHE. At 9:56:45, it increased the sell orders to 750 SPHE. This strategy was consistent with a potentially profitable negative arbitrage. At 9:56:54, Whitebox increased the sell orders to approximately 1100 SPHE. This group of orders are the May Sell Orders to which ASIC refers.
  8. After Whitebox entered the sell orders for 1100 SPHE, Fast Basis moved to 25.5, indicating (possibly) that Whitebox’s amendment was too large. At 9:57:05, Whitebox amended the sell orders to approximately 1000 SPHE. Fast Basis remained positive at 38.4, indicating that a negative arbitrage at that level would be unprofitable. Whitebox commenced to reduce its sell orders incrementally. By 9:57:51, it had reduced the sell orders to approximately 900 SPHE.
  9. However, thereafter, Whitebox commenced amending the sell orders incrementally upwards. This activity took place before the impugned May Amended Sell Orders.
  10. In the period 9:58:26 to 10:06:22, Whitebox entered sell orders for approximately 1000 SPHE, having moved (as I have said) its sell orders in the immediately preceding period incrementally upwards from 900 SPHE, generally in multiples of either 5 or 10 SPHE. For Groups 2, 3, 4 and 5 the sell orders then moved down again to approximately 900 SPHE. The sell orders in this period (9:58:26 to 10:06:22) are the impugned May Amended Sell Orders. In this period (at 9:58:31), Whitebox also entered its first buy orders for SPI Futures. This step is consistent with Whitebox seeking to execute a negative arbitrage.
  11. At 9:59:38, Whitebox entered an amended sell order in Group 1 for approximately 200 SPHE in Group 1. This was approximately 7 seconds before the earliest time for commencement of the Group 1 Rotation.
  12. Whitebox successively entered amended sell orders for approximately 200 SPHE in each of the other Groups, as follows:
(a) Group 2 at 10:01:47 (approximately 13 seconds before the earliest time for commencement of the Group 2 Rotation);
(b) Group 3 at 10:04:00 (approximately 15 seconds before the earliest time for commencement of the Group 3 Rotation);
(c) Group 4 at 10:06:19 (approximately 11 seconds before the earliest time for commencement of the Group 4 Rotation); and
(d) Group 5 at 10:08:17 (approximately 28 seconds before the earliest time for commencement of the Group 5 Rotation).
  1. The amended sell orders for approximately 200 SPHE in each of Groups 1 to 5 are the impugned May Reductions. I note, however, that, in respect of Groups 2 to 5, the May Amended Sell Orders already involved reductions of previously placed sell orders from 1000 SPHE to 900 SPHE.
  2. The defendants pointed out that, as at 9:57:35 (before the May Amended Sell Orders), there was a disparity in arbitrage baskets entered by other arbitrageurs, consistent with those participants having, by that time, entered only part of their orders. This explanation is certainly possible. The defendants argued that, at that time, it would have been logical to expect that further arbitrage buy orders would be placed by other arbitrageurs prior to each Group opening. The defendants’ argument appears to be directed to establishing an objective justification for Whitebox maintaining its sell positions of approximately 1000 SPHE or 900 SPHE immediately prior to each Group rotation in the OSPA. However, the disparity shown in the auction time line is in respect of buy orders for Group 1 securities and in respect of sell orders for Group 5 securities. It is not apparent on the evidence why there would have been an expectation, at each of those times, that further buy orders would be placed with respect to Group 2 to Group 5 securities, especially with respect to Group 5 securities given that the disparity for that Group was, as I have said, on the sell side. I am not persuaded that the evidence positively establishes the proposition for which the defendants contended. The matter is left unclear.

19 July 2012

  1. Whitebox commenced entering buy and sell orders at 7:49. The initial buy and sell orders were for nominal amounts. In its case, ASIC referred only to Whitebox’s sell orders at this time. These are the July Sell Orders to which ASIC refers.
  2. By 7:51:22, Whitebox had entered buy orders for 200 SPHE. At 9:41:42, Whitebox amended the buy orders to 800 SPHE. At this time, Fast Basis was 326.9 and Underlying Basis was 374.2, indicating the prospect of a potentially profitable positive arbitrage. At 9:41:50, Whitebox amended its buy orders to 1200 SPHE. In the period 9:41:58 to 9:43:18, Whitebox progressively increased its buy orders to 3000 SPHE. By the time it had amended its orders to 3000 SPHE, Underlying Basis was 355.5 and, although deteriorating throughout this period, Fast Basis was 38.7.
  3. At 9:43:40, Whitebox commenced to reduce its buy orders to 2500 SPHE. In the course of doing this, Fast Basis moved to a negative figure. Whitebox reduced its buy orders incrementally, initially to 2000 SPHE and finally to 1000 SPHE at 9:45:37, at which time Fast Basis was -158.7 and Underlying Basis was 84.0.
  4. In the period 9:45:41 to 9:50:34, Fast Basis moved from -132.4 to -56.0. In this period, Whitebox was not entering orders. At 9:45:41, Underlying Basis was 84.6, but diminishing. It became negative, and then swung positive again at 9:47:23.
  5. From 9:50:34, Whitebox commenced to enter a series of reductions to its buy orders. Towards the end of this activity, Whitebox commenced entering amended sell orders. Prior to this time, its sell orders for nominal amounts had remained.
  6. By the time that Whitebox had reduced its buy orders to zero (9:53:48), Fast Basis was -247.1 and Underlying Basis was -99.0. Its sell orders at this time were standing at 900 SPHE, indicating a potentially profitable negative arbitrage. However, by the time the sell orders at this level were executed, Fast Basis had moved to 32.2, while Underlying Basis had moved to -77.5, indicating that a negative arbitrage at this level was potentially unprofitable. Fast basis continued to move positively. Underlying Basis remained negative, but fluctuated.
  7. At 9:54:36, Whitebox amended its sell orders to 1100 SPHE. Following that amendment, and the entry of buy orders from other arbitrageurs, Fast Basis moved to 124.4, while Underlying Basis remained at the level of approximately -75.0. At 9:54:47, Whitebox reduced its sell orders to 1000 SPHE.
  8. By 9:55:53, Fast Basis moved downwardly to 5.6. Underlying Basis moved to -151.6, without any activity by Whitebox.
  9. At 9:55:58, Fast Basis was 4.8 and Underlying Basis was -153.1. Whitebox amended its sell orders from 1000 to 1100 SPHE. Fast Basis then moved to 21.6. Underlying Basis remained relatively constant at -154.8. Whitebox amended its sell orders to 1200 SPHE at 9:56:14 and then, at 9:56:29, to 1300 SPHE.
  10. At 9:57:04, Whitebox entered its first buy order for SPI Futures.
  11. At 9:58:12, Whitebox amended its sell orders to approximately 1500 SPHE. At this time. Fast Basis was 19.1 and Underlying Basis was -221.0.
  12. From the above it can be seen that, in the period 9:53:48 to 9:58:12, Whitebox’s sell orders moved from 900 SPHE to 1100 SPHE; back down to 1000 SPHE; then up to 1100 SPHE; then to 1200 SPHE; then to 1300 SPHE; and then to 1500 SPHE. For Groups 2 to 5, the sell orders eventually moved from 1500 SPHE to 1700 SPHE (see below).
  13. In the period 9:58:17 to 9:59:24 (a period of just over 1 minute) Underlying Basis became less negative (from -219.5 to -201.0) and there was a disparity in the basket sizes sought by other arbitrageurs across the five Groups. The defendants argued that this was consistent with the possibility that other arbitrageurs would enter further sell baskets. The defendants also argued that, viewed objectively, there was a rational basis to expect that other arbitrageurs would align their orders between the Groups, and that this order activity would cause Fast Basis to move negatively.
  14. While the possibility that other arbitrageurs might have entered further sell baskets existed at that time, other ordering activity was at least equally possible. Moreover, there was also disparity in the basket sizes on the buy side. There is no expert evidence before me to support the contention that, objectively assessed, it was rational to expect that, in that period or in the immediately succeeding period leading up to the opening of the Group 1 Rotation in the OSPA, other arbitrageurs would align their orders, or that there would have been an expectation that other arbitrageurs’ ordering activity would lead to Fast Basis moving negatively. I have no evidence as to what Whitebox’s expectations were.
  15. As events transpired, in the period 9:58:17 to 9:59:24, further sell baskets were entered, but this activity only maintained the extent of the disparity that was evident at the beginning of the period. There was no change in the basket sizes on the buy side which, at that time, were also disparate. As the opening of the Group 1 Rotation approached, there was more ordering activity by other arbitrageurs on the sell side, but, similarly, this activity maintained the extent of the disparity in basket sizes on the sell side that was evident at 9:58:17. As the opening of the Group 1 Rotation approached there was no change in the basket sizes sought by other arbitrageurs or in the disparity that had been evident on the buy side since at least 9:58:17.
  16. At 9:59:24, Whitebox amended its sell order in Group 1 from 1500 to 300 SPHE, but maintained its sell orders of 1500 SPHE in Groups 2 to 5. At this time, Fast Basis was 8.6 and Underlying Basis was -201.0. By the time this amendment was executed, Fast Basis had moved to -12.54, indicating a potentially profitable negative arbitrage. Underlying Basis was -196.8. By that time, Whitebox had acquired 90 SPI Futures. This amendment was entered approximately 21 seconds before the earliest commencement for the Group 1 Rotation in the OSPA.
  17. At 9:59:33, Whitebox amended its sell orders for Groups 2 to 5 upwards to 1700 SPHE. At this time, Fast Basis was -15.8 and Underlying Basis was -193.00, indicating the possibility of a potentially profitable negative arbitrage. However, by the time these orders were executed, Fast Basis had deteriorated and, shortly thereafter, moved from a negative value to a positive value. Fast Basis then continued to move in a positive direction. Indeed, by the time the Group 1 Rotation in the OSPA commenced on that day (10:00:13), Fast Basis was 11.1. Underlying Basis was then -171.9.
  18. ASIC characterises certain order transactions entered in this period as the impugned July Amended Sell Orders. In respect of Group 1, these comprise one set of orders, being the amendment of Whitebox’s sell orders to 1500 SPHE at 9:58:13. In respect of Groups 2 to 5, these commence with the amendment of Whitebox’s sell orders to 1500 SPHE at 9:58:13, and include the subsequent upward amendment to 1700 SPHE. The July Amended Sell Orders in respect of Groups 2 to 5 comprise a number of sets of order amendments, even though, across a number of order sets, a particular SPHE level may have been maintained for some Groups before being reduced in size. In respect of Group 2, there were 3 sets of amended sell orders; in respect of Group 3, there were 16 sets of amended sell orders; in respect of Group 4, there were 30 sets of amended sell orders; and in respect of Group 5, there were 41 sets of amended sell orders.
  19. Whitebox then entered, successively, amended sell orders in Groups 2 to 5 down to approximately 300 SPHE by a series of amendments in respect of each Group. For each Group, the amendments were entered shortly before the earliest commencement of the Group’s Rotation in the OSPA, as follows:
(a) for Group 2, 13 downward amendments were made in the period 10:00:27 to 10:01:54 (a period of 1 minute and 27 seconds), with the last amendment being entered approximately 6 seconds before the earliest commencement for the Group 2 Rotation in the OSPA. Those amendments were (with approximated SPHE figures) from 1700 to 1620 SPHE (10:00:27); 1620 to 1500 SPHE (10:00:34); 1500 to 1390 SPHE (10:00:42); 1390 to 1270 SPHE (10:00:49); 1270 to 1160 SPHE (10:00:56); 1160 to 1040 SPHE (10:01:03); 1040 to 930 SPHE (10:01:11); 930 to 810 SPHE (10:01:18); 810 to 700 SPHE (10:01:25); 700 to 580 SPHE (10:01:33); 580 to 460 SPHE (10:01:40); 460 to 350 SPHE (10:01:47); and 350 to 300 SPHE (10:01:54);
(b) for Group 3, 14 downward amendments were made in the period 10:02:19 to 10:03:49 (a period of 1 minute and 30 seconds), with the last amendment being entered approximately 26 seconds before the earliest commencement for the Group 3 Rotation in the OSPA. Those amendments were (with approximated SPHE figures) from 1700 to 1680 SPHE (10:02:19); 1680 to 1640 SPHE (10:02:26); 1640 to 1610 SPHE (10:02:33); 1610 to 1580 SPHE (10:02:43); 1580 to 1550 SPHE (10:02:48); 1550 to 1520 SPHE (10:02:55); 1520 to 1490 SPHE (10:03:02); 1490 to 1450 SPHE (10:03:10); 1450 to 1420 SPHE (10:03:19); 1420 to 1390 SPHE (10:03:24); 1390 to 1360 SPHE (10:03:32); 1360 to 1330 SPHE (10:03:39); 1330 to 1300 SPHE (10:03:46); and 1300 to 300 SPHE (10:03:49);
(c) for Group 4, 11 downward amendments were made in the period 10:05:11 to 10:06:22 (a period of 1 minute and 11 seconds), with the last amendment being entered approximately 8 seconds before the earliest commencement for the Group 4 Rotation in the OSPA. Those amendments were (with approximated SPHE figures) from 1700 to 1600 SPHE (10:05:11); 1600 to 1470 SPHE (10:05:18); 1470 to 1330 SPHE (10:05:25); 1330 to 1200 SPHE (10:05:32); 1200 to 1060 SPHE (10:05:40); 1060 to 920 SPHE (10:05:47); 920 to 790 SPHE (10:05:54); 790 to 650 SPHE (10:06:02); 650 to 510 SPHE (10:06:09); 510 to 380 SPHE (10:06:16); and 380 to 300 SPHE (10:06:22); and
(d) for Group 5, 10 downward amendments were made in the period 10:07:33 to 10:08:36 (a period of 1 minute and 3 seconds), with the last amendment being entered approximately 9 seconds before the earliest commencement for the Group 5 Rotation in the OSPA. Those amendments were (with approximated SPHE figures) from 1700 to 1590 SPHE (10:07:33); 1590 to 1440 SPHE (10:07:40); 1440 to 1290 SPHE (10:07:47); 1290 to 1140 SPHE (10:07:54); 1140 to 990 SPHE (10:08:02); 990 to 840 SPHE (10:08:09); 840 to 690 SPHE (10:08:16); 690 to 540 SPHE (10:08:24); 540 to 390 SPHE (10:08:31); and 390 to 300 SPHE (10:08:36).
  1. These downwards amendments and the amendment for Group 1 from 1500 to 300 SPHE are the impugned July Reductions.

16 August 2012

  1. Mr Boshoff was not involved in trading on this day.
  2. Whitebox commenced entering buy and sell orders at 8:09:58. These orders were for nominal amounts. The sell orders are the August Sell Orders to which ASIC refers.
  3. At 8:10:12, Whitebox entered buy orders for approximately 25 SPHE. By 8:34:42, these orders had been amended to 600 SPHE. The orders were subsequently amended down to approximately 300 SPHE and, by about 8:59:05, to buy orders for either zero or nominal values. By this time, some other arbitrageurs also had entered buy orders.
  4. At 8:59:29, Whitebox entered its first substantial sell orders for 500 SPHE. At this time, Fast Basis was -315.0 and Underlying Basis was -316.5. These orders then increased to 800 SPHE and then by a number of increments to approximately 1600 SPHE at 9:06:43. During this period, Fast Basis swung between negative and positive values. Underlying Basis was approximately 2–3% negative.
  5. At 9:11:26, Whitebox amended its sell orders down to approximately 1400 SPHE, only to amend the orders back to approximately 1600 SPHE at 9:12:14. Between 9:12:14 and 9:39:54, Fast Basis swung between negative and positive values.
  6. From 9:39:02, Underlying Basis commenced to deteriorate from -287.1. Other arbitrageurs placed sell orders between 9:39:51 and 9:39:53. At this time, Fast Basis was positive, and became even more positive. Underlying Basis was negative, but became less negative.
  7. At 9:42:57, Whitebox amended its sell orders down to approximately 1000 SPHE and then, at 9:43:24, to approximately 800 SPHE. At 9:43:24, Fast Basis was 178.5, indicating that a negative arbitrage was potentially unprofitable. Other arbitrageurs placed buy orders. At 9:43:37, Fast Basis turned to a negative value. At 9:44:03, Whitebox increased its sell orders to approximately 900 SPHE.
  8. At 9:45:57, Fast Basis moved to -90.7. Underlying Basis was -247.2. Whitebox increased its sell orders to approximately 1200 SPHE. At the time that these orders were executed, Fast Basis was -25.2 and Underlying Basis was -246.2. Whitebox then amended the sell orders to approximately 1300 SPHE, at 9:46:11.
  9. At 9:46:15, Fast Basis was 18.3. Whitebox subsequently amended its sell orders back down to approximately 1200 SPHE, at 9:50:45. At the time these orders were executed, Fast Basis had decreased to 17.8.
  10. By 9:53:11, Fast Basis was -138.3 and Underlying Basis was -303.2. Whitebox increased its sell orders to approximately 1400 SPHE. Thereafter, Fast Basis and Underlying Basis became less negative. Even so, Whitebox increased its sell orders incrementally from 9:53:25. At 9:54:13, it amended its sell orders to approximately 1800 SPHE. At that time, Fast Basis was 5.2 and Underlying Basis was -278.2, indicating a potentially profitable negative arbitrage. But, by the time these orders were executed, Fast Basis moved to 14.7, indicating a potentially unprofitable negative arbitrage. At 9:57:00, Whitebox amended its sell orders to approximately 1700 SPHE. By the time that those orders were executed, Fast Basis was -4.0 and Underlying Basis was -247.4.
  11. Earlier, at 9:53:36, Whitebox entered buy orders for SPI Futures. At 9:55:37, those orders started to be filled.
  12. The defendants argued that in the period 9:00 to 9:58, Whitebox’s ordering activity demonstrated an attempt to keep orders in line with a potentially profitable negative arbitrage where Underlying Basis was constantly negative.
  13. At 9:58:52, Whitebox amended its sell orders to approximately 1500 SPHE. At this time, Fast Basis was 12.7 and Underlying Basis was -221.9. At the time that the orders were executed, Fast Basis moved to -18.5 and Underlying Basis moved to -222.6, indicating a potentially profitable negative arbitrage. At 9:59:16, Whitebox continued to amend its Group 1 sell orders by making progressive reductions from its position of approximately 1500 SPHE to approximately 1270 SPHE (at 9:59:16), then to approximately 950 SPHE (at 9:59:23), then to approximately 620 SPHE (at 9:59:31), and finally to approximately 400 SPHE (at 9:59:38). The last amendment was approximately 7 seconds before the earliest commencement of the Group 1 Rotation in the OSPA.
  14. The auction time line shows that from 9:50:04 to 9:56:15, the “sell” volume for SPI Futures was generally larger than the “buy” volume. The defendants argued that, thereafter, the SPI Futures market “became progressively better bid” and that, after 9:59:00, the “buy volume overwhelmed sell volume”, meaning that there was significantly less liquidity in the market in which Whitebox was seeking to acquire SPI Futures.
  15. The auction time line shows that after 9:56:15, both buy and sell volumes for SPI Futures fluctuated. I do not accept that after 9:59:00 buy volumes “overwhelmed” sell volumes. In the period 9:58:43 and 9:58:58, buy volumes exceeded sell volumes, but that position reversed at about 9:59:33. From that time, the buy and sell volumes were substantially the same, but fluctuating, until about 9:59:40, when the trend was that buy volumes exceeded sell volumes. From a buyer’s perspective, this meant that there was greater competition to buy SPI Futures than there was to sell SPI Futures. At this time, Whitebox’s evident strategy was to effect a negative arbitrage and, therefore, to buy SPI Futures. Between 9:56:33 to 9:59:37, the price of SPI Futures increased by 10 points to 4262.
  16. The defendants also pointed to what they said was a decay in Underlying Basis in the period 9:53:09 to 9:58:44. They argued that this decay, taken with the greater competition to acquire SPI Futures, meant that it was commercially rational for Whitebox to reduce the size of its sell orders. In this connection, the defendants contrasted the position at 9:53:25 (when Whitebox had placed amended sell orders for approximately 1500 SPHE at a time when Fast Basis was 82.8) and 9:59:15 (at a time when Fast Basis was -22.6, and after Whitebox had, at 9:58:52 amended its sell orders from approximately 1700 SPHE to approximately 1500 SPHE). The defendants argued that, by 9:59:15, the potential return on a negative arbitrage and the latitude against risk (in the sense of the margin of profit against market risk) had diminished.
  17. The defendants also pointed to the fact that, at 9:59:16, there was a disparity in the size of sell baskets placed by other arbitrageurs. The defendants argued that any expectation that this disparity would be addressed would also lead to an expectation that the then negative value for Fast Basis (which, as I have said, was -22.6) would reduce or disappear as the market came to open. There is no evidence before me as to what, objectively, the expectation would or could have been in this regard and, on the materials provided, I am unable to make any finding to that effect. There is no evidence before me of Whitebox’s actual expectation or expectations in that regard.
  18. ASIC characterises certain order transactions entered in this period as the impugned August Amended Sell Orders. In respect of Group 1, these comprise one set of orders, being the amendment of Whitebox’s sell orders to 1700 SPHE at 9:57:00. In respect of Groups 2 to 5, these commence with the amendment of Whitebox’s sell orders to 1700 SPHE at 9:57:00, and include further subsequent downward amendments (discussed further below). The August Amended Sell Orders in respect of Groups 2 to 5 comprise a number of sets of order amendments where the trend was downward, even though, across a number of order sets, a particular SPHE level may have been maintained for some Groups before being reduced in size:
(a) in respect of Group 2, there were 7 sets of amended sell orders from 1700 to 1500 SPHE;
(b) in respect of Group 3, there were 21 sets of amended sell orders from 1700, to 1500, and to 1200 SPHE;
(c) in respect of Group 4, there were 36 sets of amended sell orders from 1700, to 1500, to 1200, and to 1000 SPHE; and
(d) in respect of Group 5, there were 47 sets of amended sell orders from 1700, to 1500, to 1200, to 1000, to 800 and to 700 SPHE.
  1. Whitebox then entered, successively, amended sell orders in Groups 2 to 5 down to approximately 400 SPHE by a series of amendments in respect of each Group. For each Group, the amendments were entered shortly before the earliest commencement of the Group’s Rotation in the OSPA, as follows:
(a) for Group 2, 13 downward amendments were made in the period 10:00:29 to 10:01:55 (a period of 1 minute and 26 seconds), with the last amendment being entered approximately 5 seconds before the earliest commencement for the Group 2 Rotation in the OSPA. Those amendments were (with approximated SPHE figures) from 1500 to 1200 SPHE (10:00:29); 1200 to 1150 SPHE (10:00:36); 1150 to 1080 SPHE (10:00:43); 1080 to 1000 SPHE (10:00:50); 1000 to 930 SPHE (10:00:57); 930 to 860 SPHE (at 10:01:05); 860 to 780 SPHE (at 10:01:12); 780 to 710 SPHE (10:01:19); 710 to 640 SPHE (10:01:27); 640 to 570 SPHE (10:01:34); 570 to 490 SPHE (10:01:41); 490 to 420 SPHE (10:01:48); and 420 to 400 SPHE (10:01:55).
(b) for Group 3, 13 downward amendments were made in the period 10:02:40 to 10:04:07 (a period of 1 minute and 27 seconds), with the last amendment being entered approximately 8 seconds before the earliest commencement of the Group 3 Rotation in the OSPA. Those amendments were (with approximated SPHE figures) from 1200 to 990 SPHE (10:02:40); 990 to 940 SPHE (10:02:47); 940 to 890 SPHE (10:02:54); 890 to 840 SPHE (10:03:01); 840 to 790 SPHE (10:03:09); 790 to 740 SPHE (10:03:16); 740 to 690 SPHE (10:03:23); 690 to 640 SPHE (10:03:31); 640 to 590 (10:03:38); 590 to 540 SPHE (10:03:45): 540 to 490 SPHE (10:03:52); 490 to 440 SPHE (10:04:00); and 440 to 400 SPHE (10:04:07).
(c) for Group 4, 10 downward amendments were made in the period 10:04:43 to 10:06:23 (a period of 1 minute and 40 seconds), with the last amendment being entered approximately 7 seconds before the earliest commencement of the Group 4 Rotation in the OSPA. Those amendments were (with approximated SPHE figures) from 1000 to 800 SPHE (10:04:43); 800 to 700 SPHE (10:04:52); 700 to 670 SPHE (10:05:32); 670 to 630 SPHE (10:05:39); 630 to 590 SPHE (10:05:47); 590 to 550 SPHE (10:05:54); 550 to 510 SPHE (10:06:01); 510 to 470 SPHE (10:06:08); 470 to 430 SPHE (10:06:16); and 430 to 400 SPHE (at 10:06.23).
(d) for Group 5, 11 downward amendments were made in the period 10:07:22 to 10:08:35 (a period of 1 minute and 13 seconds), with the last amendment being approximately 10 seconds before the earliest commencement of the Group 5 Rotation in the OSPA. Those amendments were (with approximated SPHE figures) from 700 to 680 SPHE (10:07:22); 680 to 650 SPHE (10:07:29); 650 to 620 SPHE (10:07:37); 620 to 600 SPHE (10:07:44); 600 to 570 SPHE (10:07:51); 570 to 540 SPHE (10:07:59); 540 to 510 SPHE (10:08:06); 510 to 490 SPHE (10:08:13); 490 to 460 SPHE (10:08:20); 460 to 430 SPHE (10:08:28); and 430 to 400 SPHE (at 10:08:35).
  1. These downwards amendments, and the series of amendments for Group 1 from 1700 SPHE to 400 SPHE, are the impugned August Reductions.
  2. As I have noted, ASIC’s case is that, in the period 9:57:00 to 10:06:29, Whitebox entered the August Amended Sell Orders and that, shortly before each Rotation in the OSPA, Whitebox entered the August Reductions. I also note, however, that, in respect of Groups 2 to 5, the August Amended Sell Orders themselves involved reductions from previous sell orders.
  3. Specifically, in respect of Group 2, ASIC relies on 13 sets of reductions for the August Reductions. However, the August Amended Sell Orders for this Group includes a reduction from approximately 1700 SPHE to approximately 1500 SPHE. In respect of Group 3, ASIC again relies on 13 sets of reductions, but the August Amended Sell Orders for this Group include sell orders effecting reductions from approximately 1700 SPHE to approximately 1500 SPHE, and from approximately 1500 SPHE to approximately 1200 SPHE. In respect of Group 4, ASIC relies on 10 sets of reductions, but the August Amended Sell Orders for this Group include sell orders effecting reductions from approximately 1700 SPHE to approximately 1500 SPHE, from approximately 1500 SPHE to approximately 1200 SPHE, and from approximately 1200 SPHE to approximately 1000 SPHE. In respect of Group 5, ASIC relies on 11 sets of reductions, but the August Amended Sell Orders for this Group include sell orders effecting reductions from approximately 1700 SPHE to approximately 1500 SPHE, from approximately 1500 SPHE to approximately 1200 SPHE, from approximately 1200 SPHE to approximately 1000 SPHE, from approximately 1000 SPHE to approximately 800 SPHE, and from approximately 800 SPHE to approximately 700 SPHE. Put simply, before the August Reductions on which ASIC relies, Whitebox was already progressively amending down or reducing its sell orders in Groups 2, 3, 4 and 5.

18 October 2012

  1. Whitebox commenced entering buy and sell orders at 8:03:00. The orders were for nominal amounts. The sell orders at 8:03:04 are the October Sell Orders to which ASIC refers.
  2. Between 8:03:14 and 8:51:55, Whitebox entered amended buy orders for various amounts, ranging from approximately 50 SPHE to approximately 250 SPHE, and then back down to 50 SPHE. At 8:52:01, Whitebox reduced the buy orders to zero and amended its sell orders to 50 SPHE and then, at 9:07:55, to approximately 100 SPHE. The sell orders were then amended upwards by a series of amendments which culminated in an amended sell order for 1000 SPHE at 9:17:25. By 9:17:55, the sell orders were reduced to zero and buy orders for approximately 300 SPHE were entered. At this time, Fast Basis was 406.4 and Underlying Basis was 231.9, indicating a potentially profitable positive arbitrage.
  3. In the period 9:18:09 to 9:19:30, Whitebox entered various amended buy orders. At 9:20:33, Whitebox entered a buy order for approximately 500 SPHE. At that time, Fast Basis was -16.9 and Underlying Basis was 173.8.
  4. At 9:25:57, Whitebox reduced the buy orders to zero and entered amended sell orders for approximately 300 SPHE. At that time, Fast Basis was -4.0 and Underlying Basis was 168.5.
  5. In the period 9:28:59 to 9:45:22, Whitebox amended the sell orders on a number of occasions. At 9:45:22, it entered amended sell orders for 2000 SPHE when Fast Basis was -72.2 and Underlying Basis was -371.3, indicating a potentially profitable negative arbitrage.
  6. At 9:48:14, Whitebox amended its sell orders to approximately 1900 SPHE at a time when Fast Basis had deteriorated to -8.5. At 9:51:33, Whitebox amended its sell orders back to approximately 2000 SPHE. At that time, Fast Basis was -21.5 and Underlying Basis was 319.9. At the time these orders were executed, Fast Basis had moved marginally positive to 1.0, while Underlying Basis remained relatively stable at -319.3.
  7. At 9:53:40, Fast Basis moved to 19.2, while Underlying Basis was -285.1. Whitebox amended its sell orders to approximately 1900 SPHE. At the time these orders were executed, Fast Basis was 4.4 and Underlying Basis was -284.9.
  8. At 9:55:02, Whitebox entered its first buy orders for SPI Futures. At 9:55:47, Fast Basis moved to -2.0. The evidence shows that on this morning, there were substantially greater volumes of SPI Futures on the sell side than on the buy side. This persisted until after 9:55.02, and by reference to the first 5 price levels, there was a large number of SPI Futures available to hedge a large cash order. However, by about 9:55:53, this position changed.
  9. At 9:57:14, Whitebox entered amended sell orders for approximately 1800 SPHE. At this time, Fast Basis was 11.4 and Underlying Basis was -261.2. This is one of the impugned October Amended Sell Orders.
  10. ASIC characterises the sell orders for 1800 SPHE as the impugned October Amended Sell Orders. In respect of Group 1, these comprise one set of orders (at 9:57:14). The October Amended Sell Orders in respect of Groups 2 to 5 comprise a number of sets of order amendments, even though, across a number of order sets, a particular SPHE level may have been maintained for some Groups before being reduced in size. In respect of Group 2, there were 2 sets of amended sell orders; in respect of Group 3, there were 21 sets of amended sell orders; in respect of Group 4, there were 43 sets of amended sell orders; and in respect of Group 5, there were 50 sets of amended sell orders.
  11. At 9:58:19, Woodside Petroleum Limited (WPL) entered “Notice Received” status, thereby indicating a suspension or deferral of trading in that security which constituted, by weight, 1.97% of the XJO. In this state, market participants were able to place, amend or cancel orders for WPL securities, but not trade in them. The defendants correctly pointed out that a suspension or deferral of trading in WPL meant that there would be difficulty in determining, at that time, the SPI Futures settlement price given that the first traded price of WPL (assuming it did trade on 18 October 2012) would be used in determining the level of the XJO. At about the same time (9:58:19), Fast Basis moved to -5.0 while Underlying Basis was -251.7. Fast Basis then moved to -13.4 at 9:58:25 and stayed negative until 9:59:00. Therefore, at this time, a negative arbitrage position was potentially profitable.
  12. Between 9:59:00 and 9:59:15, Fast Basis moved to a positive value and then became negative again at 9:59:16. Underlying Basis remained greater than -2%.
  13. The defendants pointed to a decay in Underlying Basis in the period from 9:58:19 to 9:59:42 and argued that if this decay continued, the observed level of Underlying Basis would not persist throughout the OSPA. However, throughout this period and up to the Group 1 Rotation at 9:59:53, Underlying Basis, although decaying, remained at a level greater than -2%.
  14. By 9:59:33, Whitebox had acquired only 90 SPI Futures. Persisting with a sell order for approximately 1800 SPHE until market opening would have resulted in Whitebox bearing a very substantial unhedged risk. At 9:59:44, Whitebox entered an amended sell order for Group 1 securities for approximately 500 SPHE. At this time, Fast Basis was -6.9, indicating a potentially profitable negative arbitrage. The defendants argued that it was rational for Whitebox to make this amendment given the relatively thin margin for Fast Basis (which might not persist until commencement of the Group 5 Rotation) and what would have been a large unhedged exposure having regard to the fact that Whitebox had acquired only 100 SPI Futures before the earliest commencement of the Group 1 Rotation. I accept that to be the case.
  15. The amendment order for approximately 500 SPHE was entered less than 1 second before the earliest commencement of the Group 1 Rotation. ASIC says this caused a spike in the price of the Group 1 securities. The defendants submitted that I should infer that the delay in entering the amendment was due to distraction caused by the WPL notice. I am not prepared to draw that inference. This explanation is no more than a speculative possibility. It is a matter on which the defendants could have adduced direct evidence, but elected not to do so.
  16. The defendants also submitted that I should accept that the lateness with which the amendment order was placed was due to human error, rather than deliberate price manipulation. This is a matter to which I shall return. However, I note once again that this is a matter on which the defendants could have adduced, but did not adduce, direct evidence.
  17. The defendants submitted that the objective data in respect of the futures market on the morning of 18 December 2012 was consistent with the pursuit of a potentially large arbitrage opportunity which diminished in the lead-up to market opening because of a diminution in the liquidity of SPI Futures after 9:55:00. The auction time line shows that, in the period 9:50:17 to 9:55:33, the volume of SPI Futures offered for sale exceeded the bid volume for those instruments. The defendants submitted that there was an objectively reasonable basis for Whitebox to form the view at 9:55:02, when it commenced to enter orders to acquire SPI Futures, that the acquisition of a large number of futures (to hedge an equally large cash order) was achievable. I do not have evidence of what Whitebox’s traders’ views were at that time. However, in cross-examination, Mr Morgan accepted that if there happened to be in excess of, say, 3000 SPI Futures available for acquisition (such as happened on 16 February 2012), it would be possible to transact 1800 futures, although, in doing so, one would expect that transacting futures at that level would affect the prices at which they could be acquired. It would be a matter of judgment for the trader to assess whether the potential profit involved in undertaking the arbitrage would justify him or her in taking the risk of increasing the prices at which the futures could be acquired. The defendants submitted that, on 18 December 2012, it was rational to maintain baskets of securities (I interpolate, in the order of 1800 SPHE) to account for the possibility that a large hedge might become available. The defendants submitted that, if the opportunity for such a hedge did not eventuate, it would be logical to amend down the sell baskets appropriately.
  18. Prior to the respective Rotations for Groups 2 to 5, Whitebox entered, successively, amended sell orders in Groups 2 to 5 down to approximately 500 SPHE by a series of amendments in respect of each Group. For each Group, the amendments were entered shortly before the earliest commencement of the Group’s Rotation in the OSPA, as follows:
(a) for Group 2, 15 downward amendments were made in the period 10:00:09 to 10:01:51 (a period of 1 minute and 42 seconds), with the last amendment being entered approximately 9 seconds before the earliest commencement of the Group 2 Rotation in the OSPA. Those amendments were (with approximated SPHE figures) from 1800 to 1740 SPHE (10:00:08); 1740 to 1650 SPHE (10:00:16); 1650 to 1560 SPHE (10:00:24); 1560 to 1470 SPHE (10:00:31); 1470 to 1380 SPHE (10:00:38); 1380 to 1290 SPHE (10:00:45); 1290 to 1200 SPHE (10:00:53); 1200 to 1110 SPHE (10:01:00); 1110 to 1030 SPHE (10:01:07); 1030 to 940 SPHE (10:01:15); 940 to 850 SPHE (10:01:22); 850 to 760 SPHE (10:01:29); 760 to 670 SPHE (10:01:32); 670 to 580 SPHE (10:01:44); and 580 to 500 SPHE (10:01:51).
(b) for Group 3, 15 downward amendments were made in the period 10:02:27 to 10:04:09 (a period of 1 minute and 42 seconds), with the last amendment being entered approximately 6 seconds before the earliest commencement of the Group 3 Rotation in the OSPA. Those amendments were (with approximated SPHE figures) from 1800 to 1710 SPHE (10:02:27); 1710 to 1620 SPHE (10:02:35); 1620 to 1530 SPHE (10:02:42); 1530 to 1440 SPHE (10:02:49); 1440 to 1350 SPHE (10:02:56); 1350 to 1260 SPHE (10:03:04); 1260 to 1170 SPHE (10:03:11); 1170 to 1080 SPHE (10:03:18); 1080 to 990 SPHE (10:03:26); 990 to 900 SPHE (10:03:33); 900 to 810 SPHE (10:03:40); 810 to 720 SPHE (10:03:47); 720 to 630 SPHE (10:03:55); 630 to 540 SPHE (10:04:02); and 540 to 500 SPHE (10:04:09).
(c) for Group 4, 11 downward amendments were made in the period 10:05:07 to 10:06:20 (a period of 1 minute and 13 seconds), with the last amendment being entered approximately 10 seconds before the earliest commencement of the Group 4 Rotation in the OSPA. Those amendments were (with approximated SPHE figures) from 1800 to 1650 SPHE (10:05:07); 1650 to 1530 SPHE (10:05:15); 1530 to 1420 SPHE (10:05:22); 1420 to 1300 SPHE (10:05:29); 1300 to 1190 SPHE (10:05:37); 1190 to 1070 SPHE (10:05:44); 1070 to 960 SPHE (10:05:51); 960 to 840 SPHE (10:05:58); 840 to 730 SPHE (10:06:06); 730 to 610 SPHE (10:06:13); and 610 to 500 SPHE (10:06:20).
(d) for Group 5, 23 downward amendments were made in the period 10:05:58 to 10:08:39 (a period of 2 minutes and 41 seconds), with the last amendment being entered approximately 7 seconds before the earliest commencement of the Group 5 Rotation in the OSPA. Those amendments were (with approximated SPHE figures) from 1800 to 1730 SPHE (10:05:58); 1730 to 1680 SPHE (10:06:06); 1680 to 1620 SPHE (10:06:13); 1620 to 1560 SPHE (10:06:20); 1560 to 1500 SPHE (10:06:27); 1500 to 1450 SPHE (10:06:35); 1450 to 1390 SPHE (10:06:42); 1390 to 1330 SPHE (10:06:49); 1330 to 1270 SPHE (10:06:57); 1270 to 1220 SPHE (10:07:04); 1220 to 1160 SPHE (10:07:11); 1160 to 1100 SPHE (10:07:18); 1100 to 1050 SPHE (10:07:265); 1050 to 990 SPHE (10:07:33); 990 to 930 SPHE (10:07:40); 930 to 870 SPHE (10:07:47); 870 to 820 SPHE (10:07:55); 820 to 760 SPHE (10:08:02); 760 to 700 SPHE (10:08:09); 700 to 640 SPHE (10:08:17); 640 to 590 SPHE (10:08:24); 560 to 530 SPHE (10:08:31); and 530 to 500 SPHE (10:08:39).
  1. These downwards amendments, and the amendment for Group 1 from 1800 SPHE to 500 SPHE, are the impugned October Reductions. As is apparent, the amendments for Group 2 to 5 were implemented in modest, staged reductions over relatively lengthy periods of time. This is to be contrasted with the amendment to Group 1. The defendants submitted that the amendment to Group 1 from 1800 SPHE to 500 SPHE is at least equally consistent with accident and mistiming as it is with ASIC’s case of a deliberate scheme to manipulate the Opening Price of XJO Securities. The defendants argued that to deliberately leave the Group 1 amendment so close to the earliest time for the Group 1 Rotation (the amendment was not fully entered until four seconds after that time) would be a very dangerous strategy because, had the OSPA commenced at or close to that time, Whitebox’s orders would have been partly filled to the extent of 1800 SPHE for Group 1, leaving that position to be unwound in open trading, attracting a very real risk of loss. The prospective size of the delta, and hence the risk, was a compelling objective reason why it was rational for Whitebox to amend the Group 1 order downwardly before the Group 1 Rotation. The defendants submitted that the inference that the amendment was made to avoid creating such a large delta is, at the very least, equally as likely as the inference of nefarious intent argued by ASIC.

Some observations on the auction time lines

  1. As I have already noted, there is a difference between ASIC and the defendants as to when the May Reductions and the August Reductions should be taken as having been commenced.
  2. For 17 May 2012, ASIC relied on the amendments made to each of Groups 2 to 5 to reduce the sell orders from 900 SPHE to 200 SPHE. However, as the defendants pointed out, ASIC’s analysis omits the reductions made for each Group which commenced earlier with amendments from 1000 SPHE to 900 SPHE.
  3. As I have noted for 16 August 2012, ASIC relied on 13 sets of order amendments for Group 2; 13 sets of order amendments for Group 3; 10 sets of order amendments for Group 4, and 11 sets of order amendments for Group 5. However, as the defendants pointed out, ASIC’s analysis omits an earlier order amendment from 1700 SPHE to 1500 SPHE for Group 2; earlier order amendments from 1700 to 1500 and from 1500 to 1200 for Group 3; earlier order amendments from 1700 SPHE to 1500 SPHE, from 1500 SPHE to 1200 SPHE and from 1200 SPHE to 1000 SPHE for Group 4, and earlier order amendments from 1700 SPHE to 1500 SPHE, from 1500 SPHE to 1200 SPHE, from 1200 SPHE to 1000 SPHE, from 1000SPHE to 800 SPHE and from 800 SPHE to 700 SPHE for Group 5.
  4. The defendants submitted that ASIC’s case presents a distorted picture of Whitebox’s order activity on 17 May 2012 and 16 August 2012. The defendants submitted that, when properly considered, Whitebox did not, in fact, reduce sell orders “shortly before” the earliest commencement of the Rotations for the various Groups in the OSPA. The defendants submitted that the auction time lines show that, for 17 May 2012, there was a series of staged amendments for all Groups commencing earlier than ASIC had contended for. The position is even starker for 16 August 2012 where, on the defendants’ argument, the reductions actually commenced some minutes before the times that ASIC contended for. In oral closing submissions, the defendants submitted that the order amendments on which ASIC relied for 16 August 2012 appear to have been selected so that they fall within a time period to accommodate Mr Graves’ 2 Minute Rule, discussed below.
  5. I accept the defendants’ submission that the Reductions on which ASIC relies do not show a complete picture of Whitebox’s order activity for 17 May 2012 and 16 August 2012 and that the order amendments omitted by ASIC should be taken into account. ASIC’s answer was to contend that these earlier order amendments were not relevant to its case and did not require explanation with respect to ASIC’s case theory. I disagree.
  6. The defendants also submitted that, when viewed against the auction time lines, there is no pattern with regard to the proximity in time at which the Cancellations and Reductions commenced with respect to the earliest time for each Rotation in the OSPA. I accept that submission on the data presented.
  7. The defendants submitted that if the object was to amend orders sufficiently close to the earliest time for the Rotation in Group 1, so that any immediate impact on the IMP would carry through to the Opening Prices of the securities concerned, there would be at least some pattern as to the time at which the Group 1 order amendments were made.
  8. Taking the May Reductions, the defendants submitted (in the same vein) that if the order amendments in respect of Groups 2 to 5 from 900 SPHE to 200 SPHE were intended to impact on the Opening Prices of the securities concerned, why wouldn’t the preceding amendment for all Groups from 1000 SPHE to 900 SPHE have been made shortly prior to the Rotation for each Group?
  9. Dealing with the position revealed by the August auction time line, the defendants submitted that the reductions effected by the order amendments with respect to Groups 2 to 5 were not “held back” until shortly before the time of the earliest commencement for each Rotation. Rather, the defendants submitted, the amendments were made “in a series of early, large steps down in the orders”, such that the final volume reductions came to be made “in a series of spaced out, modestly incremental steps, over the course of periods of 90 seconds or more”. I accept that submission.
  10. The defendants advanced the same submission with respect to 18 October 2012. In this connection, the defendants submitted, once again, that the reductions effected by the order amendments with respect to Groups 2 to 5 were implemented “in a series of early, large steps” and then “incrementally over relatively lengthy periods of time”—for Group 2, 111 seconds; for Group 3, 108 seconds; for Group 4, 83 seconds, and for Group 5, 167 seconds. I accept that submission.
  11. With respect to the Incremental Reductions, the defendants submitted that, if the various reductions in order volume were intended to impact on the Opening Prices of the securities concerned, why were they made incrementally (which would likely attenuate, if not completely eliminate, the effect of the amendments on price) and not in a “one-off” amendment?
  12. I accept that the various matters identified by the defendants are inconsistent with ASIC’s case theory and raise legitimate questions for consideration as to the timing of the amendments that were in fact made by Whitebox to either cancel or reduce its orders on the Serial Expiry Days in question. These matters are necessarily part of the mix of considerations to be taken into account when considering whether the defendants had, and acted with, the purpose ASIC alleges.

ASIC’S CASE ON THE IMPUGNED TRADING

  1. ASIC’s case is that the Buy Orders (on 19 April 2012), and the Amended Sell Orders (on 17 May, 19 July, 16 August and 18 October 2012) and the Incremental Reductions (on 19 July, 16 August and 18 October 2012) were not placed with any genuine intention that they should be traded, but rather with the knowledge that, ultimately, they would be cancelled or substantially reduced in volume, as in fact occurred. To recapitulate, ASIC submitted that when those orders are viewed in light of the ultimate Cancellations and Reductions, and the delay in making those Cancellations and Reductions, it is apparent that the purpose of the impugned orders was to manipulate the Opening Price for each of the XJO Securities in their respective OSPAs to thereby generate a profit, or to increase the profit, on Whitebox’s index arbitrage trading.
  2. ASIC sought to make good this contention by reference to (what it described as) the “pillars” of its case. In its Concise Statement filed on 18 March 2016, ASIC relied on six matters from which it said the defendants’ intentions and purposes, and likely market effects, could be inferred. In submissions, ASIC said there were eight “pillars”. The difference is really a reorganisation of the propositions in the Concise Statement, not an alteration of substance. I will address the eight “pillars” in ASIC’s closing submissions.
  3. The first pillar is that, at the times the Amended Sell Orders were placed on 17 May, 19 July, 16 August and 18 October 2012, the arbitrage position established by Whitebox would have appeared to have been unprofitable, assuming the market to have opened at those times.
  4. This pillar is based on the historical volumes of SPI Futures that were available to be traded, and Mr Morgan’s analysis of the data. I will return to Mr Morgan’s analysis and the defendants’ criticisms of it in later paragraphs of these reasons.
  5. ASIC submitted that this pillar undercuts the commercial rationale for the orders. In this connection, ASIC argued that, intentional market manipulation aside, the Amended Sell Orders could “only have made sense” if there had been an expectation of market change. However, the defendants gave no evidence of any such an expectation or any reason for there to have been such an expectation.
  6. The second pillar is that, looking forward at the time of the Buy Orders (19 April 2012), and the Amended Sell Orders and the Incremental Reductions (other Serial Expiry Days), the defendants could not have expected that sufficient SPI Futures could be traded profitably by 10:08:45 (the earliest possible commencement for the Group 5 Rotation) to match the securities that were subject of those orders.
  7. This pillar is also based (in part) on historical volumes of SPI Futures and Mr Morgan’s analysis of the data. Again, I will return to Mr Morgan’s analysis and the defendants’ criticisms of it in later paragraphs of these reasons. But, in general terms (and subject to certain exceptions), Mr Morgan’s analysis was that Whitebox would have had to have traded substantial (and, in some cases, very significant) percentages of the SPI Futures historically traded by the entire market to match the Buy Orders or Amended Sell Orders. According to Mr Morgan, this would have moved the price of the SPI Futures against Whitebox such that the transactions would have been either unprofitable or substantially less profitable than they would have been absent the price impact caused by trading a very large percentage of the market for SPI Futures.
  8. Based on Mr Morgan’s evidence, ASIC submitted more specifically that, at the time the second set of Buy Orders (19 April 2012), and the Amended Sell Orders (before or at the time of the Reductions) (other Serial Expiry Days) and the Incremental Reductions for 19 July 2012 were placed, a reasonable arbitrageur could not reasonably have expected to profitably trade a matching volume of SPI Futures.
  9. ASIC also relied on Whitebox’s stated position on its trading strategy, which (as I have recorded) was to trade SPI Futures as closely as possible to XJO Securities to avoid exposure from an unmatched position. ASIC submitted that any suggestion that Whitebox would have carried a substantial unhedged position beyond 10:08:45 would have been contrary to its stated position, and contrary to what actually happened on the days in question. ASIC submitted that the defendants must have understood that the SPI Futures necessary to match the (relevant) Buy Orders, the Amended Sell Orders and the Incremental Reductions could never have been traded. On this basis, ASIC submitted that there could not have been an intention to trade those orders.
  10. The third pillar is that, based on Whitebox’s actual trading in SPI Futures on the Serial Expiry Days in question, Whitebox did not seriously pursue the volumes of futures necessary to offset, by 10:08:45, the Buy Orders, the Amended Sell Orders or the Incremental Reductions for XJO Securities, although Whitebox’s actual trading was consistent with seeking to match, by 10:08:45, the value of XJO Securities represented by the Subsequent Buy Orders (19 April 2012) and the final set of Reductions for each Rotation Group (other Serial Expiry Days).
  11. This pillar is based on further analyses carried out by Mr Morgan—including one referred to colloquially in the course of the hearing as the “run-rate” analysis—on the pattern of Whitebox’s trading on the days in question. Once again, I will return to Mr Morgan’s analysis and the defendants’ criticism of it in later paragraphs of these reasons. It is sufficient to note at this stage that ASIC submitted that, from the outset and throughout the Serial Expiry Days in question, the defendants’ conduct was inconsistent with establishing an arbitrage position consistent with the Buy Orders, the Amended Sell Orders and the Incremental Reductions. ASIC submitted that the “natural inference” is that the orders were not intended to be traded; rather, it was anticipated that they would be reduced so as to manipulate the price for XJO Securities as part of the OSPA.
  12. The fourth pillar is that the Cancellations and Reductions were delayed until very shortly prior to the Rotation of the OSPA in which the relevant orders were due to trade. According to ASIC, this conduct was likely to affect the price at which the relevant XJO Securities would trade in the OSPA. This aspect of the fourth pillar is based largely on Mr Graves’ evidence to which I will return. I will simply note, at this stage, that Mr Graves’ evidence of likely price effect proceeded according to a number of rules of thumb, which he posited based on his experience. The soundness of these rules of thumb was the subject of significant challenge, particularly from Professor Aitken. ASIC submitted that, given (what it contended was) the likely effect of this conduct on price, it could be inferred that the defendants’ purpose was to achieve this effect.
  13. The fifth pillar is related to the fourth pillar: there was delay in making the Cancellations and Reductions for securities to be traded in the last four Rotations of the OSPA until well after the Group 1 Rotation and shortly prior to the Rotation in the OSPA in which the relevant securities were due to trade. ASIC submitted that Whitebox’s trading strategy was such that it must have known by at least the time of the Group 1 Rotation the size of the arbitrage position it wanted to trade, but it nonetheless delayed in implementing the Cancellations and Reductions. Further, by making the Incremental Reductions, Whitebox was placing reduced orders that could never have been intended to be traded. ASIC submitted that this delay is only explicable as an endeavour to preserve the effect of the order volumes to ensure that they influenced the setting of the Opening Prices of the securities in the OSPA.
  14. The sixth pillar is that the strategy of placing very large orders for XJO Securities, which were then cancelled or reduced shortly prior to the relevant Rotations in the OSPA, was repeated over a long period of time. ASIC submitted that the significance of this pillar lies in the notion that the impugned trading ought not to be seen as responsive to market conditions on the day but, rather, as a deliberate and premeditated strategy implemented in accordance with the defendants’ intentions. ASIC submitted that the deliberateness of the conduct is underscored by the consistent and uniform pattern of the size of Whitebox’s arbitrage position across the Serial Expiry Days from 19 April 2012 (where the acquired XJO Securities position was approximately 100 SPHE) to 18 October 2012 (where the acquired XJO Securities position was approximately 500 SPHE), with the intervening days (17 May 2012, 19 July 2012, and 16 August 2012) showing acquired positions of approximately 200 SPHE, 300 SPHE and 400 SPHE, respectively. According to ASIC, this was not the “amusing coincidence” that Mr de Kantzow had sought to portray in his evidence, but a pattern of suspicious trading.
  15. The seventh pillar is that the defendants were incentivised to obtain the greatest profit possible by reason of this trading. ASIC accepted that this fact alone does not establish that the impugned orders were placed with a manipulative purpose. It did, however, establish a motive for doing so.
  16. The eighth pillar is that Whitebox’s trading systems allowed it to monitor the effect on mispricing that would be caused by its orders. ASIC argued that Whitebox was keenly aware of the impact of its placement, amendment and cancellation of orders for XJO Securities upon the market and, in particular, the effect that withdrawing order volume would have. Indeed, ASIC argued that Whitebox’s contemporaneous statements bespoke the intention to affect the prices of XJO Securities in order to alter the level of the XJO. Put simply, ASIC’s case is that Basis was consciously “calibrated” or manipulated for the purpose of making Whitebox’s trading profitable or more profitable.
  17. ASIC submitted that when all these matters are considered, they lead to the “irresistible inference” that the Buy Orders, the Amended Sell Orders and the Incremental Reductions were never intended to be traded. They were simply placed to provide a “platform” from which cancellations and reductions could be made for the purpose of manipulating the price of XJO Securities in the OSPA. ASIC submitted that the defendants provided no alternative explanation for the impugned trading, let alone established such an explanation through evidence.
  18. I will return to consider ASIC’s case in the context of s 1041B(1)(b) and s 1041A of the Act. In the immediately following sections of these reasons I will deal with the evidence relied on by ASIC to establish each of the eight “pillars”.

MR MORGAN’S EVIDENCE

  1. As I have noted, the first three pillars of ASIC’s case were based wholly or substantially on Mr Morgan’s evidence and, in particular, his analyses of certain data provided to him.

The first pillar

  1. For the first pillar, ASIC relied on mathematical calculations performed by Mr Morgan using assumed values of Fair Value (in the case of 18 October 2012, using two such values (long Fair Value and short Fair Value)), actual Indicative Opening Prices for XJO Securities, and actual prices for SPI Futures.
  2. Mr Morgan’s calculations reveal that:
(a) the dollar value loss from selling the quantity of XJO Securities that were the subject of the first May Amended Sell Orders, while purchasing an equivalent number of June SPI contracts at prices which prevailed immediately before the placement of the order, would be $935,657.15;
(b) the dollar value loss from selling the quantity of XJO Securities that were the subject of the first July Amended Sell Orders, while purchasing an equivalent number of September SPI contracts at prices which prevailed immediately before the placement of the order would be $433,465.90;
(c) the dollar value loss from selling the quantity of XJO Securities that were the subject of the first August Amended Sell Orders, while purchasing an equivalent number of September SPI contracts at prices which prevailed immediately before the placement of the order, would be $190,775.33; and
(d) the dollar value loss from selling the quantity of XJO Securities that were the subject of the first October Amended Sell Orders, while purchasing an equivalent number of December SPI contracts at prices which prevailed immediately before the placement of the orders, would be either $362,060 or $387,404.20 (depending on the Fair Value chosen).
  1. In their joint report, Mr Morgan and Mr de Kantzow agreed that there was no substantial dispute between them as to the actual calculations performed by Mr Morgan. However, they also agreed that the task which Mr Morgan had been set addressed a very particular snap-shot of hypothetical profitability, and that were the snap-shot to be taken at another time, the hypothetical profitability of the trading in question might be (and often would be) significantly different. This is not surprising, given that different variables may well be in play at different times. Mr Morgan and Mr de Kantzow agreed that the Amended Sell Orders were best examined in their context of multiple amendments in a sequence as shown in the auction timelines, rather than as an arbitrary snap-shot.
  2. Even so, ASIC submitted that the task which Mr Morgan had been set was neither artificial nor irrelevant. Mr de Kantzow accepted that a competent and experienced arbitrageur, acting consistently within the normal principles and practice of securities index arbitrage, will monitor Basis for the duration of the Pre-Open Phase and, notwithstanding the phenomenon of Basis Decay, place appropriate orders so as to be in a position to capitalise on any potentially profitable remaining Basis at market opening. ASIC submitted that, given the likelihood of Basis Decay, the notional profitability of an arbitrage position in respect of an order, if placed shortly prior to the OSPA, is likely to be an absolute ceiling on its potential profitability at the time of the OSPA. Therefore, if an arbitrage position is notionally only marginally profitable or unprofitable at the time the order is placed, then, according to ASIC, it is likely that that position will be unprofitable or even more unprofitable by the time of the OSPA.

The second pillar

  1. For the second pillar of its case, ASIC relied on Mr Morgan’s analysis of data regarding the historical volume of SPI Futures traded on the eight Serial Expiry Days prior to each of the Serial Expiry Days in question in this proceeding, which was undertaken to examine the likelihood—looking forward at the time of the placement of the Buy Orders (19 April 2012) and the Amended Sell Orders (other Serial Expiry Days)—of  Whitebox trading  by 10:08:45 (the earliest commencement of the Group 5 Rotation) sufficient SPI Futures, at prices prevailing at the time of the orders, to match the securities the subject of those orders. In undertaking his analysis, Mr Morgan adjusted the historical figures to take into account the observed volumes of SPI Futures traded on the Serial Expiry Days in question up to the time of placement of the impugned orders.
  2. In substance, Mr Morgan’s analysis was that, on each of the Serial Expiry Days in question, it would have been necessary for Whitebox to have traded a large percentage of the average total SPI Futures traded by all market participants on previous Serial Expiry days (adjusted as I have said) in order to match the Buy Orders and the Amended Sell Orders for XJO Securities prior to 10:08:45, namely:
(a) 62.58% for 19 April 2012;
(b) 37.94% for 17 May 2012;
(c) 109.92% for 19 July 2012;
(d) 336.57% for 16 August 2012, and
(e) 89.29% for 18 October 2012.
  1. Mr Morgan opined that, as the 19 April 2012 trading position was a positive arbitrage, involving the sale of SPI Futures, it would have been reasonable for an arbitrageur to have expected that it was likely that the sale of the required volume of futures, in the period just prior to the placement of the second set of Buy Orders and up to 10:08:45, would likely have had a substantial downward impact on the traded price of the futures.
  2. The trading position on the other Serial Expiry Days in question were negative arbitrages, involving the acquisition of SPI Futures. Mr Morgan opined that:
(a) for 17 May 2012, it would have been reasonable for an arbitrageur to have expected that it was likely that the acquisition of the required volume of futures, in the period just prior to the placement of the first May Amended Sell Orders and up to 10:08:45, would have had an upward impact on the traded price of the futures;
(b) for 19 July 2012, it would have been reasonable for an arbitrageur to have expected that it was likely that the acquisition of the required volume of futures, in the period just prior to the placement of the first July Amended Sell Orders and up to 10:08:45, would have had a significant upward impact on the traded price of the futures;
(c) for 16 August 2012, it would have been reasonable for an arbitrageur to have expected that it was likely that the acquisition of the required volume of futures, in the period just prior to the placement of the first August Amended Sell Orders and up to 10:08:45, would have had a significant upward impact on the traded prices of the futures; and
(d) for 18 October 2012, it would have been reasonable for an arbitrageur to have expected that it was likely that the acquisition of the required volume of futures, in the period just prior to the placement of the first October Amended Sell Orders and up to 10:08:45, would have had a significant upward impact on the traded prices of the futures.
  1. ASIC submitted that Mr Morgan’s analysis showed that an arbitrageur considering whether SPI Futures could be traded in the volumes required to offset the impugned orders for the XJO Securities would have perceived that it was likely that, by trading such volumes, he or she would move the price of the SPI Futures against himself or herself.
  2. The paragraphs of Mr Morgan’s first report which expressed these opinions were subject to a particular challenge as to their admissibility. They are built upon the following paragraph in Mr Morgan’s report:
    1. A trader will have a general idea of the likely impact that the purchase or sale of a quantity of a listed security that they are contemplating trading would have on the traded prices of that security. In my opinion it is reasonable to expect to be able to trade up to a third of market volumes of a listed security in-line with the market, that is with little to no impact on traded prices. That being said, if trading exactly 33% of traded volumes, one would expect to be getting very close to the point that traded prices could be seen to be impacting traded prices.
  3. On the fifth day of the hearing (6 July 2018), the defendants objected to the admissibility of, and I rejected, this paragraph as a matter of form because, although the second sentence commenced with an apparent expression of Mr Morgan’s opinion, it was not clear to me, on reading the paragraph as a whole, that the proffered opinion was based wholly or substantially on specialised knowledge.
  4. The expression of this opinion was extended in paragraphs 130 and 131 of the report to the degree of impact that trading certain percentages of SPI Futures (expressed as ranges), relative to total market volume, would likely have on the particular price at which the futures could be traded over a future period of time. Thus, by reference to certain percentage ranges, Mr Morgan purported to express an opinion as to whether particular market volumes of traded SPI Futures would be likely to have little to no impact on traded prices (0% to 33%), some impact on traded prices (33% to 50%), substantial impact on traded prices (50% to 66%) or significant impact on traded prices (66% to 100%). It is convenient to refer to the expression of these ranges as Mr Morgan’s market impact table. On the sixth day of the hearing (9 July 2018), the defendants objected to the admissibility of these paragraphs of the report, and to other paragraphs which were based on these opinions.
  5. After I rejected paragraph 129, ASIC applied for leave to supplement Mr Morgan’s evidence in chief to support paragraph 129. I granted that leave. On 9 July 2012, ASIC filed a supplementary affidavit from Mr Morgan sworn on 8 July 2018 (Mr Morgan’s July affidavit or the July affidavit) in which he gave greater detail of his experience in securities index arbitrage trading (to support his claim to specialised knowledge) and the basis on which he expressed the opinions in paragraphs 129 to 131 of his first report.
  6. Thus, by the morning of the sixth day of the hearing, the question was whether certain identified paragraphs of Mr Morgan’s first report should be admitted or rejected. This included re-visiting the ruling I had previously made on paragraph 129.
  7. The defendants accepted (indeed, suggested) that Mr Morgan’s July affidavit should be taken as having been read and that my ruling on the challenged paragraphs (including re-visiting paragraph 129) should be deferred until the delivery of these reasons, given their view that conducting a voir dire might not be a practical way forward, and given that they intended to cross-examine Mr Morgan on his expertise in any event. ASIC did not object to this proposal.
  8. I am satisfied that Mr Morgan possesses the specialised knowledge to express the opinions in paragraphs 129 to 131 of his first report, and that the opinions so expressed are based substantially on that specialised knowledge. In light of the matters stated in his first report, as supplemented by the July affidavit, Mr Morgan’s opinions can be seen as an application of well-understood principles of supply and demand borne out by his own observations of the market when trading SPI Futures, including the effects of other traders’ order placements, and his discussion with other traders in SPI Futures, including instructions given to him by senior traders. The substance of Mr Morgan’s evidence was that, as a trader increases demand (in this case, increases the number of buy orders for SPI Futures) or increases supply (in this case, increases the number of sell orders for SPI Futures), in substantial size, the price at which the market will form a trading price will be affected. All things being equal, significantly increasing the volume of buy or sell orders will, correspondingly, affect the trading price for the futures.
  9. Mr Morgan made clear in his first report that the expected impact of an order based on its size relative to market volumes cannot be assumed to be linear or to have discrete boundaries. Nevertheless, his opinion was that one could gain a sense of the likely impact of the orders by reference to the broad ranges to which he referred. Empirical knowledge of this kind is plainly captured in the concept of “specialised knowledge” as used in s 79(1) of the Evidence Act 1995 (Cth). That said, the limitations inherent in the way in which the opinion has been formed, and the language in which the opinion has been expressed, must be taken into account when according the weight that it should be given in the context of the question to be decided.
  10. For these reasons, I will admit paragraphs 129 to 131 of Mr Morgan’s report. As I understand the agreed position between the parties, the effect of this ruling is that the other paragraphs of Mr Morgan’s first report that are dependent on paragraphs 129 to 131 will, similarly, be admitted.
  11. Objection was also taken to paragraph 132 of Mr Morgan’s first report. This paragraph is of a slightly different character to paragraphs 129 to 131. Paragraph 132 expresses the opinion that it is common practice and reasonable to use historical average volumes to reflect expected market volumes. On the evidence before me, I am satisfied that Mr Morgan has the specialised knowledge to express this opinion and that his opinion is wholly or substantially based on that knowledge. I will, therefore, admit paragraph 132. I note, however, that the defendants challenge the weight that should be given to this opinion having regard to Mr Morgan’s limited experience and the evidence adduced by them on this topic through Professor Frino.
  12. In respect of its second pillar, ASIC also relied on an analysis by Mr Morgan of the likelihood—looking forward at the time of the placement of the Buy Orders (19 April 2012) and the Amended Sell Orders (the other Serial Expiry Days)—of whether it would have appeared that Whitebox could have profitably traded by 10:08:45 (once again, the earliest commencement of the Group 5 Rotation) sufficient SPI Futures to match the impugned orders for XJO Securities. Mr Morgan’s analysis was based in part on the historical, adjusted data to which I have referred and trading data on the Serial Expiry days in question.
  13. Mr Morgan opined that:
(a) for 19 April 2012 (when there were generally lower traded volumes of SPI Futures relative to the calculated historical average volume and, according to Mr Morgan, an expectation, looking forward from the relevant time, of slightly less liquidity in SPI Futures than would be experienced on an average Serial Expiry Day), selling the required volume of futures would be likely to cause the average selling price of the futures to be lower and so cause the posited securities index arbitrage trading to be unprofitable looking forward at the time of the second set of Buy Orders (Mr Morgan was unable to make an assessment looking forward from the time of placement of the first set of Buy Orders);
(b) for 17 May 2012 (when there were generally higher traded volumes of SPI Futures relative to the calculated historical average volume and, according to Mr Morgan, an expectation, looking forward from the relevant time, of a liquidity in SPI Futures that would be in-line with or greater than that which would be experienced on an average Serial Expiry Day), acquiring the required volume of futures would be likely to cause the average selling price of the futures to be higher and so cause the posited securities index arbitrage trading to be more unprofitable (it having been observed by Mr Morgan that, at the time of the placement of each of the May Amended Sell Orders, the implied profitability of a negative arbitrage, as calculated from traded and indicative match prices and the traded level of SPI Futures, was always a negative figure, and hence unprofitable);
(c) for 19 July 2012 (when there were generally lower traded volumes of SPI Futures relative to the calculated historical average volume and, according to Mr Morgan, an expectation, looking forward from the relevant time, of less liquidity in SPI Futures than would be experienced on an average Serial Expiry Day), acquiring the required volume of futures would be likely to cause the average selling price of the futures to be higher and so cause the posited securities index arbitrage trading to be unprofitable;
(d) for 16 August 2012 (when there were generally lower traded volumes of SPI Futures relative to the calculated historical average volume and, according to Mr Morgan, an expectation, looking forward from the relevant time, of substantially less liquidity in SPI Futures than would be experienced on an average Serial Expiry Day), acquiring the required volume of futures would be likely to cause a substantial upward impact on the traded prices for the futures and result in the posited securities index arbitrage trading to be unprofitable at the time of placement of the initial August Amended Sell Orders but profitable at the time of placing the subsequent August Amended Sell Orders, although the profitability would be substantially less profitable than that shown by a calculation of implied profitability he performed; and
(e) for 18 October 2012 (when there were generally lower traded volumes of SPI futures relative to the calculated historical average volume and, according to Mr Morgan, an expectation, looking forward from the relevant time, of a liquidity in SPI Futures that would be in-line with that which would be experienced on an average Serial Expiry Day), acquiring the required volume of futures would be likely to cause the futures to trade at significantly higher prices and result in the posited securities index arbitrage trading to be more unprofitable at the time of the placement of the initial October Amended Sell Orders (it having been observed by Mr Morgan that, at the commencement of the placement of the October Amended Sell Orders, the implied profitability of a negative arbitrage would be a negative value) but profitable at the time of placing the subsequent October Amended Sell Orders, although substantially less profitable than that shown by a calculation of implied profitability he performed.
  1. ASIC submitted that the effect of Mr Morgan’s analysis was that, at least at the time of placing the second set of Buy Orders, placing the Amended Sell Orders before or at the time the Reductions began, and placing the Incremental Reductions for July, a reasonable arbitrageur could not have expected to trade profitably the required amount of SPI Futures to establish a matched arbitrage position.

The third pillar

  1. For the third pillar of its case, ASIC relied on Mr Morgan’s analysis of Whitebox’s trading patterns in relation to SPI Futures on the Serial Expiry Days in question against certain benchmarks he posited, including:
(a) the rate (expressed as a constant rate) at which SPI Futures would need to have been sold/acquired to match the Buy Orders (19 April 2012) or the Amended Sell Orders (other Serial Expiry Days) by 10:08:45;
(b) the SPI Equivalent of Whitebox’s orders for XJO Securities that ultimately entered the OSPA; and
(c) the rate (expressed as a constant rate) at which SPI Futures would need to have been sold/acquired to match Whitebox’s orders for XJO Securities that ultimately traded in the OSPA.
  1. Mr Morgan observed that:
(a) for 19 April 2012, an overall comparison of Whitebox’s trading against the benchmarks suggested that its trading and order activity were not commensurate with selling sufficient SPI Futures to match the Buy Orders by 10:08:45, although the actual rate at which SPI Futures were sold was substantially in line with the rate at which futures would need to have been sold (expressed as a constant rate) to match the actual value of the XJO Securities that traded in the OSPA;
(b) for 17 May 2012, an overall comparison of Whitebox’s trading against the benchmarks suggested that its trading and order activity were not commensurate with acquiring sufficient SPI Futures to match the Amended May Sell Orders by 10:08:45, although the actual rate at which SPI Futures were acquired (ignoring a spike in Whitebox’s acquisition) was substantially in line with the rate at which futures would need to have been acquired (expressed as a constant rate) to match the actual value of the XJO Securities that traded in the OSPA;
(c) for 19 July 2012, an overall comparison of Whitebox’s trading against the benchmarks suggested that its trading and order activity were not commensurate with acquiring sufficient SPI Futures to match the Amended July Sell Orders by 10:08:45, although the actual rate at which SPI Futures were acquired was substantially in line with the rate at which futures would need to have been acquired (expressed as a constant rate) to match the actual value of the XJO Securities that traded in the OSPA;
(d) for 16 August 2012, an overall comparison of Whitebox’s trading against the benchmarks suggested that its trading and order activity were not commensurate with acquiring sufficient SPI Futures to match the Amended August Sell Orders by 10:08:45, although, despite a rising trend in acquisition for a limited period of time, the actual rate at which SPI Futures were acquired was substantially in line with the rate at which futures would need to have been acquired (expressed as a constant rate) to match the actual value of the XJO Securities that traded in the OSPA; and
(e) for 18 October 2012, an overall comparison of Whitebox’s trading against the benchmarks suggested that its trading and order activity were not commensurate with acquiring sufficient SPI Futures to match the Amended October Sell Orders by 10:08:45, although the actual rate at which SPI Futures were acquired closely matched the rate at which futures would need to be acquired (expressed as a constant rate) to match the actual value of the XJO Securities that traded in the OSPA.
  1. Mr Morgan also observed the pattern of Whitebox’s trading in SPI Futures on the Serial Expiry Days in question, and the data on traded prices for those days to see, for example, the level of Whitebox’s “bids” (for buy orders) and “asks” (for sell orders) in relation to the market “spread”; the extent to which its trading in SPI Futures “crossed” the spread; and the timing of its orders. His examination led him to opine that the data does not indicate that, on 19 April 2012, Whitebox was seeking, by 10:08:45, to sell a substantially greater number of SPI Futures than it did, in fact, sell on that day and that, in relation to each of the other Serial Expiry Days in question, Whitebox was seeking, by 10:08:45, to acquire a substantially greater number of SPI Futures than it did, in fact, acquire on those days. ASIC submitted that this pattern of trading reveals that Whitebox was not seeking to trade SPI Futures to match the Buy Orders, the Amended Sell Orders and the Incremental Reductions.

THE CHALLENGE TO MR MORGAN’S EVIDENCE

Introduction

  1. The challenge to Mr Morgan’s evidence had a number of strands. There was the challenge to the admissibility of certain paragraphs of his first report. There was a challenge to the weight that should be placed on his evidence given the extent and level of his expertise in securities index arbitrage trading. There was a challenge to the validity of his analyses and, hence, whether his opinions, based on those analyses, should be accepted. This section of my reasons deals, principally, with the last matter, although my conclusions are affected by the second matter, namely the weight to be given to Mr Morgan’s evidence. This challenge was mounted through Professor Frino’s evidence.

Professor Frino’s evidence

  1. In connection with the second pillar of ASIC’s case, Professor Frino disagreed with Mr Morgan’s evidence and analysis as to the likely impact that the size of an order for SPI Futures, relative to market volumes, would have on the price of those futures over time, and on the question of whether it would have appeared that Whitebox could have profitably traded by 10:08:45 sufficient SPI Futures to match the impugned orders for XJO Securities.
  2. Professor Frino considered Mr Morgan’s market impact table to be overly simplistic and not to be an accurate or reliable general statement of trading effects. Professor Frino said that due to the complexities of order execution and the volatile nature of traded prices, there is not a meaningful relationship between the size of an order and the price impact of the order or price movements around that order, as Mr Morgan had stated. In this connection, Professor Frino pointed to a research paper he co-authored in 1996 (admittedly in respect of trading in securities) which concluded that the size of large trades explains no more than 2% of the variation in the market impact of trades.
  3. To test Mr Morgan’s statements, Professor Frino carried out an analysis using ASX24 trading log data for 2012 and 2013. This analysis showed that there were large variations in the price impact of buy and sell trade packages (the total volume of SPI Futures bought and sold over 5minute time intervals between 9:50 and 10:10) of similar size executed by a given broker. It also illustrated that there is no clear relationship between the size of trades and the price impact of each trade package in the adjacent order size ranges posited by Mr Morgan.
  4. As to the last matter, Professor Frino considered whether there was a statistically significant difference between the price impacts of the adjacent ranges. From the data he analysed, he concluded that there was no evidence that the market impact of trades which represent 33% to 50% of trading volume are associated with a statistically higher price impact than trade packages which represent 0% to 33% of trading volume. He said that a similar conclusion can be drawn for trade packages which represent 50% to 60% of trading volume and those which represent 66% to 100% of trading volume.
  5. Overall, Professor Frino concluded that the ranges adopted in Mr Morgan’s market impact table cannot reasonably or accurately be said to classify the likely impact that orders would have on market prices for SPI Futures over a period of time. That said, Professor Frino accepted in cross-examination that, in a purely or completely static market, where one assumes everything is constant and all else is equal, larger orders will, theoretically, have larger price impacts. He did not accept, however, that this idealised or theoretical world reflected the operation of the market for SPI Futures in which trading is dynamic and trading times can be measured in micro- and nanoseconds. Professor Frino gave this evidence:
... I have been doing more than 20 years of research in this area looking for a meaningful and material relationship between order size and the price impact of orders on the marketplace and I have not found a meaningful and material relationship between the two.
  1. Professor Frino was challenged on this statement and, relatedly, on the validity of his statement (referred to above) that the size of trades explains no more than 2% of the variation in the market impact of trades. In challenging the latter statement, Professor Frino was taken to other literature he had co-authored in respect of trading in SPI Futures, which (it was suggested) was more pertinent. Although much was made in cross-examination, and subsequently in submissions, about what could be drawn from this other literature and the answers which Professor Frino gave, I do not think that these challenges to Professor Frino’s evidence materially undermine the thrust of his evidence or, more importantly, advance the acceptance of Mr Morgan’s market impact table. Indeed, in the end, this challenge secured no more than a heavily qualified acceptance that, at a theoretical level, larger volume trades in SPI Futures might have a larger impact on the price of those futures.
  2. More significant was ASIC’s challenge to Professor Frino’s analysis of the ASX24 trading data log for 2012 and 2013, which led him to opine that there is no clear relationship between the size of trades and the price impact of each trade in the adjacent order size ranges posited by Mr Morgan. ASIC submitted that Professor Frino’s analysis was fundamentally flawed because Mr Morgan’s opinions were concerned with the price impact of a single order and Professor Frino’s analysis aggregated orders by the same broker over five-minute time intervals. The point of this challenge was ASIC’s assertion that Professor Frino’s analysis did not test the propositions in Mr Morgan’s market impact table because Professor Frino’s analysis was likely to have captured packages of orders which, purposely, had been broken up to avoid the very price impacts about which Mr Morgan was opining. The substantial answer to this criticism is that Mr Morgan’s market impact table did not, itself, say anything about the number of contracts that might underpin the orders falling within the ranges he selected. I do not accept that Professor Frino’s analysis is not apposite to test the soundness and reliability of the opinions expressed in Mr Morgan’s market impact table.
  3. In connection with the second pillar of ASIC’s case, Professor Frino also criticised Mr Morgan’s forecasting method using average historical volumes of traded SPI Futures to determine the likely availability of SPI Futures to be traded on a given Serial Expiry Day. Professor Frino described this as a “naïve average model” that ignored “the rich set of information cues” that traders have at their disposal (such as the likely price volatility of the underlying securities over which the futures contract is written, the preceding days trading activity, activity in the roll over market, and visible volumes on the particular trading day) which, according to Professor Frino, a trader would typically take into account when determining a likely level of trading. In short, Professor Frino said that traders would not be making decisions about the extent of their trading in SPI Futures by reference only or primarily to the historical average of futures that had been traded on previous days.
  4. Professor Frino said:
In my experience, futures traders are constantly assessing the futures market and the events that are happening in the futures market and elsewhere, to assess trading opportunities. Their expectations and intentions for their trading and their approach to trading will likely change as information becomes available to them. Mr Morgan’s approach appears to proceed on an assumption that at the beginning of the periods in which he has been referred to, the Whitebox trader’s intentions and expectations are set and his trading strategy proceeds accordingly. I consider such an assumption to be inconsistent with how futures are traded in the market. ... [P]atterns of trading in the market may fluctuate depending on whether the market is liquid or perceived to be liquid. It is usual to see large volumes of futures traded in short periods and not in others with futures traders choosing to leave trading until the last possible time or to times when they see or are expecting to see increased liquidity.
  1. Further, according to Professor Frino, it cannot be expected that, on any given day, the liquidity of the market for SPI Futures will mirror the historical average volumes of the futures traded. In this connection, Professor Frino carried out an analysis of the volumes of SPI Futures traded on every Serial Expiry Day since 15 July 2010 to calculate, on each day, the volume of futures traded between 9:55 and 10:10. For each Serial Expiry Day, he then used the previous eight Serial Expiry Days to “forecast” the volume traded between 9:55 and 10:10. He compared the forecast with the volumes actually traded on a given day. From this, Professor Frino concluded that there is a poor relationship between actual traded volumes and historical average volumes, with actual traded volumes deviating from forecast volumes by more than 1,000 contracts or 30% on various days. His analysis also illustrated that actual traded volumes can at times be equal to either the maximum or the minimum volumes traded in the previous eight Serial Expiry Days.
  2. In light of his analysis, Professor Frino concluded that it is not reasonable to use historical average volumes of traded SPI Futures to forecast expected volumes. Contrary to Mr Morgan’s evidence, he also said that, based on his experience, it is not common practice for traders to do so given the other information cues available to them.
  3. Professor Frino also criticised related aspects of Mr Morgan’s analysis based on average historical volumes of traded SPI Futures.
  4. First, Professor Frino said that he disagreed with Mr Morgan’s “assumption” that the volume of SPI Futures traded during the course of the morning would evolve in a linear and completely predictable way, such that the volume at the start of the morning would necessarily correlate with the volume later in the morning. This observation by Professor Frino is also relevant to that part of Mr Morgan’s evidence that underpins ASIC’s third pillar, which I discuss below. For present purposes, it is sufficient to note that, within the particular time periods considered by Mr Morgan on each Serial Expiry Day in question, Professor Frino pointed to periods when traders may have viewed the potential liquidity of SPI Futures differently to the way presented by Mr Morgan. In other words, there were periods within Mr Morgan’s examined periods that showed the liquidity in SPI Futures to be significantly greater than that revealed by historical averages (assuming historical averages to be a reasonable forecast of actual volumes to be traded on the day).
  5. Secondly, Professor Frino criticised Mr Morgan’s adoption, in his analysis, of a cut-off time of 10:08:45, which Professor Frino considered to be artificial. To illustrate his point, Professor Frino extended Mr Morgan’s analysis on the assumption that, on the Serial Expiry Days in question, Whitebox extended its trading in SPI Futures by up to one hour after the ASX24 opened for trading (i.e., up to 10:50). This time was chosen because, in Professor Frino’s experience, there can be extensive trading activity in SPI Futures on Serial Expiry Days well after the markets for SPI Futures and for XJO Securities have opened. The apparent rationale of this approach is that it can be assumed that futures traders will (or might) avail themselves of this liquidity in the market, should they wish to do so or need to do so. I pause to note that whether Whitebox would have done so, in light of its stated position on its trading strategy, is another matter.
  6. Professor Frino’s analysis shows that by extending Mr Morgan’s cut-off time by approximately 42 minutes, but otherwise adopting his methodology (including his market impact table), a different conclusion about price impacts is evident. Indeed, on this analysis, Mr Morgan’s analysis breaks down because the size of all orders relative to volume up to 10:50 on the previous eight Serial Expiry Days are all 33% or less (i.e., according to Mr Morgan, all order sizes that would have little to no impact on traded prices). In Professor Frino’s words:
... even if one were to accept the approach used by the Morgan Report to determine whether the relative size orders could be executed (ie. a naïve averaging approach and assumed market impact table) but assume that Whitebox had up to 10:50 to execute its orders, the conclusions drawn by the Morgan Report cannot be sustained.
... On this basis, the Morgan Report would have concluded that it is likely that the specified quantity of futures could have been acquired at prices that prevailed immediately prior to the start of order execution.
  1. Thirdly, Professor Frino observed that Mr Morgan’s analyses and opinions expressed with respect to the likelihood that, by 10:08:45 on the Serial Expiry Days in question, Whitebox could have profitably traded sufficient SPI Futures to match the Buy Orders (19 April 2012) or the Amended Sell Orders (other Serial Expiry Days), carried forward the assumptions and deficiencies he had identified. Further, the implied profit figures that Mr Morgan had calculated were expressed in index points. Mr Morgan did not express an opinion on how his predicted price effects on Whitebox’s hypothetical trading would translate into index points. For this reason, Professor Frino felt unable to comment further on this aspect of Mr Morgan’s evidence, beyond saying that a substantial movement in the futures price may have been necessary in order to make the hypothetical trades unprofitable.
  2. As to ASIC’s third pillar, Professor Frino returned to what he saw as the assumption underlying Mr Morgan’s analysis in respect of the rate at which Whitebox sold or acquired SPI Futures on the Serial Expiry Days in question: Whitebox should have traded SPI Futures at a relatively constant rate from open of trading to 10:08:45 in order to match the Buy Orders (19 April 2012) or the Amended Sell Orders (other Serial Expiry Days). Professor Frino described this as a “simplistic strategy” which ignored the fact that the liquidity of the futures limit order book changes throughout the morning. He said:
9.4 It is widely known that the liquidity of the limit order book of any market changes through time, and therefore a strategy which trades a proportion of the expected volume through time may be followed to potentially reduce market impact (Bialkowski et. Al (2008)). Such a strategy, which is widely used, is known as a VWAP (volume weighted average price) strategy where a trader seeks to minimise market impact by trading a constant portion of the total volume expected to be traded in each time interval [Bialkowski et. Al (2008)] thereby trading a large volume when the market is liquid and a smaller volume when the market is less liquid. One would not therefore usually see a futures trader buying or selling of futures in a constant linear pattern.
9.5 As a general rule, the liquidity in a future market does fluctuate and is not constant. Fong and Frino (2001) document that the liquidity of stock index futures is relatively higher all else being equal when the underlying market is open. This implies that SPI futures volume should increase as the time of day approaches the set of opening rotations on ASX.
(Emphasis in original)
  1. Professor Frino performed an analysis of the average volume of SPI Futures traded minute-by-minute in the period 9:55 to 10:09 (straddling 10:08:45) on each of the 14 Serial Expiry Days from 15 July 2010 (the first Serial Expiry Day ever) to 16 February 2012 (the last Serial Expiry Day before 19 April 2012). His analysis illustrated that the liquidity in SPI Futures is lowest between approximately 9:55 and 9:58 before it commences increasing until the Group 1 Rotation at approximately 10:00, and that the liquidity of the market is much higher from 10:00 to 10:09 than from 9:55 to 10:00. Professor Frino argued that the implication of this pattern is that a trader seeking to trade a given order, while at the same time minimising the market impact of the trading, should seek to trade small quantities of SPI Futures earlier in the morning (prior to 10:00) and significantly larger quantities later in the morning (after 10:00). He said:
9.9 The trading activity in the futures market is not such that allows one to assume that a futures trader, seeking to trade a certain number of futures, should have reached [a] particular milestone of volumes traded by certain times within that period based on a constancy analysis of how futures would be traded over that period. That is not what happens in the futures market.
9.10 It is perfectly consistent with what happens in the market to see large volumes of futures traded in very short periods and none in others. A futures trader seeking to trade a particular volume of futures may choose to leave any futures trading until the last possible time or when he expects to see, or is watching to see, increased liquidity.
  1. Professor Frino considered Mr Morgan’s “run-rate” approach to be flawed because it assumed that Whitebox should be trading SPI Futures in a somewhat linear and constant fashion over a defined period. Because he did not agree with this assumption, Professor Frino argued that comparisons between such a pattern and actual futures executions should not be relied upon to form a judgment as to whether the trading activity was intended to trade any particular volume of SPI Futures by any particular point in time. He said that the level and extent of futures trading activity will fluctuate with the changing perceptions of the trader of the market conditions being experienced. The trader’s expectations will also likely change as assessments are made and re-made as to the anticipated liquidity of the futures. Further, as I have already noted, Professor Frino did not agree that the futures trades had to be completed by no later than 10:08:45.
  2. Professor Frino concluded:
10.38 I accordingly do not believe that adopting the analysis applied by Mr Morgan ... enables any sound conclusions to be made about the Whitebox trader’s intentions or expectations as to volumes of SPI that may be traded over each defined future period; or whether the  Whitebox trading  activity was commensurate or not with selling or buying a sufficient number of SPI contracts by some pre-determined point in time.
  1. With respect to Mr Morgan’s observations about the pattern of Whitebox’s trading in SPI Futures on the Serial Expiry Days in question, Professor Frino noted that, with the exception of 19 April 2012, Mr Morgan’s tables start at a time when Whitebox had already placed orders for SPI Futures (with the consequence that one does not have a picture of trading beforehand); none of the tables identify the volume of orders being placed; none of the tables provides information as to the activity of other participants in the market at the time or provide any information as to the liquidity that might have been seen; and the tables cover periods of 11 to 13 minutes, within which Mr Morgan seems to assume that Whitebox had the intention to acquire a specific number of SPI Futures, which carried through to the whole period being reviewed.
  2. Professor Frino said:
10.45 For reasons that I have stated above, I do not believe that any trader would be trading in such a period with a linear expectation as to volumes to be acquired. To the contrary, I believe that a trader would have expected a small amount to be traded, with the uplift in volume expected when the underlying market opened ... Trading decisions and expectations would be made and change in reaction to evolving market conditions and other information available to the trader. I would expect that over the period, depending on changes in perceived liquidity, the Whitebox trader’s expectations or intentions regarding the number of SPI contracts to trade or how to be traded, i.e. how aggressively or passively, may have changed.

CONCLUSIONS ON MR MORGAN’S EVIDENCE

The first pillar

  1. The first pillar is that, at the time the Amended Sell Orders were placed on 17 May, 19 July, 16 August and 18 October 2012, the arbitrage position established by Whitebox would have appeared to have been unprofitable, assuming the market to have opened at those times. This is a critical assumption.
  2. In my view, this pillar is of very limited assistance to ASIC’s case. No witness suggested that it provided any particular, let alone useful, insight into Whitebox’s conduct. It speaks of nothing more than a snapshot, which Mr Morgan agreed was arbitrary. He accepted that Whitebox’s Amended Sell Orders should be viewed in the context of the auction time lines.
  3. ASIC’s response to its own expert’s apparent rejection of the usefulness of this snapshot was to contend that if an order is either marginally profitable or unprofitable at the time it is placed then, having regard to the phenomenon of Basis Decay, the arbitrageur could expect that the order will be unprofitable or even more unprofitable by the time of the OSPA.
  4. I do not accept that contention or the reasoning on which it is based. It is simplistic. It assumes away the dynamics of Pre-Open Phase. These dynamics are the very thing that the experts, by their rejection of a snapshot approach, acknowledged should not be assumed away.
  5. Orders placed in the Pre-Open Phase do not trade and cannot trade until the market opens. The Pre-Open Phase commences at 7:00 am, approximately three hours before Normal Trading commences. It is a period in which traders and investors commonly place orders for the purpose of gaining time/price priority in the Central Order Book. These orders can be amended and cancelled up to the time when they participate in the OSPA. Whether any of these orders are maintained or amended or cancelled, will depend on how conditions vary and evolve as the morning progresses.
  6. It is, of course, a period of time when arbitrageurs will seek to identify and capture mispricing. All things being equal (in particular, a given level of mispricing), the larger the order that is placed, the greater the potential profit that can be made. But conditions can change markedly and, when they do, arbitrageurs will tailor their position size to meet these changing conditions having regard to their own changing perceptions of the level of mispricing and the level of profit that can be captured, and their own appetite for risk. Within seconds or fractions of seconds, potentially profitable arbitrage positions, reflected in cash orders placed in the Pre-Open Phase, can become potentially unprofitable, obliging an arbitrageur to cancel or amend its orders.
  7. The dynamism of order activity in the Pre-Open Phase is reflected in one of Professor Aitken’s analyses to which I will make further reference below. In essence, he analysed orders placed for XJO Securities between 9:00 and the OSPA on “20 Serial Expiry Days” and found that around (slightly below) 50% of the orders were amended or deleted and that 52% of the amendments that reduced volume were placed during the last five minutes before the OSPA.
  8. Further, it should be recognised that arbitrageurs place orders to pursue the possibility of profit arising from mispricing. Orders may well be placed in contemplation of a perceived opportunity that the order will be or will become profitable based on the arbitrageur’s assessment of the ordering activity or trends in ordering activity in the Pre-Open Phase.

The second pillar

  1. The second pillar is that, looking forward at the time of the Buy Orders (19 April 2012), and the Amended Sell Orders and the Incremental Reductions (other Serial Expiry Days), the defendants could not have expected that sufficient SPI Futures could be traded profitably by 10:08:45. The time limitation of 10:08:45 was one that Mr Morgan was instructed to adopt when carrying out the analyses that ASIC relies on to support this pillar. It is not a limitation that he, himself, imposed; nor is it a limitation whose appropriateness he expressly considered.
  2. In cross-examination, Mr Morgan accepted that, in carrying on securities index arbitrage trading, an arbitrageur trading SPI Futures in the Pre-Open Phase will carry the risk of an unhedged position until at least the commencement of the Group 5 Rotation. He accepted that if an arbitrageur wished to minimise the unhedged risk, it would be rational to wait as long as possible before the commencement of the OSPA to acquire SPI Futures so as to minimise the time that the risk is carried.
  3. Mr Morgan also accepted that, if an arbitrageur was willing to accept unhedged risk up to the time of commencement of the Group 5 Rotation, there was no reason why the arbitrageur would not also accept unhedged risk after that time. He accepted that whether or not the arbitrageur would run such a risk was dependent on a number of matters which could include the arbitrageur’s mandate; the size of Basis (or, in other words, the extent of the profit opportunity, which might mean that the arbitrageur not only has an incentive to bear that risk but also a degree of flexibility to take on the risk); questions of subjective judgment; and factors that have nothing to do with a particular trade (such as the desire to offload accumulated inventory or to move large orders because of a liquidity event).
  4. This evidence supports Professor Frino’s contention that the selection of 10:08:45 in Mr Morgan’s analyses is an arbitrary cut-off point. The artificiality of this cut-off is illustrated by Professor Frino’s extended analysis using Mr Morgan’s methodology (including Mr Morgan’s market impact table) which shows that, depending on the time of the cut-off point chosen, trading certain volumes of SPI futures may well have little or no impact on SPI Futures prices. Put another way, adopting Mr Morgan’s methodology and assuming that it is valid to use his market impact table (a matter in contention but, in this exercise, assumed in ASIC’s favour), an arbitrageur might well have concluded, as Professor Frino argued, that it was likely that the volume of SPI Futures required to match the Buy Orders (19 April 2012), and the Amended Sell Orders and the Incremental Reductions (other Serial Expiry Days), could have been traded at prices prevailing immediately prior to the execution of those orders. This simply depended on the cut-off point chosen for trading the required volume of SPI Futures.
  5. Consistently with Professor Frino’s evidence, Mr Morgan also accepted that, based on data he himself had used, the period from 9:59:45 to 10:09:15 (once continuous trading in XJO Securities had begun) was one in which there were, generally, higher volumes of SPI Futures traded than in earlier periods.
  6. The rationale for choosing the cut-off point as 10:08:45 is Whitebox’s stated trading strategy of not carrying unhedged risk into intraday trading. However, the evidence shows that this trading strategy was not always observed by Whitebox in practice. In this connection, the defendants relied on the auction timeline of Whitebox’s trading on 16 February 2012 (not one of the Serial Expiry Days in question in this proceeding). On that day, Whitebox had not acquired the SPI Futures to match its orders for XJO Securities by the commencement of the Group 5 Rotation. It had acquired only 150 SPI Futures to cover its sell orders for XJO Securities standing at 171.2 SPHE (i.e., approximately 87% of the SPI Futures required to effect the hedge). Mr Morgan accepted that, in order to address the unhedged position (of approximately $2 million), Whitebox needed to buy XJO Securities which it had previously sold or acquire additional SPI Futures it had not already acquired. Mr Morgan accepted that it was reasonable to carry this unhedged position (approximately 20 SPI Futures) into open trading given the Basis indicated (-12.1 SPHE)—I would add, at a time when, generally speaking, there are higher volumes of SPI Futures traded than in earlier periods of the trading day.
  7. I am not persuaded that the cut-off point of 10:08:45, which is integral to Mr Morgan’s analyses, should be accepted as a barrier against which to assess Whitebox’s expectations of the profitability of its trading on the Serial Expiry Days in question. While I certainly do not disregard Whitebox’s statement of its trading strategy, I do not consider it to be other than a statement of general approach which might yield to the trading conditions which Whitebox faced on a particular day. This is illustrated by its trading on 16 February 2012. I accept the possibility—as a real possibility—that, on any given trading day, including the Serial Expiry Days in question, Whitebox would be prepared to trade SPI Futures after the market for XJO Securities had opened—for example, to address an unhedged position or to capture a profit it perceived was capable of being captured without discarding entirely its generally expressed aversion to the risk involved in intraday trading. Much would depend on how Whitebox actually perceived trading conditions to be on the day in question, and at the point in time in which it became necessary to make a decision whether to continue trading futures through and after the OSPA. Whitebox’s order activity and trading on 16 February 2012 show that it was not as risk averse as, say, the Schedule A document or the 2009 document, on first reading, suggest. This finding is not gainsaid by the fact that, on the Serial Expiry Days in question, Whitebox did complete the futures leg of its trading before 10:08:45. That shows nothing more than that Whitebox had achieved a hedged position by that time. It says nothing specific about Whitebox’s expectations with respect to profit looking forward from the point in time that the Buy Orders (19 April 2012) and the Amended Sell Orders (other Serial Expiry Days) were placed.
  8. Thus, Mr Morgan’s model with a cut-off time of 10:08:45 for trading SPI Futures is an idealised one based on a stated trading strategy that was not always adhered to. Once the 10:08:45 barrier in Mr Morgan’s analysis is removed, then ASIC’s case in respect of the second pillar commences to unravel because, as Professor Frino’s analysis shows, a different picture of liquidity in SPI Futures can emerge using Mr Morgan’s methodology. This, in my view, is a weakness in the model which affects the weight I can place on it for the purpose of determining whether ASIC has proved, on the balance of probabilities, the purpose with which it says the defendants acted in placing the impugned orders on the Serial Expiry Days in question.
  9. There are, however, other more significant weaknesses exposed in Mr Morgan’s analysis which further undermine ASIC’s proof of the second pillar. As I have discussed, Mr Morgan’s opinion on likely profitability also involved his analysis of the likely price effects that would be brought about in respect of SPI Futures if Whitebox were to trade the volumes of futures required to match the Buy Orders (19 April 2012) and the Amended Sell Orders (other Serial Expiry Days) within the posited timeframes on the Serial Expiry Days in question. Mr Morgan’s analysis of the likely price effects depended on the validity of, and then the application of, his market impact table. While I accept, at the level of principle, the theory that underpins the expression of Mr Morgan’s opinions in the market impact table, I am unable to accept, with any real confidence, the ultimate expression of those opinions.
  10. First and foremost, I cannot help but be struck by the fact that the opinions expressed in the market impact table are based on what is, essentially, a subjective view (albeit said to be based on Mr Morgan’s experience) as to the appropriateness of the ranges that have been selected. Further, the likely market impacts are expressed in generalised and somewhat impressionistic terms—“little to no impact”; “some impact”; “substantial impact”; and “significant impact”, which lack the precision required to be useful in the present inquiry, particularly in inferring the defendants’ intentions. Even the adoption of “likely” impact, without elaboration, imports further imprecision and compounds the overall problematic nature of the table. The table is, of course, the key driver of Mr Morgan’s assessment of whether or not buying or selling the required hypothetical volumes of SPI Futures in a given time period would be “likely” to have an impact on the traded prices of SPI Futures and, ultimately, whether the hypothetical index arbitrage transactions based on the Buy Orders (19 April 2012) and the Amended Sell Orders (other Serial Expiry Days) could be expected to have been profitable.
  11. Secondly, my concerns are supported by Professor Frino’s conclusion that the ranges adopted in the market impact table cannot reasonably or accurately be said to classify the “likely” impact that orders would have on market prices for SPI Futures over a period of time. Professor Frino’s analysis provides an objective test which assists in gauging the reliability of the market impact table. I do not accept that ASIC’s criticism of Professor Frino’s analysis deprives it of utility for this purpose. Professor Frino’s analysis and conclusion strengthens the conclusion I have reached, independently of his evidence, that the market impact table is not reliable for the purpose for which Mr Morgan has used it, and for which ASIC seeks to rely on it.
  12. Mr Morgan’s opinion on likely profitability also involved recourse to average historical volumes of traded SPI Futures as a forecast of likely volumes of SPI Futures available to be traded on the Serial Expiry Days in question. It is from these forecasted volumes (adjusted in part for actual trading on the day in question) that Mr Morgan arrived at a percentage order size to which he then applied the likely market impacts expressed in the market impact table.
  13. I accept that historical volumes of traded orders do provide information that can be used as an input in assessing possible future volumes of such orders. GN 32 indicates this to be the ASX’s view. I repeat part of the extract quoted at [83] above:
The size of an index arbitrage order can cause imbalance to the existing market price levels. Prior to effecting an index arbitrage order the Participant should assess how much volume has been traded on previous expiries, and together with more recent volumes, assess whether the size of the order can be reasonably absorbed by the market given current trading conditions.
  1. This extract does not support the position that historical volumes alone can predict future volumes or provide the sole basis on which an assessment of intended order size (on the day) should be made. The extract makes clear, for example, that an assessment should take into account “more recent volumes” and “current trading conditions”.
  2. This is, in fact, one of the important points made by Professor Frino when he stated that the expectations and intentions of those trading in the futures market, and their approach to trading, will likely change as information becomes available to them, and that trading patterns will fluctuate depending on traders’ perceptions of the market at the time. Professor Frino’s analysis of the SPI Futures traded on every Serial Expiry Day since 15 July 2010 shows that there was, indeed, a poor relationship between actual traded volumes and historical average volumes. I accept, therefore, that historical average volumes of SPI Futures alone provide a poor forecasting tool of future volumes that might be available to be traded.
  3. Mr Morgan’s evidence was that it was common practice to use historical average volumes to reflect expected market volumes. Professor Frino disagreed. While I do not doubt the sincerity of Mr Morgan’s opinion, I am not satisfied on the basis of his evidence that, at the times relevant to this case, it was common practice to use historical average volumes to reflect expected market volumes if, by his evidence, Mr Morgan intended to convey that it was common practice to use this information alone to forecast future market volumes. In reaching this view, I take into account Mr Morgan’s relatively limited experience in securities index arbitrage trading. While I accept that he has sufficient expertise to support the admissibility of the opinions he has expressed, the weight I am prepared to attribute to those opinions, and my acceptance of them, is, as I have said, another matter. I take Professor Frino’s opinion into account when reaching a view on this question, recognising that Professor Frino has not himself engaged in securities index arbitrage trading. Nevertheless, his expertise arises, in part, from study, research and the observation of trading patterns in the futures market, and forms part of the evidential fabric from which I must decide this question. I have not, however, decided this question solely on the expression of Professor Frino’s contrary opinion.
  4. I should note that, in carrying out his analyses, Mr Morgan also had regard to what Whitebox would need to trade to match the relevant orders as a percentage of the maximum historical volume from among the eight preceding Serial Expiry Days. Professor Frino acknowledged that this addressed his criticism of using historical average volumes. However, the general criticism of using historical volumes remains.
  5. In light of these findings, I am not persuaded, on the evidence adduced, that the second pillar should be accepted. I express my conclusion in this way because the second pillar is premised on an acceptance that it is appropriate to evaluate the defendants’ likely expectations on profitability at one point in time (the placement of the Buy Orders (19 April 2012) and the Amended Sell Orders (other Serial Expiry Days)) looking forward to a fixed cut-off point (the earliest commencement of the Group 5 Rotation in the OSPA at 10:08:45). The limitation placed by this fixed point should not be accepted as simulating what the defendants’ actual expectations would have been, or were likely to have been, when the impugned orders were placed.
  6. The second pillar should not be accepted for two further reasons. First, the methodology on which it is based is not sufficiently reliable, for the reasons I have expressed above. It does not enable me to make sound conclusions on the balance of probabilities about the defendants’ expectations concerning the profitability of the impugned orders at the time they were placed.
  7. Secondly and relatedly, even though the methodology seeks to provide an objective benchmark against which it is said the defendants’ conduct should be assessed, I am simply not persuaded that it provides a model of the likely real-time perceptions, let alone expectations, of index arbitrageurs in general or, specifically, the defendants with respect to the profitability of trading particular orders. My conclusion in this regard is also informed, in part, by the fact that the second pillar also includes the vice of the first pillar in that it ignores the unfolding dynamics of the markets involved. The methodology adopted by Mr Morgan represents an idealised and somewhat abstracted and artificial model of perceived likely profitably that simply fails to account for those dynamics.
  8. There is a final matter to which I should refer. As I understand his evidence in respect of this pillar, Mr Morgan expresses opinions about the profitability of the Buy Orders (19 April 2012) and the Amended Sell Orders (other Serial Expiry Days) at the point in time they were placed. I base this conclusion on the text of his first report (regardless of the somewhat confusing questions he was asked). However, in its second pillar, ASIC includes reference to the profitability of the Incremental Reductions. In [195] of its written submissions, ASIC says that Mr Morgan considered profitability looking forward at the time of the Buy Orders, the Amended Sell Orders and the Incremental Reductions. But, so far as I can see, ASIC’s proposition is referenced only to Mr Morgan’s opinions about the profitability of the Buy Orders and the Amended Sell Orders. Even more confusingly, [196] of ASIC’s submissions appears to limit its contention with respect to Incremental Reductions to the Incremental Reductions on 19 July 2012. I mention these matters because I doubt that Mr Morgan even addressed the alleged profitability of the Incremental Reductions as such.
  9. I do accept on the balance of probabilities that, in making the Incremental Reductions, it would have been in the defendants’ contemplation that there was a real possibility that each successive reduction (other than the last made) might not trade and that the most likely scenario was that a hedge would be achieved with respect to the futures leg which, by then, Whitebox had traded. But this is far from accepting that the Incremental Reductions were placed with the purpose of manipulating the Opening Prices for XJO Securities.

The third pillar

  1. The third pillar is that, based on its actual trading in SPI Futures on the Serial Expiry Days in question, Whitebox did not seriously pursue the volume of futures necessary to offset, by 10:08:45, the Buy Orders, the Amended Sell Orders or the Incremental Reductions for XJO Securities, although Whitebox’s actual trading was consistent with seeking to match, by 10:08:45, the value of XJO Securities represented by the Subsequent Buy Orders (19 April 2012) and the final set of Reductions for each Rotation Group (other Serial Expiry Days).
  2. I accept that the analysis undertaken by Mr Morgan of Whitebox’s actual trading on the Serial Expiry Days in question shows, with hindsight, that this trading was consistent with Whitebox seeking to match, by 10:08:45, the value of XJO Securities represented by the Subsequent Buy Orders (19 April 2012) and the final set of Reductions for each Rotation Group (other Serial Expiry Days). The evidence also shows, as a matter of fact, and with hindsight, that Whitebox’s actual trading was undertaken at a somewhat linear rate.
  3. While I do not reject Professor Frino’s evidence that it should not be assumed that a particular arbitrageur would trade the futures leg of a securities index arbitrage at a relatively constant rate or that, because of the increased liquidity of SPI Futures at around the time of the Group 1 Rotation, it would be rational for the arbitrageur to acquire larger quantities of futures from around this time, I note that his evidence in chief was given without the benefit of Whitebox’s actual trading patterns on the Serial Expiry Days in question.
  4. ASIC submitted that it would be implausible to suggest that Whitebox would have tracked so closely to the linear rate shown in Mr Morgan’s charts, particularly on consecutive Serial Expiry Days, unless the defendants had in mind, from the start of trading on those days, a particular target (with respect to the volume of SPI Futures) to be achieved by 10:08:45. While I accept that the pattern of Whitebox’s trading on the days in question is consistent with such a target being in contemplation, I am not persuaded that it necessarily follows that the defendants had in mind, from the start of trading on those days, a particular target of SPI Futures to be achieved by 10:08:45, and I do not so find.

An aspect of Mr Morgan’s evidence

  1. It is convenient at this point to address a submission made by the defendants about the weight that should be given to an opinion expressed by Mr Morgan in the Joint Report he prepared with Mr de Kantzow and Professor Frino.
  2. In the course of the hearing (Whitebox No 5 at [15] – [37], especially at [32] – [36]), I admitted the following passage in the Joint Report:
Alternatively, the Whitebox traders would have been conscious of the market impact of removing all or part of their orders as evidenced by their use of underlying basis. With reference to the timelines, it is apparent that their orders were positioned and sized during the auction period to maximise the difference between basis and underlying basis and therefore the profit that could be realised through the market impact of the reduction of their stock orders. The sudden removal of the majority of their orders enabled the capture of this profit. That it was done just prior to the market opening limited the opportunity for other market participants to respond to this market impact before the relevant rotations occurred.
  1. The defendants submitted that this passage should be given no weight because it is not supported by reasoning and it would be unsafe to place reliance on it. The defendants also submitted that it was inconsistent with certain concessions made by Mr Morgan in cross-examination concerning the use of Underlying Basis and the rationality of arbitrageurs placing large cash orders in the Pre-Open Phase to position themselves to profit from mispricing.
  2. The concessions made by Mr Morgan in cross-examination put in context some of the remarks made in this passage to neutralise what might be seen as an implicit suggestion of wrongdoing based on the size of the orders placed. As to the balance, I am persuaded, on final analysis, that Mr Morgan’s statements about the defendants’ consciousness of the market impact of removing all or part of Whitebox’s orders, and their motives with respect to the timing of that removal (to limit market response), are really no more than his speculation about these matters. For these reasons, I am not prepared to place any weight on these parts of the quoted passage.

MR GRAVES’ EVIDENCE

  1. As I have noted, the fourth pillar of ASIC’s case is based on Mr Graves’ evidence and, in particular, the application of his “rules of thumb” to establish the likely immediate and continuing effects of the Cancellations and the Reductions on the Indicative Match Price (IMP) of the XJO Securities to which they related and on the Opening Prices (also referred to as the Match Prices) of those securities in the OSPA.
  2. In order to put Mr Graves’ evidence and the fourth pillar in the correct setting, it is necessary to say something about the ASX Algorithm.
  3. The ASX Algorithm determines the IMP of XJO Securities (in the Pre-Open Phase) and their Opening Prices (in the OSPA). It does this by a four-step process.
  4. The quantity of a security that would trade at a particular price is the executable volume of that security at that price. The first step in the operation of the ASX Algorithm is the determination of the price or prices at which the highest volume or Maximum Executable Volume (MEV) occurs. If there is only one price at which the MEV would occur, that price becomes the IMP (in the Pre-Open Phase) or the Opening Price (in the OSPA), as the case might be.
  5. In the event that the first step produces more than one possible result for the IMP or the Opening Price of a security, the algorithm proceeds to the second and, if necessary, the third and fourth steps to establish one, unique IMP or Opening Price for that security.
  6. The second step is based on the principle that the price should be the one that leaves the least quantity of securities in unfilled orders. In this step, the ASX Algorithm ascertains the volume of unfilled orders at the prices determined under the first step and determines which of those prices leave the least quantity of securities in unfilled orders. If the same quantity of securities in unfilled orders would arise at more than one of those prices, the algorithm moves to the third step.
  7. The third step is based on the principle that the highest potential price should be used if pressure is only on the buy side (there is a greater volume of Bids (or buy orders) than Asks (or sell orders)), and the lowest if the pressure is on the sell side (there is a greater volume of Asks than Bids). If the pressure is on both the Buy side and the Sell side, the algorithm moves to the fourth step.
  8. The fourth step is based on the principle that the price should be set with reference to the last traded price.
  9. The Indicative Match Volume (IMV) is the quantity of a security, calculated by the ASX Algorithm, that would be traded at the IMP if, hypothetically, the market were to open at the time of the calculation. The IMV is recalculated each time an order is entered, amended or cancelled.
  10. Mr Graves’ opinion was that the placement of the Cancellations (19 April 2012) and the Reductions (other Serial Expiry Days) was likely to have an impact on the Opening Prices of XJO Securities in the OSPA on the days in question, other than the following securities:
(a) 19 April 2012: ANN, EHL, ERA, IVC, MDL, MML, MSB, NAB, NUF, NWH, ORG, ORI, PAN, PBG, PTM, RRL, SHL, STO, SYD, TPI, and TSE;
(b) 17 May 2012: CTX, DTE, ERA, HDF, IDL, MQA, OGC, SAR, SDL, SFR, SLR, SVW, SXY and TOL;
(c) 19 July 2012: AGO, BTU and MDL;
(d) 16 August 2012: ILU, JBH, MGR, NAB, PAN, PDN, RHC, RMB, QBE and S-Z Securities; and
(e) 18 October 2012: BBG and the Excepted S-Z October Reductions (Order Index 77 to 83, placed prior to the commencement of the Final Pre-OSPA Period for S-Z Securities).
  1. In its written submissions, ASIC conveniently summarised Mr Graves’ reasoning, on which his opinion was based, as follows:
(a) The difference between the executable volume at the next price step above or below the IMP and the executable volume at the IMP is generally less than 10% of the IMV (10% IMV Rule).
(b) Therefore, the placement, cancellation or amendment of an order for XJO Securities that results in either adding or removing the equivalent of 10% or more of the IMV for the security is likely to cause an immediate change in the IMP (10% Order Rule).
(c) A combination of placements, cancellations or amendments are likely to have an immediate effect on the IMP if they occur at the same time and cumulatively result in the addition or removal of 10% or more of the IMV (10% Simultaneous Combined Order Rule).
(d) On a Serial Expiry Day, a combination of placements, cancellations or amendments to reduce volume that occur within the two minutes immediately before the earliest time for the OSPA (Final Pre-OSPA Period) may be treated as if they occurred at the same time because of the relative stability of the IMV in that period on Serial Expiry Days, and therefore a combination of that kind that results in the addition or removal of 10% or more of the IMV is likely (once the 10% threshold is reached) to have an immediate effect on the IMP (10% Sequential Combined Order Rule).
(e) A change in IMP will cause people to change their orders. Those who use computer-directed order management can respond to changes in the IMP within fractions of a second, whereas orders that are placed manually may take significantly longer.
(f) By the start of the Final Pre-OSPA Period (that is, the last two minutes before the earliest possible time for the rotation of the OSPA) some manual order managers are already unable to respond to any changes and, as the time gets closer to the earliest start time for the Rotations of the OSPA, the number of manual order managers who can respond decreases rapidly;
(g) Therefore, not all users in the market are likely to place, cancel or amend orders in response to a Significant Order Event (any of the events described in either (b) to (d) above) that occurs within the Final Pre-OSPA Period.
(h) For those investors who are able to change their orders in response to a change in the IMP of a security that occurs within the Final Pre-OSPA Period, they will only do so gradually, if given sufficient time, as follows:
(i) Typically, there is an initial wave of buyers/sellers whose orders have been, or are in danger of being, displaced from the indicative match as a result of the increase/decrease of the IMP and who therefore amend (subject to order instructions) the price of their orders to put them back, or deeper, in the match to avoid the “opportunity cost” of their orders not trading in the OSPA. This puts further pressure on the IMP in the same direction as the initial change in the IMP and may push it further than it was as a result of that initial change.
(ii) After this, there is typically a wave of sellers/buyers who place further sell/buy orders in the indicative match, or by amendment increase the volume of their orders in the indicative match, seeking to take advantage of a higher IMP (which is favourable to sellers) or a lower IMP (which is favourable to buyers). This puts downward/upward pressure on the IMP and may push it back towards what it was before the initial IMP change. However, those buyers and sellers seeking to take advantage of the price change will necessarily seek to avoid or minimise dissipating it so they will not initially place buy orders for their highest acceptable price but, rather, will only gradually place or amend their orders towards those prices.
(i) If a Significant Order Event occurs within the Final Pre-OSPA Period, it is therefore likely to have an ongoing effect on the IMP up to the OSPA and, accordingly, on the Opening Price itself (2 Minute Rule).
  1. The definitions used in this summary are ASIC’s. They do not replicate the definitions used in Mr Graves’ report. Nevertheless, they capture the concepts and notions addressed in Mr Graves’ report without altering the substance of his reasoning or opinions.

THE CHALLENGE TO MR GRAVES’ EVIDENCE

The 10% Order Rule

  1. There was no real dispute about the validity of the 10% IMV Rule, taken as a rule of thumb. There was, however, a dispute about the 10% Order Rule which, on Mr Graves’ reasoning, followed from the 10% IMV Rule.
  2. The defendants submitted that the effect of the placement, cancellation or amendment of an order for an XJO Security that results in either adding or removing the equivalent of 10% or more of the IMV at the security’s existing IMP will depend upon what else is in the Order Book at the time. Mr Graves agreed with that proposition, but did not agree that this meant that the statement of the 10% Order Rule should be qualified. According to Mr Graves, this is because he was looking at (what he described as) the immediate effect of the placement of the order. As he put it in cross-examination:
... So all of the other orders that in the order book are there. I’m looking at the next step, which is the placement of that order. So it’s my opinion that the logic holds.
  1. However, the defendants’ point was that, whether or not the addition or removal of 10% of the IMV will cause a change in the IMP depends on the operation of the four-step process of the ASX Algorithm, and does not simply or necessarily flow from the fact that 10% or more of the IMV has been added or removed. The burden of this submission appears to be that the 10% Order Rule is an oversimplification of the ASX Algorithm and does not adequately account for how the algorithm works.
  2. Professor Aitken was asked whether there was any empirical support for Mr Graves’ 10% Order Rule. He conducted tests using ASX order and trading data relating to 50 days in 2012 and 2013, 20 of which he defined as “20 Serial Expiry Days” (which erroneously included four Quarterly Serial Expiry Days) and 30 of which he defined as “30 Non-Serial Expiry Days”. His analysis showed that the percentage of “10% orders” that changed the IMP was approximately 46% on the “20 Serial Expiry Days” and 34% on “Non-Serial Expiry Days”—in other words, much less than a chance.
  3. ASIC criticised the results of this analysis because Professor Aitken treated all amendments that increased the volume of an order, or that amended price alone, as a cancellation and a subsequent new order for the amended price or volume.
  4. Professor Aitken’s justification for treating the data in this way reflected his understanding of how ASX’s order priority rules operate (which govern the order in which queued orders will trade):
... [It] is my understanding of the operation of the ASX Market (now “Operating”) Rules and which were in place in 2012, [sic] that an order amendment that changes price or price and volume, as opposed to just volume, becomes a new order whose order priority changes. In effect, such an amendment results in the original order being cancelled, and a new order being placed. Further, if an order is not amended as to price but is amended as to volume, and volume goes up, the same applies; the amendment gives rise to a cancellation and a new order placement. Under ASX Operating Rules, the only time an amend order does not give rise to a new order is when price does not change and order volume is reduced. ...
  1. ASIC submitted that the effect of Professor Aitken treating the data this way meant that if an order was in the indicative match, had a volume of more than 10% of the IMV and was amended as to price alone (which price was also in the indicative match), or increased its volume by as little as one share, it was treated as a cancellation followed by a placement, even though the amendment did not substantively change the volume in the indicative match.
  2. ASIC submitted that this was a flawed approach because order priority has no bearing on the 10% Order Rule, which is solely concerned with the addition or removal of volume equivalent to at least 10% of the IMV, and whether that addition or removal is likely to have an immediate effect on the IMP.
  3. Moreover, ASIC submitted that Professor Aitken’s analysis presumably counted as significant cancellations and significant placements amendments that made little or no change to volume in the indicative match. The burden of this submission, as I apprehend it, is that, by treating the data in this way, Professor Aitken obtained deflated percentages which did not truly test, empirically, the 10% Order Rule.
  4. ASIC endeavoured to draw support for the 10% Order Rule from a table prepared by one of Professor Aitken’s assistants, Dr Ji: Exhibit P56. This table appears to be part of Professor Aitken’s working papers. ASIC attempted to elicit an explanation of the table from Professor Aitken in cross-examination. The explanation was, with respect, utterly confusing.
  5. The cross-examination was directed to two rows in the table. The point sought to be made by ASIC was that the second row in the table most appropriately covered the amendments addressed by Mr Graves in the 10% Order Rule. That row recorded that 65% of the amendments made on Serial Expiry Days moved the IMP.
  6. Professor Aitken dismissed any reliance on the table, ultimately because the percentage figures shown in it did not reflect his view of how the word “likely” should be construed (which he equated with an outcome that has a 66% or greater likelihood or probability). I do not think that, by dismissing reliance on the table in this way, Professor Aitken was necessarily agreeing with ASIC’s proposition that it was recording data addressing Mr Morgan’s 10% Order Rule. That said, it appears that the first row might have done so. If that be so, it records that 39% of orders on “Non-Serial Expiry Days” and 50% of orders on “Serial Expiry Days” moved the IMP. These results do not provide empirical support for the 10% Order Rule. The second row records that 39% of orders on “Non Serial Expiry Days” and 65% of orders on “Serial Expiry Days” moved the IMP. The trouble is that I do not know what the second row of the table is addressing, either from Professor Aitken’s cross-examination or from the text of the table itself, which is cryptic.
  7. In closing submissions, ASIC relied on the first and second rows in the table, and the combined percentage figure of 61% recorded in the third last line of the document. ASIC submitted that this showed the percentage of Significant Reductions defined by Mr Graves (orders at a price better than the IMP which are reduced by at least 10% of the IMV) that changed the IMP on the “20 Serial Expiry Days”. Apart from the obscurity of the second row of the table, it is not clear to me what the 61% figure is addressing or how it is calculated from the body of the table.
  8. In the end result I do not place any reliance on the table. It is confusing and not properly explained. It does not enable me to conclude that there is empirical support for the 10% Order Rule.

The 10% Simultaneous Combined Order Rule and the 10% Sequential Combined Order Rule

  1. As the ASX Algorithm is a mathematical one, Mr Graves reasoned that the effect of many accumulated small orders can be the same as one large order.
  2. The 10% Simultaneous Combined Order Rule is based on Mr Graves’ notion that two or more orders placed at the same price (being at least one price step better than the IMP) that do not individually have an immediate effect on the IMP are likely to have an immediate effect as soon as their cumulative addition to volume is sufficiently large to change the price at which the MEV occurs. Similarly, amendments or cancellations that remove volume, but do not have an immediate effect on the IMP, are likely to have an immediate effect as soon as their cumulative removal of volume is sufficiently large to change the price at which the MEV occurs.
  3. Mr Graves reasoned that if a combination of placements, amendments or cancellations of the kind referred to above were all made at the same time, then the addition of volume caused by those placements, or the removal of volume caused by those amendments or cancellations, is likely to be sufficiently large to change the IMP when the sum of the individual additions to or removal from volume, expressed as a percentage, is at least 10% of the IMV.
  4. However, Mr Graves also said that if the placements, amendments or cancellations are separated in time then, generally speaking, the more time that elapses between the individual events in the combination, and between the first and last events in the combination, the more likely it is that other events will occur in the market and that the IMV will change in the intervening period. In those circumstances, Mr Graves said that the sum of the individual additions to or removal from volume would have to be a percentage other than 10% to have the same effect as one addition to or removal of volume equivalent to 10% of the IMV at that time. In cross-examination, Mr Graves accepted that it was crucial to his reasoning that placements, amendments and cancellations can be combined, that those placements, amendments and cancellations take place at the same time. This was because other, intervening orders can affect the IMV and the IMP.
  5. Relatedly, Mr Graves considered that every order placed in the last two minutes prior to the OSPA for each Rotation—the Final Pre-OSPA Period referred to above—should be taken as having been placed at the same time for the purpose of applying the 10% Simultaneous Combined Order Rule to Whitebox’s trading on the Serial Expiry Days in question. This is what ASIC defined as the 10% Sequential Combined Order Rule. It is also an aspect of the 2 Minute Rule. Whitebox submitted that the proposition that every order that is placed in the last two minutes prior to each rotation in the OSPA can be combined, is manifestly absurd.
  6. Professor Aitken said that two or more orders placed over a two minute period could not be considered to be orders placed at the same time. He examined the data for the “20 Serial Expiry Days” and noted that, in Mr Graves’ Final Pre-OSPA Period, the average and maximum number of orders placed in that period were 180 (90 per minute) and 2755 (1,377 per minute). He said:
... This is consistent with my experience that many hundreds of Orders can be placed for any individual XJO Security by participants on the ASX in a 2-minute period. There would be a multitude of orders from different participants in that time period (a.k.a. Interposed Orders) and an almost infinite number of different influences that may exist affecting why Orders are placed, not least of which is that parties react to [others] Orders. Indeed, even in a period of one second, there could be multiple Orders placed by different participants and also from the same participants, including Orders reacting to Orders placed within that same second. The ASX is a market that has participants reacting and placing Orders within a second. In such a market, to talk about Orders that are placed at the “same time”, one would have to be talking about Orders placed within milliseconds, not seconds of each other, and certainly not within minutes of each other. I would say therefore, in the context of Orders being placed on ASX as being at the “same time”, one would need to be looking at a time frame of some milliseconds and those Orders not interspersed with Orders from other participants.
  1. Mr Graves accepted in cross-examination that in the Final Pre-OSPA Period on the Serial Expiry Days in question there were many thousands of orders that had been interposed between the Whitebox orders in question. He also accepted that, in the two minute period of his Final Pre-OSPA Period, traders can and do respond to changes in the IMP. Critically, he agreed that it was not correct to treat this two minute period as being “at the same time”. He also accepted that two minutes was “a long time in the order book” in which “thousands of orders can go through a security in that time”. He agreed that the two minute period was an arbitrary choice. It was put to him that he did not have “any reasoned basis for distinguishing between two minutes and 10 seconds, on the one hand, and one minute 50 seconds, on the other”, and he accepted that, indeed, there was no empirical basis to support the selection of the two minute period. The support he proffered was his experience in observing the market based on the practices of manual order managers.
  2. In his calculations Mr Graves took a series of Whitebox orders that had been placed in the Final Pre-OSPA Period, none of which would have individually met the 10% Order Rule, and combined them so that, if taken together, they did meet that rule. He thus treated these orders as if they had been placed at the same time. But, as Professor Aitken observed, the IMV is necessarily affected by the large number of orders that were interposed with the Whitebox orders, as identified in Professor Aitken’s and Associate Professor Prokhorov’s Joint Report. Professor Aitken said that the effects of the interposed orders on the IMV have not been, and cannot be, removed from Mr Graves’ calculations and that, for this reason, these calculations cannot be relied on.
  3. ASIC submitted that Professor Aitken’s position was “patently artificial”. It argued that Professor Aitken’s view was founded on an opinion that any effect on the second order in the proposed sequence on the market would, by definition, be caused by the interposed order and not any earlier order. ASIC illustrated its criticism of Professor Aitken’s approach as follows:
It is akin to saying that, if a builder builds a wall seven bricks high and another builder later adds or removes a layer of bricks, then the height of the wall no longer had anything to do with the first builder’s work.
  1. I do not think that this analogy answers the point that Professor Aitken’s criticism addresses. Mr Graves’ opinion was that, other than for certain securities (those noted at [388] above), the placement of the Cancellations (19 April 2012) and the Reductions (other Serial Expiry Days) was likely to have an impact on the Opening Price of XJO Securities in the OSPA on the days in question. It is to be borne in mind that, in the context of the fourth pillar, ASIC alleges that the placement of the Cancellations and Reductions was deliberately undertaken to affect the price at which relevant XJO Securities would trade in the OSPA. But the price at which relevant XJO Securities would trade in the OSPA is a function of the state of the Order Book at the commencement of each Rotation in the OSPA and, on the Serial Expiry Days in question, the state of the Order Book was not simply a function of Whitebox’s placements, cancellations and amendments from time to time, but a function of all the orders in the book. The effects of others’ trading cannot be ignored because that trading will also affect the IMV and, hence, the IMP. It is also to be borne in mind that the defendants had no control over, or foreknowledge of, the placements, cancellations and amendments that other traders might make with a view to profiting from their own trading.
  2. The criticism of Mr Graves’ approach is that the 10% Sequential Combined Order Rule in particular purports to say something about Whitebox’s conduct, whereas that rule does not and cannot say anything specific about Whitebox’s conduct at all.
  3. I accept that the 10% Sequential Combined Order Rule does not say anything specific about Whitebox’s conduct because, in his calculations seeking to apply that rule to Whitebox’s trading, Mr Graves did not account for, mathematically, a very large number of interposed orders placed by others, which also would have affected the IMV and, hence, the IMP.
  4. I also accept that, at a conceptual level, the trading of others—the interposed orders—cannot be ignored so far as their effect on the IMV is concerned. Moreover, one cannot know how the trading by others was affected by the placement of Whitebox’s orders. At the empirical level, Mr Graves did not show the effect that the trading of others had on the IMV of the securities in question on the days in question. Without knowing this, one simply cannot know what effect the Cancellations (19 April 2012) and Reductions (other Serial Expiry Days) had on the IMV and, hence, the IMP.
  5. As Professor Aitken put it:
5.74 It is extremely difficult to say with any degree of conviction what the opening match price of XJO Securities would have been absent the [Whitebox orders] or if those Orders were placed earlier in time, because one cannot simply remove them from the market and run the marketplace. This is because the market consists of a large array of traders who will be reacting to orders placed in the market. Reaction and activity may be different if orders are not placed. One cannot therefore just remove or move one participant’s orders and expect all other activity to have remained the same. To do that will give artificial results. Exactly how other participants would have acted if another participant’s orders were not placed is impossible to say with any degree of certainty.
  1. Mr Graves’ rationale for the 10% Sequential Combined Order Rule was that the IMV of XJO Securities in the Final Pre-OSPA Period on Serial Expiry Days is relatively stable. Professor Aitken agreed that there is less volatility in percentage changes in the IMV in Serial Expiry Days compared to non-Serial Expiry Days. However, he did not see this as supporting the 10% Sequential Order Rule because it did not follow that there was less order activity on Serial Expiry Days. He therefore disagreed with the suggestion that the cumulative effect from a series of orders in the Final Pre-OSPA Period on a Serial Expiry Day will “act like one simultaneous order to affect the IMP” (Professor Aitken’s words).
  2. I prefer and accept Professor Aitken’s evidence. I would add that it is hard to see how relative stability provides substantial support for the 10% Sequential Combined Order Rule once it is accepted that there is movement in the IMP in the Final Pre-OSPA Period. I accept Whitebox’s submission that the concept of relative stability of the IMV ceases to be a viable measure of whether two orders can be treated as placed, amended or cancelled at the same time.
  3. As to the validity of the 10% Simultaneous Combined Order Rule itself, Professor Aitken carried out an analysis that led him to conclude that there was no empirical support for it. His analysis proceeded on the basis that orders could be combined only if they were placed by the same participant; were of the same order type (placement, amendment or cancellation); were placed at the same time (within 2 milliseconds, although Professor Aitken ran the same analysis for 5 and 10 milliseconds with no substantive change in results); and were from the same side of the Order Book (i.e., were either of a Bid or an Ask).
  4. Although there was some disagreement between Professor Aitken and Associate Professor Prokhorov on the matter, the outcome of this particular analysis was that, in respect of the “20 Serial Expiry Days”, 38% of the “combined orders” caused the IMP to move (Associate Professor Prokhorov, 35%). In respect of the “30 Non-Serial Expiry Days”, 27% of the “combined orders” caused the IMP to move (Associate Professor Prokhorov, 30%).
  5. ASIC criticised Professor Aitken’s analysis because, as with his analysis of the 10% Order Rule, Professor Aitken treated all amendments that increased the volume of an order, or that amended price alone, as a cancellation and a subsequent new order for the amended price or volume. For this reason, ASIC submitted that Professor Aitken’s test was not a legitimate test of the rule. I have already rejected ASIC’s criticism of Professor Aitken’s evidence on this score

The 2 Minute Rule

  1. The immediately preceding discussion concerns the significance of the two minute period of Mr Graves’ Final Pre-OSPA Period for the purpose of combining orders. This two minute period had further significance for Mr Graves’ analysis.
  2. In his report, Mr Graves opined that it is likely that the placement, amendment to reduce volume or cancellation of an order for an XJO Security in the Pre-Open Phase on a Serial Expiry Day will have a continuing effect on the IMP of the security throughout the remaining period until the OSPA Rotation in which the security is due to trade if it is a Significant Order Event and it is placed after the start of the Final Pre-OSPA Period (i.e., within two minutes of the earliest possible time of the rotation of the OSPA in which that security is due to trade).
  3. As Whitebox correctly submitted, Mr Graves’ formulation of the 2 Minute Rule depended heavily on his notions of the personal response time of manual order managers, as described in the following paragraphs of Mr Graves’ report:
    1. ... I have described actions (namely placing, amending or cancelling orders) that certain Order Managers and other users will take in response to a change in the indicative match price or the displacement of their orders from the indicative match, provided they have sufficient time both to observe the change or displacement and take that responsive action.
    2. As described further below, many Order Managers manage their orders “manually”, that is, they personally monitor the market and enter orders by human action (Manual Order Managers). Manual Order Managers who are managing orders on behalf of other investors rather than themselves (i.e. certain brokers and investment advisors) enter orders either in response to specific instructions regarding the order or within the parameters of standing instructions from the investor. Other market users employ computer-directed order management. In contract to manual order management, computer-directed order management involves using automated processes for monitoring and responding to movements in the market and placing orders, amendment and cancellations in the market through the use of algorithms and automated processes that are set according to pre-determined instructions.
    3. Computers can respond within fractions of a second to events in the market. Therefore, users who employ computer-directed order management can respond almost instantly in the kinds of ways described in paragraphs 47 and 50 above to a change in the indicative match price of a security or the displacement of their orders from the indicative match.
    4. Manual Order Managers need substantially longer than computers to respond to events in the market. The time needed by particular Manual Order Managers to take any such responsive action (Personal Response Time) depends on several things, including the following (in no particular order):
a. the market monitoring tools to which they have access;
  1. any need to obtain new instructions or clarify existing instructions in order to respond;
  1. the procedures and processes they are required to follow in managing orders;
  1. the number of clients for whom they are managing orders and monitoring securities;
  2. the numbers of orders they are managing and securities they are monitoring;
f. the experience and skill of the Order Manager.

(Emphasis in original)

  1. However, in cross-examination, Mr Graves accepted the following propositions:
(a) Of all persons placing orders in the order book, very few of them were subject to the constraints that Mr Graves identified in [56] of his report (quoted above).
(b) Retail investors are a small fraction of the volume in the ASX Order Book.
(c) By 2012, many of those within the universe of retail investors traded directly through online brokers, such as CommSec.
(d) If a retail investor was using an online broker like CommSec, that investor was able to make his or her own decisions and was only constrained by the time it took to observe the market and then to place an order via a keyboard on a computer.
(e) In respect of retail clients who were using a traditional broker, manual order managers often did not need to take instructions in order to respond to developments in the market because their mandate permitted them to respond without doing so.
(f) Even within the universe of retail investors, manual order managers who had to make telephone calls, before they could respond to developments in the order book, were in 2012 a small proportion of the total of retail investors.
  1. Professor Aitken said that he was not aware of how one could accurately count how many manual order managers placed orders in Mr Graves’ Final Pre-OSPA Period so as to gauge their importance in maintaining an equilibrium of demand and supply. But, he said, given automation of the marketplace and his experience of “how the market operates”, the traders meeting Mr Graves’ description of manual order managers “would account for a very small proportion of turnover in the market and their activity would have little or no influence on price”.
  2. The 2 Minute Rule proceeds on the theory that any change in the IMP in the Final Pre-OSPA Period will not represent a genuine balance of supply and demand because, until all manual order managers respond, the IMP is likely to be different to the price it would be if all such managers had, in fact, responded.
  3. I should say at once that there is no evidence to support such a surprising proposition. Mr Graves accepted that it could make no difference to the IMP that some manual order managers might be slow to respond in circumstances where another, faster participant in the book moved the IMP back to its previous level. He accepted that such a change by another participant who was not a manual order manager can happen in “a very, very short period of time” and that, in respect of “large institutional players with computers driven by algorithms”, this can happen in the space of milliseconds.
  4. In addition, an order placed by a manual order manager would only make a difference to the IMP if it was of sufficient volume under the ASX Algorithm to move that price. Mr Graves accepted that he did not have any basis for saying that it was likely, as a general proposition, that manual order managers would have been placing orders in excess of 10% of the IMV.
  5. Finally on this score, Mr Graves accepted that, in continuous trading, the fact that a manual order manager did not have time to respond to a market event did not mean that the prices set did not represent a genuine balance of supply and demand. As Whitebox correctly submitted, there is no logical basis to distinguish between the market realities of price formation in continuous trading and those forces acting to form the IMP in the Pre-Open Phase.

Other analyses carried out by Professor Aitken

  1. It is convenient at this juncture to record the results of other analyses carried out by Professor Aitken. These analyses relate to what Mr Graves referred to as “Significant Placements”, “Significant Reductions” and “Significant Cancellations”. Professor Aitken referred to these as “Significant Orders”.
  2. In his report, Mr Graves included these as “Significant Order Events”. He also included a “Significant Placement Combination”, a “Significant Reduction Combination” and a “Significant Cancellation Combination” as “Significant Order Events”.
  3. Professor Aitken’s other analyses did not address, therefore, all the order “events” captured by Mr Graves’ “Significant Order Events”. It will be apparent that Mr Graves’ opinion regarding the likely impact of the placement of the Cancellations (19 April 2012) and the Reductions (other Serial Expiry Days) on the Opening Price of XJO Securities in the OSPA was based on combining Whitebox’s orders in the two minutes captured by his Final Pre-OSPA Period, in the way I have described. Professor Aitken’s additional analyses (which I accept) are, nevertheless, instructive.
  4. By way of background to this section of the reasons, Mr Graves expressed the opinion that each of the following is likely to have an immediate effect on the IMP of a security to which it relates:
(a) a Significant Placement (a placement that is at a price better than the IMP for the security immediately before the placement and for a volume that is equivalent to at least 10% of the IMV of the security immediately before the placement);
(b) a Significant Reduction (an amendment to the volume of an order that immediately before the amendment is at a price that is better than the IMP for the security at that time and reduces the volume of the order by an amount that is equivalent to at least 10% of the IMV of the security immediately before the amendment); and
(c) a Significant Cancellation (a cancellation of an order that is at a price better than the IMP for the security immediately before the cancellation and is for a volume that is equivalent to at least 10% of the IMV of the security immediately before its cancellation).
  1. As I have said, Professor Aitken referred to these as “Significant Orders”.
  2. Professor Aitken’s analyses showed that 47% of “Significant Orders” placed during the Final Pre-OSPA Period for XJO Securities across the “20 Serial Expiry Days”, and 45% of such orders across the “30 Non-Serial Expiry Days”, moved the IMP. He also analysed the Serial Expiry Days in question and found that 49% of the “Significant Orders” placed in the Final Pre-OSPA Period moved the IMP. He expressed this conclusion:
5.51 This evidence showing Significant Orders placed during the Final Pre-OSPA Period for XJO Securities will have an immediate impact on the IMP across the sample is consistently below 50% for the 20 Serial Expiry Days, the 5 Serial Expiry Days and the 30 Non-Serial Expiry Days. That is inconsistent with there being empirical support for the general proposition that Mr Graves has stated. The results show that occurrence has a probability less likely than chance. An alternative way of interpreting these results is that if I tossed a coin to decide whether Mr Graves’ proposition was true or not, that the coin toss would be no less accurate. Accordingly, the correlation between a Significant Order placed during the Final Pre-OSPA Period and being an immediate change on the IMP is no more than chance and nowhere near likely, defined to be greater than 66%.
(Emphasis in original)
  1. Based on the results of these analyses, Professor Aitken said that there is no empirical support for the proposition that a “Significant Order” placed in the Final Pre-OSPA Period will likely have a continuing effect on the IMP at the OSPA to affect the Opening Price.
  2. In addition he took from his analyses the “Significant Orders” that did move the IMP in the Final Pre-OSPA Period and examined whether it could be said that these orders were likely to have caused the change in the IMP to carry through to the Opening Prices in the OSPA. In this connection, he looked at how many of these “Significant Orders” were followed by non-“Significant Orders” which further moved the IMP before the OSPA. He concluded that, for the “20 Serial Expiry Days”, 96.6% of the “Significant Orders” were followed by non-“Significant Orders” that further moved the IMP, and for the “30 Non-Serial Expiry Days” 96.8% of the “Significant Orders” were followed by non-“Significant Orders” that further moved the IMP.
  3. Professor Aitken then expressed the following overall conclusions:
5.56 Therefore, with the exception of 3.4% of Significant Orders that moved the IMP on Serial Expiry Days, there is no empirical evidence to suggest that Significant Orders placed in the Final Pre-OSPA that did move the IMP, did cause a change in the IMP that was carried through to the OSPA.
5.57 For Non-Serial Expiry Days, the position is similar. With the exception of 3.2% of the Significant Orders that moved the IMP on Non-Serial Expiry Days, there is no empirical evidence to suggest that Significant Orders placed in the Final Pre-OSPA that did move the IMP, did cause a change in the IMP that was be carried through to the OSPA.
5.58 Collectively, my work suggests that there is no empirical support for the conclusion that Significant Orders placed for XJO Securities in the Final Pre-OSPA Period were likely to lead to a change in the IMP for XJO Securities in the OSPA to carry through to the opening match price for XJO Securities in the OSPA. I have only identified approximately 3% of Significant Orders where that might be said to have occurred.
5.59 Accordingly, there is no empirical support for the proposition from paragraphs 100 to 102 of Mr Graves’ Report that Significant Orders placed in the Final Pre-OSPA Period on a Serial Expiry day were likely to have a continuing effect on the IMP of XJO Securities until the actual OSPA to carry into the opening match price for XJO Securities in the OSPA.
(Emphasis in original)
  1. Professor Aitken also analysed the orders placed for XJO Securities between 9:00 and the OSPA on the “20 Serial Expiry Days” and found that:
(a) slightly below 50% of the orders were amended or deleted; and
(b) 52% of the amendments that reduced volume were placed during the last five minutes before the OSPA.
  1. On the Serial Expiry Days in question, Professor Aitken determined that other traders placed more “Significant Orders” in the Final Pre-OSPA Period than Whitebox, and that less than 50% of the “Significant Orders” placed by other traders, and by Whitebox, moved the IMP. Professor Aitken said that this suggested that the proportion of “Significant Orders” placed by Whitebox during the Final Pre-OSPA Period on the Serial Expiry Days in question were not different from those placed by other traders.
  2. Professor Aitken also looked at the movement of the IMP leading up to the OSPA on the five Serial Expiry Days in question. He reasoned that if Whitebox’s orders were to have an impact on price, they should be the orders driving the price in the minute or so leading up to the OSPA. Alternatively, if other orders were impacting the prices in this period, it would make it much more difficult to conclude with confidence that Whitebox’s orders were affecting the OSPA.
  3. To this end, Professor Aitken considered the top 30 stocks that constitute 70% of the XJO. He did so because of the time-consuming nature of the task of dealing with all stocks in the index. His analysis showed that there were only 12 out of 150 cases where Whitebox’s orders were the last to have impacted the OSPA—in other words, in most cases, other parties were affecting the OSPA towards the close. He looked at the 12 Whitebox orders and could discern no pattern that, in his view, raised any suspicion that the orders had been placed with the purpose of significantly affecting the OSPA for the stocks in question, or that the orders had that effect.
  4. Professor Aitken said:
5.88 On the basis of the above evidence I have some difficulty with Mr Graves’ assertion that it was [Whitebox’s orders], as opposed to any other participants’ orders, that impacted or were likely to have impacted on the opening match price for XJO Securities on the 5 Serial Expiry Days. There is no convincing empirical evidence that I could find to suggest that was the case. Also, for reasons I have noted above (namely, that there is frequently a range of other orders in between Significant Orders, many of which affect IMP), it is difficult to conclude one way or the other whether Significant Orders placed by [Whitebox] that moved the IMP have had or were likely to have had a continuing impact on the IMP until the actual OSPA to carry through to the opening match price. Finally, the charts for the top 30 stocks over the 5 relevant serial expiry days suggest that [Whitebox] orders are unlikely to have significantly impacted either the stocks in question or the index.
5.89 In conclusion therefore, the empirical evidence does not support the general proposition proposed by Mr Graves, that Significant Orders placed during the Final Pre-OSPA Period are likely to have a continuing effect on the IMP of XJO Securities until the actual OSPA. There is also no empirical support for the more specific conclusion that any of the [Whitebox orders] that were Significant Orders placed in the Final Pre-OSPA Period had or were likely to have a continuing effect on the IMP of XJO Securities until the actual OSPA. The key reason for this conclusion is that Mr Graves does not account for interposed orders, either generally or in the more specific case of [the Whitebox] orders.
(Emphasis in original)
  1. I accept the evidence presented by Professor Aitken’s other analyses.

CONCLUSIONS ON MR GRAVES’ EVIDENCE

  1. For the purpose of considering the likely effects on the IMP and the Opening Prices of the Cancellations and Reductions, Mr Graves was instructed, firstly, to consider the likelihood or otherwise of those effects looking forward from the time the Cancellations and Reductions were placed. Secondly, he was instructed to compare the position that was likely to prevail with the placement of the Cancellations or Reductions occurring when they did, with the position likely to prevail if the Cancellations or Reductions had been placed at certain earlier times.
  2. ASIC submitted that these tasks do not lend themselves to conducting empirical analyses or expressing outcomes of likelihood in percentage figures which, it argued, “would have been artificial in a forward-looking task that depended on any number of variables given the dynamic nature of the ASX market in the pre-open phase”.
  3. These submissions were made in justification of Mr Graves’ use of rules of thumb. It is not, however, necessary for me to decide whether Mr Graves’ approach was justified. My concern must be whether the rules of thumb he did use represent, individually and collectively, reliable simulations of how the market for XJO Securities operated by ASX actually works and, if so, whether Mr Graves’ analyses and opinions are valid and support ASIC’s fourth pillar in the context of a case that raises serious allegations of commercial impropriety attracting significant sanctions.
  4. In my view, Mr Graves’ rules of thumb are not reliable and no reliance should be placed on them for the purpose of determining the likely price effects of the Cancellations (19 April 2012) or the Reductions (other Serial Expiry Days); still less should reliance be placed on them to infer that the defendants’ purpose was to achieve the price effects that ASIC alleges.
  5. In the case of the 10% Order Rule, I accept that its reliability depends on the actual state of the Order Book at a given point in time. The fact that the ASX Algorithm operates by a four-step process shows that merely adding or removing the equivalent of 10% or more of the IMV for a given security is a somewhat simplistic, and potentially misleading, approach to determining whether there is likely to be an immediate change in the IMP by reason of that change in volume. Professor Aitken’s evidence demonstrated the likely reliability of this rule, which was no better than chance. I do not accept ASIC’s criticism of Professor Aitken’s analysis in this regard, which took into account ASX’s order priority rules. In my view, it was appropriate that Professor Aitken did take those rules into account. After all, the reliability of the 10% Order Rule should be considered from the perspective of how the market operated by ASX (ASX Trade) actually works.
  6. The 10% Order Rule underpins the 10% Simultaneous Combined Order Rule and the 10% Sequential Combined Order Rule. Thus, the reliability of the 10% Simultaneous Combined Order Rule and the 10% Sequential Combined Order Rule is qualified accordingly. Further, the 10% Simultaneous Combined Order Rule and the 10% Sequential Combined Order Rule depend on the appropriateness of combining separate order events. As I have noted, Mr Graves accepted that it was crucial to his reasoning that placements, amendments and cancellations can be combined, that these order events occur at the same time.
  7. I do not accept that it is appropriate to treat every order placed by a particular trader in the last two minutes prior to each relevant Rotation of the OSPA as having been placed by that trader at the same time, simply because the placement, amendment or cancellation was made in that period. As I have noted, the evidence shows that, in that period, many thousands of orders can be placed, including orders that are interposed between the order events that Mr Graves’ rules combine. I accept the criticisms of the 10% Simultaneous Combined Order Rule and the 10% Sequential Combined Order Rule, which I have summarised at [400] – [418] above, including Mr Graves’ own evidence in cross-examination, which substantially undermined any reliance that could be placed on these rules.
  8. With specific reference to the 10% Simultaneous Combined Order Rule, Professor Aitken provided an example of what, in my view, could be considered to be orders placed at the same time and demonstrated, once again, the likely reliability of that rule (considered on that basis), which was less than a chance. Once again, I do not accept ASIC’s criticism of that analysis, which also took into account ASX’s order priority rules.
  9. As to the 2 Minute Rule, I do not accept that one can assume that it is likely that the placement, amendment to reduce volume or cancellation of an order for an XJO Security in the Pre-Open Phase on a Serial Expiry Day will have a continuing effect on the IMP of the security throughout the remaining period until the OSPA Rotation in which the security is due to trade if it is a Significant Order Event (as defined by Mr Graves) and it is placed after the start of the Final Pre-OSPA Period. As I have noted, this rule is based on the surprising proposition, for which there is no evidence, that any change in the IMP in the Final Pre-OSPA Period will not represent a genuine balance of supply and demand until all manual order managers can respond. Once again, the propositions accepted in cross-examination by Mr Graves, summarised at [422] – [427] above, substantially undermined the validity of this rule and the reliance that could be placed on it.
  10. My rejection of the reliability of the 10% Order Rule, the 10% Simultaneous Combined Order Rule, the 10% Sequential Combined Order Rule and the 2 Minute Rule deprive ASIC’s fourth pillar of the evidentiary support on which it purports to rest. Accordingly, by reference to Mr Graves’ evidence, ASIC has not established that, at the time they were placed, the Cancellations and Reductions were likely to affect the price at which the relevant XJO Securities would trade in the OSPA.

ADDITIONAL EVIDENCE RELIED ON TO SUPPORT PRICE EFFECTS

The auction time lines and Mr Clifford’s tables

  1. ASIC argued that Mr Graves’ evidence concerning the likely effects of the Cancellations (19 April 2012) and Reductions (other Serial Expiry Days) was complemented by the conclusions that could be drawn from the data provided by the auction time lines and from tables prepared by Mr Clifford. As to the latter, reference must be made to the rulings I made in Australian Securities and Investments Commission, in the matter of  Whitebox Trading  Pty Ltd v  Whitebox Trading  Pty Ltd (No 6) [2018] FCA 1077 (Whitebox No 6) as to the confinement of the use which could be made of Mr Clifford’s tables.
  2. ASIC submitted that compelling evidence of the immediate and ongoing effects of the Cancellations (19 April 2012) and Reductions (other Serial Expiry Days) is contained in the auction time lines. ASIC sought to illustrate this proposition by reference to movements in Underlying Basis, Fair Basis and Auction Basis summarised in tables that extracted data from the auction time line for 18 October 2012: Exhibits 50 and 51. ASIC pointed to the relative differences in movement between Fast Basis and Underlying Basis from the start to the end of the Reductions, and the commencement of each rotation in the OSPA. ASIC argued that this relative movement was the consequence of the placement, amendment or cancellation of Whitebox’s orders, based on Mr de Kantzow’s acceptance that comparing the two movements will give “some idea” of the impact that Whitebox’s orders had on Basis. It is to be noted that any movement in Basis is because there is a movement in price of the underlying securities.
  3. ASIC argued that the slower progression of Fast Basis (compared to Underlying Basis) towards an unprofitable position, or the progression of Fast Basis (compared to Underlying Basis) towards a profitable or more profitable position, can be attributed to Whitebox’s orders. Therefore, on ASIC’s argument, and by reference to Exhibits 50 and 51, Whitebox’s October Reductions can be seen to have caused a movement in the price of the securities the subject of those Reductions.
  4. By comparing Auction Basis with Fast Basis at the time of each rotation in the OSPA on that day, ASIC submitted that Auction Basis became substantially more negative than Fast Basis, so that by the time of the Group 5 Rotation, Auction Basis was -76 whereas Fast Basis was only -1.9. ASIC submitted that this indicated that, although Fast Basis had decayed to “almost zero”, Whitebox had traded a substantially profitable position in the OSPA overall.
  5. As to Mr Clifford’s tables, ASIC commenced with the proposition that one would expect that any immediate price impact of the Cancellations (19 April 2012) would be a decrease in prices, on the principle that cancelling buy orders will reduce demand for the securities concerned. Correspondingly, one would expect that any immediate price impact of the Reductions (other Serial Expiry Days) would be to increase prices, on the principle that reducing volumes of sell orders will reduce supply for the securities concerned.
  6. Taking certain tables (referred to in the evidence as Tabs 20 – 24 (19 April 2012), 57 – 61 (17 May 2012), 92 – 96 (19 July 2012), 127 – 131 (16 August 2012) and 163 – 167 (18 October 2012), ASIC pointed to the percentage changes in the last column of each table—“Per-cent change OSPA index”. The data in this column recorded the percentage change in the weighted IMP of groups of securities (according to Rotation Group) against their weighted closing price on the previous day (a fixed reference point). ASIC pointed, firstly, to a general but not invariable change in a negative direction at the start of the Cancellations (19 April 2012) and then throughout the period of the Cancellations for each Rotation Group (showing, on ASIC’s argument, decreasing prices) and a change in the positive direction at the start of the Reductions (other Serial Expiry Days) and then throughout the period of the Reductions for each Rotation Group (showing, on ASIC’s argument) increasing prices.
  7. With reference to the same tables, ASIC then pointed to what it described as a general correlation between the size of individual Cancellations (19 April 2012) and individual Reductions (other Serial Expiry Days) (gleaned from the buy or sell columns in the tables) and the size of immediate movements in the percentage figures recorded in the last column.
  8. ASIC submitted that a similar phenomenon could be observed in other groups of tables being those found at Tabs 28, 65, 100, 135 and 171. These tables record the size of Whitebox’s buy and sell orders and the IMP or, after the OSPA for the Rotation Group, the latest traded price (in the column “Indicative XJO”).
  9. ASIC argued that, in principle, one would expect any immediate price impact of the Cancellations (19 April 2012) and Reductions (other Serial Expiry Days) to be reflected to some degree in immediate changes to the Indicative XJO (a decrease in the case of the Cancellations and an increase in the case of the Reductions). ASIC argued that these changes can be seen in these particular tables, although “the correlations are subtler and less consistent than those seen in the Group Weighted Prices Tables”.
  10. It is important to note that ASIC’s submissions with respect to Mr Clifford’s tables were confined to asserted impacts on the IMP, following my rulings in Whitebox No 6.

Conclusions on additional evidence

  1. Except in a clear and simple case, I am not prepared to draw conclusions from raw data simply on the basis of arguments advanced by one party or another about it, unaided by appropriate expert evidence. To do so carries the significant risk, in a complex case such as this, of uninformed misinterpretation compounded by the further risk of false or misleading conclusions.
  2. The extent to which conclusions can be drawn from the raw data in Mr Clifford’s tables is a case in point. ASIC’s reliance on the directional change in prices shown in the data in Tabs 20 – 24 (19 April 2012), 57 – 61 (17 May 2012), 92 – 96 (19 July 2012), 127 – 131 (16 August 2012) and 163 – 167 (18 October 2012) was qualified by the statement that “generally (although not invariably)” the figures move in a direction supporting the proposition that the Cancellations (19 April 2012) caused a decrease in the prices of the XJO Securities concerned, and that the Reductions (other Serial Expiry Days) caused an increase in the prices of the XJO Securities concerned. However, much resides in the statement “generally (although not invariably)”. The defendants demonstrated that, in the case of the Group 2 securities on 16 August 2012, some of Reductions reflected increased prices while some of the Reductions reflected decreased prices, thereby showing in that series of orders that ASIC’s general proposition about directional change did not hold good. What is more, there appeared to be no correlation between the size of the individual groups of Reductions and the size of the movements that were said to have been caused by them, contrary to the other proposition that ASIC advanced with respect to these tables.
  3. As to the other tables referred to by ASIC, I have already noted that ASIC advanced them as showing correlations that were “subtler and less consistent” than the correlations I have just addressed. No more need be said about the degree to which reliance can be placed on them.
  4. Similar difficulties abound with the auction time lines. Dealing with the 18 October 2012 auction time line itself, the first thing to note about ASIC’s tables (Exhibits 50 and 51) is the difference in the line numbers between the start and end of the Reductions in each Rotation Group, and the difference in line numbers between the end of the Reductions for a particular Rotation Group and the commencement of the OSPA for that Group. Reference to the auction time line shows that, in these intervals, the direction in movement of Underlying Basis, and the direction in movement of Fast Basis, is not linear but quite often changes direction in the course of the Reductions. Further, there is no necessary correlation in either the direction or the size of the stepwise movements in Underlying Basis compared with the stepwise movements in Fast Basis (although some correlations exist). Further, the auction time line indicates that, in the course of some of the Reductions, new baskets were entered by other traders. In one case, new baskets were entered after the Reductions ended and the OSPA for that Group Rotation began. I should add that the auction time lines do not attempt to record the very large number of interposed orders that one can expect to have been placed at around this time, as identified by Professor Aitken’s and Associate Professor Prokhorov’s respective analyses.
  5. The fact that such variations exist was confirmed by Mr Morgan in cross-examination, but that is as far as his evidence went on this score. Given that Basis can be taken to be a reflection of price, these matters call for interpretation and elucidation. They presage the fact that the untutored analysis that ASIC invites is far more complex than its submissions suggest. The possible significance of these matters for the IMP needs to be explained through appropriate expert evidence. They have not been explained.
  6. There are, however, more fundamental difficulties with ASIC’s submissions based on the auction time lines. In oral closing submissions, the defendants illustrated a particular difficulty with ASIC’s foundational proposition that any relative movement between Fast Basis and Underlying Basis in the auction time lines is the consequence of the placement, amendment or cancellation of Whitebox’s orders.
  7. First, contrary to ASIC’s submissions, Mr de Kantzow did not give this proposition the unconditional confirmation that ASIC’s submissions might suggest. Mr de Kantzow went no further than to accept the inherent limitation in the question put to him—that looking at the difference between Fast Basis and Underlying Basis will give “some idea” of the impact of Whitebox’s orders.
  8. Secondly, and critically, ASIC’s foundational proposition is not sound. This was convincingly demonstrated by the defendants’ reference to the 16 August 2012 auction time line which shows that, without any order activity by Whitebox in the 50 second period before the earliest commencement of the Group 1 Rotation (a not insignificant period of time for this particular form of trading), Fast Basis and Underlying Basis moved contrariwise to each other. The point of this example was to show that, not only did Fast Basis and Underlying Basis move away from each other significantly (rather than relative to each other), but they moved without any order activity whatsoever by Whitebox.
  9. For these reasons, I am not prepared to draw the conclusions from the auction time lines and Mr Clifford’s tables that ASIC invites me draw. I do not accept that the auction time lines or Mr Clifford’s tables necessarily complement or support Mr Graves’ opinion that the placement of the Cancellations (19 April 2012) and the Reductions (other Serial Expiry Days) was likely to have an impact on the Opening Price of XJO Securities on the Serial Expiry Days in question (in the sense in which this inquiry was understood and treated by Mr Graves). All that one can say confidently is that Whitebox’s orders were part of the Order Book to which the ASX Algorithm was applied as the orders of all traders were placed, amended or cancelled, and thus contributed to the formation of the IMP from time to time with respect to the securities in question.

THE REMAINING PILLARS OF ASIC’S CASE

The fifth pillar

  1. The fifth pillar has two aspects. The first is that there was a delay in making the Cancellations and Reductions for securities to be traded in the last four Rotations of the OSPA in which the relevant securities were due to trade. The second is that, by making the Incremental Reductions, Whitebox was placing orders that the defendants did not intend to trade.
  2. The nub of this pillar is ASIC’s submission that there was an awareness on the part of the defendants that the Buy Orders (19 April 2012) and the Amended Sell Orders (other Serial Expiry Days) would never trade and that the Cancellations (19 April 2012) and the Reductions (other Serial Expiry Days) are explicable only by reference to the fact that the defendants were seeking to preserve the effect of Whitebox’s order volume withdrawal on the market so as to ensure that the withdrawal exerted an influence on the setting of Opening Prices for the relevant XJO Securities in the OSPA. In short, the Buy Orders and the Amended Sell Orders were never intended to be traded but were, instead, a mechanism by which to manipulate the market.
  3. In its Revised Consolidated Particulars, ASIC stated that the precise time when Whitebox’s index arbitrage position was set was the placement of the orders constituting the Cancellations and the Reductions for the Group 1 securities (ASIC said that, for 16 August 2012, the precise time was the placement of the last set of August Reductions for the Group 1 securities).
  4. The defendants submitted that the evidence does not support the conclusion that the defendants definitively set the size of Whitebox’s orders by the time of the Rotation for Group 1 securities. They argued that ASIC’s case proceeds on the false assumption that mispricing does not change after Group 1 opens and that “an arbitrageur’s calculus of opportunity and risk is necessarily fixed after that time”.
  5. In this connection, the defendants pointed, once again, to Whitebox’s trading on 16 February 2012 to illustrate the general proposition (which Mr Morgan accepted) that, in principle, index arbitrage trading can be undertaken by trading each leg in different sizes and at different times. On 16 February 2012 (which, as I have said, was also a Serial Expiry Day, although not one of the days in question here), Whitebox held a position size of 100 SPHE in sell orders at the opening of the Rotations for Groups 1 to 4. However, by the time of the commencement of the Rotation for Group 5, it had increased its order size to 171 SPHE by selling, in live trading, further baskets of securities in Groups 1, 2 and 3, and increasing its sell order for securities in Group 5. Throughout this time, mispricing was such that a negative arbitrage remained profitable, and it seems that Whitebox pursued that profit opportunity.
  6. In cross-examination directed to Whitebox’s order activity on 16 February 2012, Mr Morgan accepted that it was rational for an arbitrageur to adjust his or her orders depending on market conditions as they existed throughout the OSPA. However, in his report, Mr Morgan advanced the contrary proposition:
    1. Having bought or sold XJO Securities in the first rotation of the OSPA for the purposes of executing an index arbitrage strategy a trader would then need to continue to buy or sell XJO Securities in the subsequent rotations in the same size relative to their weight in the XJO index, as those securities already traded. If subsequent rotations were traded in different relative sizes, then the position would be unmatched without further trading in the XJO Securities that had already begun trading and could expose the trader to market risk due to the unbalanced legs.
  7. Professor Frino disagreed with this part of Mr Morgan’s report and advanced a view similar to that which Mr Morgan came to accept in cross-examination:
6.62 Paragraph 102 suggests that an index arbitrage trader having executed orders in the first rotation for ASX 200 Stocks would continue to execute orders in the other ASX 200 Stocks in the same size relative to their weight in the Index as the securities already traded. That could be the position, but I do not agree that it is a universal proposition that would apply to all index arbitrage trading. Index arbitrageurs can and do behave differently. If an opportunity arises for there to be greater trading for securities in the later rotations, those opportunities may be taken up which may lead to adjustments to the level of the securities traded in the earlier rotations through later trading. Also, the approach that Mr Morgan has set out seems to be predicated on traders trading to perfectly replicate the Index in executing an arbitrage position. If a proxy portfolio is used to execute an arbitrage position, then this description is also not correct.
  1. The defendants submitted that there is nothing unusual in the fact that the Buy Orders were not cancelled and the Amended Sell Orders were not reduced until shortly before each relevant Rotation. They submitted that, for the Serial Expiry Days in question, the position at the Group 1 Rotation should be taken as Whitebox’s then best estimate of the “appropriate overall position” in light of likely mispricing through the Rotations and the availability of SPI Futures to hedge its position at favourable prices. But, they submitted, Whitebox should not be taken as having “set” its hedge position at the Group 1 Rotation, bearing in mind the example provided by Whitebox’s trading on 16 February 2012. The defendants submitted that it was entirely rational and commercially legitimate for Whitebox to have maintained the Buy Orders and the Amended Sell Orders against the possibility that conditions could move in favour of executing a larger position size than had been executed for the Group 1 securities.
  2. The defendants also submitted that the making of staged reductions, by way of the Incremental Reductions, was inconsistent with ASIC’s case theory that the defendants were seeking to manipulate the price of XJO Securities. They submitted that, if that had been the purpose, it would have been best served by making the reductions as large and as late as possible.
  3. It is incontrovertible that the Cancellations and Reductions for securities in Groups 2 to 5 were placed close to the earliest commencement of the Rotations in the OSPA for those Groups. This fact is consistent with ASIC’s fifth pillar, but it does not establish it.
  4. It is also incontrovertible that, on the Serial Expiry Days in question, the overall arbitrage position that Whitebox effected was one that reflected its arbitrage position at the Group 1 Rotation. This fact is also consistent with ASIC’s fifth pillar but, on the evidence before me, I am not persuaded that Whitebox’s position was necessarily “set” at that Rotation, such that the Cancellations and Reductions effected with respect to the Groups 2 to 5 securities close to the earliest commencement of the Rotations in the OSPA for those Groups are explicable only by the defendants seeking to preserve order volume so as to affect the Opening Prices for those securities in the OSPA. Based on the example provided by Whitebox’s order activity on 16 February 2012, it is at least a real and equal possibility that Whitebox was doing no more than keeping its options open, should there be a change in market conditions during the course of the OSPA. The fact that no changes eventuated, such as to move Whitebox to change its eventual position from that adopted at the time of the Group 1 Rotation, does not gainsay my acceptance of that possibility as a real and equal one.
  5. I accept that the Cancellations and the Reductions, particularly those Reductions not effected by the Incremental Reductions, involved, in a number of cases, significant volumes of securities. This fact is also consistent with the fifth pillar. But ASIC’s submissions proceed on an acceptance that the defendants had an awareness that, by making the Cancellations and Reductions at the time they were made, they were likely to affect the Opening Prices at which the securities traded in the OSPA. I am not persuaded on the evidence that the defendants necessarily had this awareness and then acted on it for the purpose of affecting the Opening Prices when Whitebox placed the Cancellations and Reductions. I am certainly not prepared to make such a finding based on Mr Graves’ rules of thumb or the unexplained data which ASIC advanced, for the reasons I have given above.
  6. For completeness, I am not persuaded on the evidence that the defendants had an awareness that, at the time of their placement, the Buy Orders and the Amended Sell Orders would never trade. The evidence falls short of such a finding. The defendants would have known that the Buy Orders and the Amended Sell Orders could be amended or cancelled up to the commencement of the relevant Rotations for the securities in question. The possibility that the Buy Orders and the Amended Sell Orders might be amended or cancelled, as events unfolded on the days in question, is likely to have been in their minds. But, as I have found in relation to the first pillar, arbitrageurs place orders to pursue the possibility of profit arising from mispricing. Orders may well be placed in contemplation of a perceived opportunity.
  7. Finally, as to the Incremental Reductions, it can be accepted that at the time of each subsequent reduction in the series of reductions, the defendants knew that the earlier reductions would not trade. That observation is a truism. Further, as I have said, I do accept on the balance of probabilities that, in making the Incremental Reductions, it would have been in the defendants’ contemplation that there was a real possibility that each successive reduction (other than the last made) might not trade and that the most likely scenario was that a hedge would be achieved with respect to the futures leg which, by then, Whitebox had traded. But I repeat: this is far from accepting that the Incremental Reductions were placed to maintain order volume with the purpose of manipulating the Opening Prices of the relevant XJO Securities in their respective OSPAs.

The sixth pillar

  1. The sixth pillar directs attention to what ASIC described as a pattern of trading—the placement of large orders for XJO Securities which were then cancelled or reduced shortly prior to the relevant Rotations in the OSPA which were repeated “over a long period of time”. ASIC submitted that, because of this pattern, Whitebox’s placement of large orders, and then the removal of order volume shortly prior to the commencement of the Rotations in the OSPA, should not be seen as mere happenstance. Rather, on ASIC’s case, this should be seen as “a deliberate and concerted approach adopted on multiple trading days”. ASIC submitted that it can be inferred from the fact that Whitebox never implemented or sought to implement an arbitrage position commensurate with the large orders it had placed that the defendants’ never had a genuine intention to do so.
  2. As I have noted at [282] above, ASIC relied, relatedly, on a pattern in Whitebox’s XJO Securities positions from April to October 2012—100 SPHE (April), 200 SPHE (May), 300 SPHE (July), 400 SPHE (August) and 500 SPHE (October)—that, in its submission, was consistent and uniform across the relevant Serial Expiry Days.
  3. I accept that, at a relatively high level of generality, the patterns that ASIC identified are discernible. However, in light of the findings I have made, I am not persuaded that these patterns should bear the character, or have the significance, that ASIC attributes to them.
  4. First, in light of my findings in relation to the fifth pillar, the fact that Whitebox placed orders that could be described as “large orders” which were then cancelled or reduced shortly prior to the relevant Rotations in the OSPA, is explicable by reference to the nature of index arbitrage trading, the nature of order activity in the Pre-Open Phase (particularly by arbitrageurs); and the conditions existing on the days in question in the Pre-Open Phase, as revealed by the auction time lines. As I have said, I am not persuaded on the evidence that the defendants had an awareness that, at the time of their placement, the Buy Orders and the Amended Sell Orders would never trade, although the possibility that they might not trade was likely to have been in their contemplation. I am also not persuaded on the evidence that the defendants necessarily had an awareness that the Cancellations and Reductions were likely to affect the Opening Prices at which the relevant securities would be traded in the OSPA and acted on that awareness when placing the Cancellations and Reductions for the purpose of affecting the Opening Prices. Once that is accepted, the fact that a pattern of the kind identified by ASIC can be observed is a factor that is relevant but, overall, one of diminished significance.
  5. Secondly, while the pattern that ASIC has identified in Whitebox’s XJO Securities position on the Serial Expiry Days in question is observable, its consistency and uniformity is not as neat as ASIC would have it. Further, as the defendants submitted, this “pattern” (the defendants say there is no real pattern) must be seen in the context of other expiry days, as recorded in the defendants closing written submissions:
Expiry
Cash Orders in SPHE
Type of expiry
Feb 2012
171
Serial
Mar 2012
107
Quarterly
April 2012
138
Serial
May 2012
200
Serial
June 2012
296
Quarterly
July 2012
299
Serial
Aug 2012
402
Serial
Sept 2012
44
Quarterly
Oct 2012
500
Serial
  1. While the broad pattern that ASIC has identified for Serial Expiry Days in question is capable of raising suspicion about the size of Whitebox’s XJO position on those days, it would be of greater significance if other elements of ASIC’s case were established on the evidence. As matters stand on the findings I have made, I do not attribute to this particular matter the weight that ASIC asks me to place on it.

The seventh pillar

  1. The seventh pillar directs attention to the fact that the defendants were incentivised to obtain the greatest profit possible by reason of Whitebox’s trading. I accept that that is likely to have been the case on the Serial Expiry Days in question, as no doubt it also was on other days when Whitebox traded on behalf of NAB. As ASIC properly accepted, this fact alone does not establish the impugned orders were placed with a manipulative purpose.

The eighth pillar

  1. The eighth pillar is directed to the contention that Whitebox’s trading systems allowed it to monitor the effect on mispricing that would be caused by the withdrawal of its orders for XJO Securities. ASIC submitted that the defendants were keenly aware of the impact that the placement, amendment and cancellation of Whitebox’s orders for XJO Securities had on the market, particularly through the withdrawal of order volume.
  2. ASIC submitted that Whitebox’s orders for XJO Securities were adjusted so as to affect Basis, which necessarily bespeaks an intent to affect the prices of XJO Securities in order to alter the level of the XJO. It submitted that this was evident from Mr Boshoff’s email to Mr Stevenson (NAB) on 26 October 2012 in respect of Whitebox’s trading on 18 October 2012. It is convenient to repeat part of the email I have already quoted above:
Referring to table 1 provided by Mallesons the amend that we did at 9.48:14 from nominal 2000 futures was a response to basis spiking up to 17bps. Our amendment to 1900 nominal futures brought basis back to 7.2bps and our system indicated that an amendment of a further 100 lots would take it to -14bps. Our experience is that as the market opening approaches basis generally becomes less volatile and we are then able to accurately calibrate it to provide us with enough mispricing to commercially justify our basket...
(Errors in original; emphasis added)
  1. ASIC submitted that this passage is significant because it illustrates that the defendants were regularly monitoring the effect that removing order volume would have on Basis, and thus the likely profitability of the trading that Whitebox was conducting. It showed that Whitebox’s systems were programmed and operated to reveal how Basis would be affected if order volume was removed.
  2. ASIC submitted:
In addition, and critically, this email reveals that the Defendants sought to exploit this capability in order to generate mispricing within the market that they could exploit for profit. In particular, Mr Boshoff’s references to “accurately calibrat[ing] [basis] to provide us with enough mispricing to commercially justify our basket” make clear that the Defendants were not merely attempting to profit from mispricing that existed in the market. Rather, they were nakedly attempting to manipulate Basis for the purpose of making their trading profitable or more profitable.
  1. ASIC also submitted that these matters put in context the large orders that Whitebox placed in the period leading up to the OSPA which were then cancelled or reduced in volume shortly prior to each Rotation. ASIC submitted that the orders were not placed with any realistic expectation that they might be traded. Instead, they were the means by which the defendants were able to “calibrate” Basis and so generate a profit or a greater profit and “commercially justify” Whitebox’s basket.
  2. I have already observed that Mr Boshoff’s use in the email of the word “calibrate” is strange. I have accepted the defendants’ submission that Whitebox could not “calibrate” Basis: see at [128].
  3. Whitebox’s ability to detect Underlying Basis meant that it could identify the mispricing that would exist in the absence of its own orders because of the operation of the ASX Algorithm. The calculation of Underlying Basis is necessarily hypothetical because, in fact, the orders in question are in the Order Book. Moreover, the calculation is undertaken only at a particular point in time. If, for example, a particular trader’s order is removed, Underlying Basis does not, and cannot, reveal what will happen in the market. It cannot predict how others in the market will react and thereby affect Basis.
  4. That said, an ability to observe Underlying Basis will give some insight into the direction of the market and the flow of orders. For example, Mr Morgan accepted the following propositions in cross-examination:
(a) If an arbitrageur is not changing his or her own orders, but Underlying Basis is moving, that gives the arbitrageur some insight as to the direction of the market and the flow orders.
(b) Underlying Basis is a useful information cue to understand what is happening in the Order Book.
(c) Over time, Underlying Basis will show the direction, trend and persistence of mispricing, particularly where Basis has been swinging between substantially positive and substantially negative.
  1. As the defendants correctly submitted, Whitebox’s ability to observe Underlying Basis did not also give it the ability to generate mispricing. Further, Underlying Basis did not give it the ability to know what would happen in the market if its orders were removed (cancelled or amended). It could not tell Whitebox the unknowable—that is, how, and the extent to which, other traders and arbitrageurs would or might react to Whitebox’s own order amendments.
  2. I accept the defendants’ submission that no inference of nefarious intent can be drawn simply from Whitebox’s use of Underlying Basis.

THE DEFENDANTS’ CRITICISM OF ASIC’S CASE THEORY

  1. As I have noted, ASIC’s case is that, on the Serial Expiry Days in question, the defendants sought to manipulate the Opening Price of XJO Securities by placing large orders that were never intended to be traded, and then cancelling or amending those orders shortly before the Rotations for each Group in the OSPA. The defendants submitted that such a scheme would be irrational and, therefore, improbable because it would have required Whitebox to place order amendments at the right time and in the right volumes to bring about this price effect. But, the defendants submitted, it would be practically impossible to predict, with any accuracy or certainty, the impact that particular orders or amendments might have, particularly on the larger securities that make up the XJO. This is because the lead up to the OSPA, particularly in the last two minutes, is “a hive of activity”, when many thousands of orders can be placed, including those placed by large, computerised traders. These orders can be placed within very short periods of time, sometimes within the same second. Amongst other evidence, the defendants pointed to the analyses undertaken by Professor Aitken and Associate Professor Prokhorov. Further, the time of each Rotation is not known and the potential impact of a trader’s orders in respect of individual XJO Securities is both unknown and practically unknowable.
  2. The defendants further submitted that, if one were seeking to achieve a price effect on the XJO itself, it would be necessary to affect a sufficient number of the individual securities that make up the XJO to achieve that outcome. However, some securities are less liquid than others. Mr Graves accepted in cross-examination that, across the 200 securities, and on a given trading day, the liquidity of particular securities will differ, and not necessarily in a way that reflects their proportion in the XJO. Mr Graves accepted that it would be perfectly possible that, on a given day, a particular security might have unusually low liquidity, even by reference to its own typical pattern. The defendants submitted that, because some securities are less liquid than others, if one were to make the same proportionate change across all the XJO Securities or across all of one Group, there would be a real chance that there would be a larger effect on the price of some securities than on others, because of variations in liquidity between them on a particular day. Further, if one were to leave amendment orders to very late in the piece, and amend across a whole Group, there would be a real chance of causing a price spike in a particular security, especially a smaller one that was not very liquid. Therefore, if one were attempting to manipulate price by making amendments across whole Groups—and Whitebox’s cancellations and amendments were across whole Groups—“one would risk failing to move the index very much, but bringing about a significant price spike in less liquid stocks, which would result in one’s conduct coming to the attention of ASIC”. This scenario was put to Mr Graves, who agreed it would not be a rational strategy.
  3. The defendants further submitted that if, as ASIC alleged, Whitebox had, in fact, embarked on a scheme to deliberately amend (reduce) orders so close to the earliest time of each Rotation that the market would not have time to react, then it could be expected that there would have been a consistent pattern of amending very close to the earliest time for each Rotation so as to maximise the likelihood of a price impact carrying through to the opening for each Group. The defendants submitted that, on the objective evidence, there was no such pattern.
  4. I have, in this connection, already referred to the differences between ASIC and the defendants on when Whitebox’s order reductions should be seen as having commenced. I have also summarised Whitebox’s order activity on the Serial Expiry Days in question. The defendants submitted that, when regard is had to the auction time lines for those days, there was considerable variation in both the size and timing of Whitebox’s order amendments which was consistent with Whitebox having responded to market conditions initially for Group 1 and then, in the absence of any favourable market movement, successively (and, in some cases, incrementally) for the other Groups as the OSPA progressed. Correlatively, the defendants submitted that this activity was not consistent with conduct that was “simply in order to have an impact on the immediate Indicative Match Prices or opening prices”. The defendants further submitted that ASIC’s “central proposition” that Whitebox cancelled or amended orders shortly prior to the respective Rotations of the OSPA in which the XJO Securities the subject of those orders were due to trade, elided the fact that, in most cases, Whitebox’s order amendments were made in “a staggered series of relatively modest increments over relatively lengthy periods”.
  5. The defendants illustrated the latter submission by reference to Whitebox’s order reductions in Group 4 on 16 August 2012. This involved a reduction from 1700 SPHE to 400 SPHE by 13 steps over nearly seven and a half minutes such that, with almost 45 seconds until the earliest commencement of the Rotation for that Group, all that remained to be made of the overall reduction of 1300 SPHE was 230 SPHE.
  6. The defendants also illustrated their submission by reference to Whitebox’s order reductions in Group 2 on 18 October 2012. This involved a reduction from 2000 SPHE to 500 SPHE by 17 steps over nearly eight minutes such that, with 45 seconds until the earliest commencement of the Rotation for the Group, all that remained to be made of the overall reduction of 1500 SPHE was 440 SPHE.
  7. The defendants asked rhetorically: How could the Court be satisfied that these order amendments would be likely to have an effect on the IMP of the securities in those Groups, let alone one that would be likely to endure into the Opening Prices so as to establish an “artificial” price?
  8. In this connection, the defendant also pointed to ASIC’s acceptance in opening address that the potential impact on Opening Prices of a staggered reduction was obviously less than for a single, one-off reduction for the same volume. The defendants submitted that this must be so as a matter of common sense because a staggered reduction “would necessarily give the market more time to react to the change which was occurring”.

THE “MARKET FOR” CASE

Introduction

  1. Relevantly to this limb of ASIC’s case, s 1041B(1)(b) provides that a person must not do an act that has or is likely to have the effect of creating or causing the creation of a false or misleading appearance with respect to the market for financial products on a financial market operated in this jurisdiction. The acts which ASIC relies on are the placements of the Buy Orders (19 April 2012), and the Amended Sell Orders (other Serial Expiry Days) and the Incremental Reductions (19 July, 16 August and 18 October 2012). ASIC submitted that the placement of these orders had, or was likely to have, the effect of creating a false or misleading appearance as to the market because they appeared to the market to be genuine expressions of interest to trade the securities the subject of those orders.
  2. This limb is grounded on the central tenet of ASIC’s case that the Buy Orders, the Amended Sell Orders and the Incremental Reductions were not placed with any genuine intention that they should be traded but, rather, with the knowledge that, ultimately, they would be cancelled or substantially reduced in volume so as to manipulate the Opening Price for each of the XJO Securities in their respective Rotations in the OSPA.
  3. Consistently with the main object of Ch 7 of the Act, and the statements made in Lloyd, Fame Decorator and Marra Developments quoted at [16] – [19] above, ASIC submitted that an order placed on the ASX is taken to be a genuine expression of an intention to buy or sell securities in accordance with the order. However, if it is not a genuine reflection of supply and demand, then the order will necessarily create a false or misleading appearance with respect to the market.
  4. In this connection, in Marra Developments, Mason J said (at 59):
I agree with Hope and Samuels JJA in rejecting the suggestion that the section strikes only at fictitious or colourable transactions. Transactions which are real and genuine, but only in the sense that they are intended to operate according to their terms, like fictitious or colourable transactions, are capable of creating quite a false or misleading impression as to the market or the price. This is because they would not have been entered into but for the object on the part of the buyer or of the seller of setting and maintaining the price, yet in the absence of revelation of their true character they are seen as transactions reflecting genuine supply and demand and having as such an impact on the market.
When purchases have been made of shares in a company at or about a particular level for the purpose of setting and maintaining a market price for those shares, there is a breach of the statutory prohibition. At the very least purchases have then been made which are calculated to create “a false or misleading appearance with respect to the market for, or the price of” the shares. In reality the purchases are calculated to create a false market or false price. The false or misleading appearance is that the market, in the absence of any disclosure that a market support operation is on foot, appears to be real or genuine, there being no overt sign of market support or manipulation.
  1. Also, in Australian Securities and Investments Commission v Nomura International plc [1998] FCA 1570; (1998) 89 FCR 301 (Nomura), Sackville J said (at 411 – 412):
... Nomura neither intended nor expected that any of the offers contained in the Ask Basket would be accepted by bidders on the ASX. At the time the Ask Basket was placed, Nomura had already decided to sell the securities by means of the aggressive strategy to be embodied in the March Sale Orders. Nomura's only reason for placing the Ask Basket was to conceal from a particular broker its intention of placing the March Sale Orders in the terms already discussed. I do not think that this is a case of confusion of the kind dealt with in Puxu. Nomura avowedly set out to confuse Were. While Mr Channon and Mr Moss had a number of concerns, including the fear that Were or its clients might engage in front running, their overriding motivation was to protect the strategies to be implemented later that day. The placement of the Ask Basket represented to Were that Nomura, insofar as it was a seller of securities, wished to maximise the price it would receive for those securities. The placement of the Ask Basket represented to Were that Nomura was a seller of securities seeking to maximise the price of those securities. Nomura's intentions were quite different. It intended to sell the securities comprised in the Ask Basket aggressively at the close of trading, in order to drive down the closing price of the All Ords, in the process hitting the Bid Basket.
The central object of Nomura's strategies was to produce a fall in the All Ords near the close of trading on the ASX on 29 March 1996. The placement of the March Sale Orders, with the intention of hitting the Bid Basket in a large number of securities, was designed to ensure that the object was achieved. The misleading of Were as to Nomura's true intentions was an integral element in a series of strategies designed to “move the close”. Nomura was not simply using accepted or standard market techniques to achieve legitimate commercial objectives. Nomura engaged in deliberately misleading conduct as part of strategies designed to achieve illegitimate ends.
It is true that there was no evidence as to whether Were's conduct was influenced by the representation implicit in the placement of the Ask Basket. However, it is not difficult to infer that Nomura's deliberate conduct was likely to achieve the desired result. Were did not learn that it was to sell portion of Nomura's basket of securities until 2:30 pm on 29 March 1996 and did not receive precise instructions (for the sale of liquid securities) until 3:38 pm on that day. It is uncertain what Were would have done had it appreciated before noon on 29 March 1996 that Nomura intended to sell aggressively near the close into its Bid Basket. As it happened, Mr Crabb explicitly adverted to that possibility in his conversation with Mr Moss at 3:38 pm. However, in consequence of the placement of the Ask Basket, Were laboured for three and a half hours under the misapprehension that Nomura intended to maximise the returns from its holdings of securities. If it matters, there was at least a real chance that Were, had it known of Nomura's true intentions much earlier, would have appreciated the regulatory consequences and taken appropriate action.
In these circumstances, Nomura engaged in conduct which was misleading or deceptive, or likely to mislead or deceive, in contravention of s 52(1) of the TP Act.

An issue of construction

  1. The parties were at issue as to both the construction of s 1041B(1)(b) and as to the approach that should be adopted to its application in respect of ASIC’s “market for” case.
  2. The issue of construction was the meaning to be given to the word “likely” in the phrase “has or is likely to have the effect”. ASIC contended that the word “likely” bears the meaning of “a real and not remote chance” of an outcome being achieved. “Likely” does not mean “more probable than not”. This construction is supported by Westpac at [2105] (Beach J) (in the context of s 1041A, at [1961]) and, in the context of s 1041E of the Act, by Casaclang v WealthSure Pty Ltd [2015] FCA 761; (2015) 238 FCR 55 at [257] (Buchanan J).
  3. The defendants submitted that “likely” must be construed as “more probable than not”. This construction is supported by Nomura where Sackville J (at 396) held that “likely” in the context of s 998(1) of the Corporations Law (“... likely to create a false appearance of actual trading ...”) means “more probable than not”. In coming to this conclusion, Sackville J was influenced by the fact that contravention of s 998(1) was a criminal offence and that, where ambiguity is present, the provision should be construed in a manner more favourable to a defendant. In the present case, the defendants submitted that the same approach should be adopted in relation to s 1041B(1)(b) and s 1041A, for the same reason.
  4. The defendants also referred to the line of authority under the Trade Practices Act 1974 (Cth) established in Tillmanns Butcheries Pty Ltd v Australasian Meat Industry Employees’ Union [1979] FCA 85; (1979) 42 FLR 331 at 346, which continues to prevail under the Competition and Consumer Act 2010 (Cth), that “likely” means “a real and not remote chance”. The reason for going to this line of authority was to challenge the correctness of Beach J’s conclusion in Westpac. The defendants argued that the “real chance” test was criticised by Buchanan J in Australian Competition and Consumer Commission v Metcash [2011] FCAFC 151; (2011) 198 FCR 297 (Metcash) and that it appeared that, in Westpac, Beach J was not taken to this criticism which, the defendants said, was “compelling”. Further, the defendants argued that Beach J overlooked the importance of the fact that, after Nomura had been decided, Parliament enacted ss 1041A and 1041B using the same language of “likely”. The defendants submitted that where language in a statute has received judicial interpretation and the Parliament employs the same language in a subsequent statute on the same subject matter, it is presumed that Parliament intended the same interpretation to be given to the words so used: Salvation Army (Vic) Property Trust v Ferntree Gully Corporation [1952] HCA 4; (1952) 85 CLR 159 at 174; R v Reynhoudt [1962] HCA 23; (1962) 107 CLR 381 at 395 – 396; 407 – 408.
  5. Given that this question has been recently considered by Beach J in Westpac, and given that I am not persuaded that his Honour was clearly wrong in the conclusion to which he came, I propose to give “likely” the meaning of “a real and not remote chance”.

The approach to applying s 1041B(1)(b)

  1. The defendants did not dispute that s 1041B(1)(b) has application to the Pre-Open Phase in which the impugned orders (i.e., the Buy Orders, the Amended Sell Orders and the Incremental Reductions) were placed. They did submit, however, that there is no actual “market” in existence at that time and that the “appearance” created by an order placed in the Pre-Open Phase is necessarily fundamentally different to the “appearance” of an order placed in the Open Phase.
  2. The defendants’ line of argument challenged ASIC’s contention that this limb of its case should be approached by recognising that the notion of “genuine supply and demand” is formed by forces that are created in a market by buyers whose purpose is to acquire at the lowest possible price and sellers whose purpose is to sell at the highest realisable price.
  3. ASIC’s contention in this regard was based on the following passage in JM at [71]:
    1. The forces of “genuine supply and demand” are those forces which are created in a market by buyers whose purpose is to acquire at the lowest available price and sellers whose purpose is to sell at the highest realisable price. The references in s 1041A to a transaction which has, or is likely to have, the effect of creating an “artificial price”, or maintaining the price at a level which is “artificial”, should be construed as including a transaction where the on-market buyer or seller of listed shares undertook it for the sole or dominant purpose of setting or maintaining the price at a particular level. It is, however, important to emphasise that whether there are other kinds of transaction which have the effect of creating or maintaining an artificial price in a market for listed shares is not, and, given the terms of the case stated, should not be, decided.
  4. Plainly, these observations were made with respect to the application of s 1041A in the context of on-market (or “live”) trading, even though the case stated under s 302 of the Criminal Procedure Act 2009 (Vic) from which the appeal to the High Court was brought makes clear that the conduct impugned in that case was the placing of a buy order in the Pre-Open Phase after Normal Trading (which had the effect of lifting the price of a security in the Closing Single Price Auction): see Director of Public Prosecutions (Cth) v JM [2012] VSCA 21; (2012) 37 VR 1 at [16] – [18].
  5. The defendants submitted that ASIC’s approach to s 1041B (and also s 1041A) was “misconceived” because ASIC framed the issue for determination by reference to “forces of genuine supply and demand”, such that any order not actuated by an immediate intention to buy or sell in accordance with the order, was not formed by these forces, with the consequence that the order, so placed, did not represent “genuine supply and demand”. And, if the order did not represent “genuine supply and demand”, it necessarily created a false or misleading appearance with respect to the market.
  6. The defendants submitted that this approach was simplistic because it ignored the inherently contingent nature of orders in the Pre-Open Phase and thus divorced the statutory language of s 1041B(1)(b) from its relevant context for this limb of ASIC’s case. The defendants submitted that the correct approach was to understand what, if anything, a reasonable observer of, or participant in, the Pre-Open Phase would reasonably understand or assume the placement or amendment of an order to convey: Rosenberg v ASIC [2010] AATA 654; (2010) 117 ALD 582 at [90] – [95]. This, in turn, directs attention to the assumptions on which those participating in the Pre-Open Phase operate, because it is only conduct that falsifies those assumptions that gives rise to a false or misleading appearance with respect to “the market” in which the securities are traded: see Fame Decorator at 63.
  7. Following this line, the defendants submitted that “the range of ordinary or legitimate commercial strategies and behaviours in the Pre-Open Phase” comprehended a number of matters.
  8. First, the defendants submitted that no orders placed in the Pre-Open Phase can be executed until the commencement of the Rotation(s) in the OSPA to which the order relates. Therefore, any notion of supply and demand is, at most, “contingent”.
  9. Secondly, the defendants submitted that all orders placed in the Pre-Open Phase are necessarily contingent. There is no promise or representation made that an order will be maintained up to market opening and whether an order will be maintained depends entirely on the person placing the order. In that latter regard, “any number of matters may affect (the) decision”. These include matters that are not even observable to other participants. As the defendants put it:
They extend to any matter which might affect the decision to amend the order—that is, any matter relevant to the decision whether, and where, to deploy capital. For example, an order might be amended because the person placing the order perceives a more favourable opportunity relating to another security on the ASX, or another security on another market. Or the person might simply change his or her mind about the order.
  1. The defendants submitted that this was particularly so in relation to orders forming one leg of a securities index arbitrage:
The economics of such a position ultimately depend upon the existence and scale of any mispricing which exists between the SPI Futures price and the Opening Price of the XJO Securities. Whether such an order will be maintained into the opening must depend on the trader’s understanding of the degree of indicative mispricing (if any) as the opening approaches and his judgment as to whether or not mispricing might endure through the opening rotations.
  1. I accept the defendants’ submissions in this regard. These characteristics of orders placed in the Pre-Open Phase were not in dispute. This distinguishes them from orders placed in the Open Phase (or Normal Trading). I also accept that reasonably informed observers or market participants understand the contingent nature of orders placed in the Pre-Open Phase and that it certainly does not follow from an order placed in the Pre-Open Phase that the order will remain in the Order Book until the commencement of the OSPA. I accept that the reasonably informed observer or participant understands that an order placed in the Pre-Open Phase can be cancelled or amended after it has been placed, and may be followed by other orders and even further cancellations and amendments, and that this might occur for any number of reasons. This is borne out by Professor Aitken’s analysis of order amendments, and the timing of order amendments, in respect of the “20 Serial Expiry Days” where he found that slightly below 50% of all orders placed between 9:00 and the OSPA were amended or deleted and 52% of the amendments that reduced volume were placed during the last five minutes before the OSPA.
  2. Thirdly, the defendants pointed to the fact that, on days when an SPI Futures contract is expiring, many of the orders placed are “expiry baskets” which are placed “deep” in the Order Book to ensure execution at the Opening Price so as to close out a futures position. Arbitrageurs placing such orders are generally agnostic as to the Opening Price. This is because any “loss” experienced by the arbitrageur in buying or selling the component securities at the Opening Prices is made good by the concomitant adjustment to the expiring futures’ settlement price which is determined by the Opening Prices on that day. Such arbitrageurs are therefore only likely to be concerned that their cash position is able to be fully traded at the Opening Price for the securities.
  3. The defendants submitted that two consequences followed. The first is that, because some arbitrageurs are indifferent as to price, they have no motivation to obtain the “best” price as either buyer or seller. They are only interested in ensuring that their orders are filled at market opening, at whatever the Opening Price might be. The second is that, because orders are placed “deep in the book”, they are placed well beyond the IMP. The effect of placing such orders is to push prices above the lowest available price for a buyer or below the highest available price for a seller.
  4. The extent of this activity in closing out positions is a likely cause of the very mispricing that other arbitrageurs seek to profit from by seeking to establish index arbitrage positions against the next quarterly expiring futures contract. This then draws attention to the other side of the coin. These other arbitrageurs, who are seeking to establish positions, are also largely indifferent to whether they buy XJO Securities at the lowest available price or sell at the highest available price. Their concern is to capture the mispricing between the price of the constituent XJO Securities and the price of a SPI Futures contract. Their decision to place, amend, cancel or maintain orders will necessarily be affected by their individual perceptions of relative conditions in the market for SPI Futures. Conditions in that market will be constantly changing from 9:50:00 (when the futures market opens) up to and through the Opening Phase of the ASX. Once again, these characteristics of the Pre-Open Phase are not in dispute.
  5. The defendants further submitted that these characteristics of the Pre-Open Phase affect, equally, other investors who wish to ensure execution of an order at market opening. Therefore, observers of or participants in the Pre-Open Phase are not constituted solely by other investors whose orders necessarily reflect the purpose of buying (or selling) at the lowest (or highest) realisable prices.
  6. I accept these submissions. I also observe that these characteristics of the Pre-Open Phase run counter to the paradigm of “genuine supply and demand” formed by forces that are created in a market by buyers whose purpose is to acquire at the lowest possible price and sellers whose purpose is to sell at the highest realisable price.
  7. Fourthly, the defendants submitted, and I accept on the evidence before me, that it is an entirely legitimate and conventional strategy for arbitrageurs to place orders in the Pre-Open Phase against the mere possibility of profitable trading. As the defendants submitted:
The arbitrageur’s decision to place, amend, cancel or maintain such orders will depend upon, amongst other things, the arbitrageur’s perceptions of expected market conditions (both on the ASX and in ... SPI Futures), of opportunity to profit, of applicable limits including inventory constraints, and of risk (including the risk of entering sufficient futures contracts, at favourable prices, to hedge the cash position). Most if not all of those matters are rapidly changeable and involve matters of subjective judgment, impression, intuition or even “gut feel”, particularly with respect to risk.
  1. Fifthly, the defendants submitted, and I accept on the evidence before me, that it is a conventional and legitimate feature of the behaviour of some market participants to disguise, by the manner in which they place orders, their trading intentions. Mr Graves gave this evidence:
    1. As a general statement, investors and traders aim to buy or sell the maximum desired order volume for the minimum price disruption (smallest impact) during the time period they have to execute their trade as this improves their overall financial performance. Managers of substantial orders employ strategies such as splitting their orders for a security into multiple orders with different volumes, prices and times of placement to disguise their actions, thus making it difficult for users and observers to know if there are a number of like-minded investors or traders placing orders, suggesting a change in the buy and sell balance, or if there is a single dominant order placer who is having an effect on the indicative match price. Some users who have substantial orders or orders that are part of an ongoing program, with instructions that they be done over the whole day, execute a portion of their orders in the OSPA as it is typical that a significant portion of the total volume on the day is executed in this single price auction. In my opinion orders received from retail investors prior to the market opening with an instruction that they be executed "at market" are most likely to be placed at a price "deep" enough in the indicative match to ensure full execution during the relevant OSPA. Orders initiated by retail investors in the less liquid securities of the XJO are likely to be more carefully managed as the desired volume may not be able to be traded without moving their price and therefore increasing the total cost of a trade resulting from such orders.
  2. Mr Morgan gave this evidence:
MR McHUGH: But it’s rational for an arbitrager to try to break up their orders, to make it harder for other people in the marketplace to work out what they’re doing.
MR MORGAN: Yes. Sure. Yes.
MR McHUGH: Do you agree?
MR MORGAN: Yes. I agree with that. Yes.
  1. Given the characteristics of orders placed in the Pre-Open Phase as discussed above, the defendants submitted that the placement or amendment of an order conveys only that the person placing the order is, at the time of the order or amendment, pursuing some ordinary or legitimate commercial strategy or behaviour. It does not convey that the person placing or amending the order has anything more than a contingent intention to execute the order at market opening.

Conclusion

  1. I accept that the particular characteristics of the Pre-Open Phase must be recognised and taken into account when applying s 1041B(1)(b) in this case. This means that, for the reasons I have discussed above, the paradigm of “forces of supply and demand” discussed in JM cannot be directly applied to orders, amendments and cancellations in this phase. I accept the defendants’ submission that the placement or amendment of an order in the Pre-Open Phase conveys only that the person placing the order or amendment is, at that time, pursuing some ordinary or legitimate commercial strategy or behaviour. I accept further that the person placing or amending the order does not convey anything more than a contingent intention to allow the order or amendment to execute at market opening.
  2. As the defendants correctly submitted, in order for ASIC to establish the “market for” case it has brought under s 1041B(1)(b) of the Act, it is necessary for it to establish, by evidence, that the defendants were not, at the time that Whitebox placed the impugned orders, pursuing an ordinary or legitimate commercial strategy or behaviour. Critically, ASIC’s “market for” case turns on it establishing that the Buy Orders, the Amended Sell Orders and the Incremental Reductions were not placed with any genuine intention that they should be traded but, rather, with the knowledge that, ultimately, they would be cancelled or substantially reduced in volume so as to manipulate the Opening Price for each of the XJO Securities in their respective Rotations in the OSPA. It is this alleged lack of genuine intention which, on ASIC’s case, had, or was likely to have had, the effect of creating a false or misleading appearance as to the market for financial products on a financial market operated in this jurisdiction.
  3. ASIC’s contention that the Buy Orders, the Amended Sell Orders and the Incremental Reductions were not placed with this genuine intention rests on an acceptance of the cumulative effect of its eight “pillars”. In oral closing submissions, ASIC argued that these “pillars” provided a “sufficient and compelling evidentiary basis” to draw the inferences it urged on the Court. I have found that:
(a) The first pillar (the Amended Sell Orders would have appeared to have been unprofitable at the time they were placed, assuming the market to have opened at those times) is of very limited assistance to ASIC’s case.
(b) The second pillar (the defendants could not have expected that sufficient SPI Futures could have been profitably traded by 10:08:45 to match the Buy Orders, the Amended Sell Orders and the Incremental Reductions) should not be accepted because it proceeds on a basis and model that I do not accept, and by methodology which I consider to be not sufficiently reliable, for the reasons I have given.
(c) In respect of the third pillar (Whitebox did not seriously pursue the volume of SPI Futures necessary to offset, by 10:08:45 the Buy Orders, the Amended Sell Orders and the Incremental Reductions), the analysis undertaken by Mr Morgan of Whitebox’s actual trading on the Serial Expiry Days in question shows that it was consistent with Whitebox seeking to match, by 10:08:45, the value of XJO Securities represented by the Subsequent Buy Orders (19 April 2012) and the final set of Reductions for each Rotation Group (other Serial Expiry Days) and that, with hindsight, Whitebox’s actual trading was undertaken at a somewhat linear rate. However, I am not persuaded that it necessarily follows that the defendants had in mind, from the start of trading on those days, a particular target (with respect to the volume of SPI Futures) to be achieved by 10:08:45.
(d) In respect of the fourth pillar (the Cancellations and Reductions were delayed until very shortly prior to the Rotation of the OSPA in which the relevant orders were due to trade so as to affect the price at which they would trade in the OSPA), Mr Graves’ rules of thumb are not reliable and no reliance should be placed on them for the purpose of determining the likely price effects of the Cancellation and the Reductions or to infer that the defendants’ purpose was to achieve the price effects that ASIC has alleged.
(e) In respect of the fifth pillar (the Cancellations and Reductions in respect of the last four Rotations were delayed until after the Group 1 Rotation and shortly prior to the rotation in the OSPA in which the relevant securities were due to trade so as to preserve order volume and thus affect the Opening Prices in the OSPA), there are aspects of Whitbox’s order activity that are consistent with ASIC’s case, but are ambiguous and do not establish it. I am not persuaded on the evidence that Whitebox’s position was necessarily “set” at the time of the Group 1 Rotation. Further, I am not persuaded on the evidence that by making the Cancellations and Reductions, at the time they were made, the defendants necessarily had an awareness that those Cancellations and Reductions were likely to affect the Opening Prices at which securities would trade in the OSPA and acted on that awareness, or that the defendants had an awareness that, at the time of their placement, the Buy Orders and the Amended Sell Orders would never trade, although the possibility that the Buy Orders and the Amended Sell Orders might be amended or cancelled, as events unfolded on the days in question, is likely to have been in their minds. In respect of the Incremental Reductions, I accept that it would have been in the defendants’ contemplation that there was a real possibility that each successive reduction (other than the last reduction made in each case) might not trade and the real likelihood was that a hedge would be achieved with respect to the futures leg which, by then, Whitebox had traded. However, this is not to say that the Incremental Reductions (other than the last Reduction made in each case) were placed to maintain order volume with the purpose of manipulating the Opening Prices of the relevant XJO Securities.
(f) In respect of the sixth pillar (Whitebox’s placement of large orders and subsequent reductions in order volume constituted a pattern capable of supporting an inference of a lack of intention to trade), a pattern of placing large orders for XJO Securities which were then cancelled or reduced shortly prior to the relevant Rotations in the OSPA is discernible on the Serial Expiry Days in question, and is relevant. But I am not persuaded that this pattern should bear the character, or have the significance, that ASIC attributes to it given my other findings. Further, the consistency and uniformity of the pattern is also discernible and relevant. Although capable of raising suspicions about Whitebox’s order activity, these matters would be of greater significance if other elements of ASIC’s case were established. As matters stand, I do not attribute to these matters the weight that ASIC asks me to place on them.
(g) In respect of the seventh pillar (an incentive to profit), I accept that it is likely that the defendants were incentivised to obtain the greatest profit possible through Whitebox’s trading, but this fact alone does not establish that the impugned orders were placed with the manipulative purpose that ASIC has alleged.
(h) In respect of the eighth pillar (Whitebox was keenly aware of the impact of its orders on the market), Whitebox’s ability to observe Underlying Basis gave it insight into the direction of the market and the flow of orders. But it did not give it the ability to generate mispricing or to know what would happen in the market if its orders were removed (amended or cancelled).
  1. In light of the evidence before me, and my evaluation of the cumulative effect of the findings I have summarised, I am not persuaded on the balance of probabilities that the Buy Orders, the Amended Sell Orders and the Incremental Reductions were not placed with any genuine intention that they should be traded but, rather, with the knowledge that, ultimately, they would be cancelled or substantially reduced in volume so as to manipulate the Opening Price for each of the XJO Securities in their respective Rotations in the OSPA.
  2. For this reason, the “market for” case has not been established, and fails.
  3. Given this conclusion, it is not necessary for me to address in any detail the defendants’ overall criticism of ASIC’s case theory: see at [503] – [510] above. However, on the evidence before me I accept that it would have been practically impossible for the defendants to predict, with any accuracy or certainty, the price impact that particular orders or amendments by Whitebox might have had on Opening Prices and that, when regard is had to the auction time lines for the Serial Expiry Days in question, there was considerable variation in both the size and timing of Whitebox’s order amendments which was consistent with Whitebox having responded to market conditions initially for Group 1 and then, in the absence of any favourable market movement, successively for the other Groups as the OSPA progressed.
  4. There is a further matter to which I should refer. The defendants submitted that the word “market” in the phrase “false or misleading appearance ... with respect to the market”, as used in s 1041B(1)(b), can only refer to the period of Normal Trading because only then can securities be bought and sold and the forces of actual supply and demand engage with each other. In other words, in the defendants’ submission, the Pre-Open Phase is not a “market”. Proceeding on this basis, the defendants submitted that ASIC must also establish a sufficient connection between the “appearance” likely to have been created in the Pre-Open Phase by Whitebox’s placement or amendment of the impugned orders and the “market” in Normal Trading which, at the relevant times, was not yet open. The defendants submitted that ASIC has not, and cannot, establish this connection because the “appearance” created by the impugned orders did not involve any promise or representation about the future.
  5. Given my conclusion that ASIC’s “market for” case has not been established, it is also not necessary for me to decide whether the Pre-Open Phase constitutes a “market” for the purposes of s 1041B(1)(b) of the Act. However, contrary to the defendants’ submission, I am inclined to the view, and if necessary would have decided, that it is. The “market” in s 1041B(1)(b) is a reference to a market for “financial products on a financial market operated in the jurisdiction”. The ASX is such a “market”. Although this “market” has phases, there is no reason to delineate or segment it so that the “market” is constituted by one phase but not another. A “market” is a broad notion that contemplates potential as well as actual transactions. It is appropriate, therefore, to the treat the “market” as including the Pre-Open Phase even though, in that phase, no order can execute until the Rotations in the OSPA for the securities concerned actually commence. In other words, the conditional nature of an order does not deny the existence of a “market” for the purposes of s 1041B(1)(b).

THE “PRICE FOR TRADING” CASE

Introduction

  1. Relevantly to this limb of ASIC’s case, s 1041B(1)(b) provides that a person must not do an act that has or is likely to have the effect of creating or causing the creation of a false or misleading appearance with respect to the price for trading in financial products on a financial market operated in this jurisdiction. The acts which ASIC relies on are the placement of the Buy Orders and Cancellations (19 April 2012) and the placement of the Amended Sell Orders and Reductions (other Serial Expiry Days). ASIC submitted that the likely effect of the Buy Orders and Cancellations on 19 April 2012 was to decrease the price of XJO Securities in the OSPA. It submitted that the likely effect of the Amended Sell Orders and Reductions on each of the other Serial Expiry days was to increase the price of XJO Securities in the OSPA.
  2. ASIC submitted that a false or misleading appearance as to the price for trading in those securities was created because the Buy Orders and the Amended Sell Orders did not reflect the forces of genuine supply and demand. Once again, ASIC relied on the notion of “genuine supply and demand” referred to in JM—namely, forces created in a market by buyers whose purpose is to acquire at the lowest available price and sellers whose purpose is to sell at the highest realisable price. On ASIC’s case, the defendants had no intention to trade the Buy Orders or the Amended Sell Orders. Rather, these orders, together with the Cancellations and Reductions, were placed for the purpose of manipulating the price for XJO Securities. This is the same manipulative purpose which underpins the “market for” case.
  3. ASIC accepted that the “price for trading” in financial products contemplates the price at which financial products can be bought and sold. Thus, on this limb of its case, s 1041B(1)(b) is directed to prices formed through the OSPA or in open or “live” trading. It is not directed to, and does not cover, the IMP for securities in the Pre-Open Phase.
  4. Therefore, the “price for trading” case implies a sufficient causal connection between the impugned orders and the likelihood that they affected the prices set for the XJO Securities in question. In other words, this limb of ASIC’s case is only made out if ASIC establishes not only that the defendants placed the Buy Orders and the Amended Sell Orders with the intention to manipulate the Opening Prices of the XJO Securities concerned, but also the fact that those orders, together with their related Cancellations or Reductions, were likely to have so affected the Opening Prices. I accept, of course, that likely effect can be inferred from the existence of a manipulative purpose such as alleged by ASIC in the present case: see JM at [72] – [73].
  5. ASIC submitted that, in this particular application, s 1041B(1)(b) extends beyond perpetrators who are parties to transactions that set the market price; it catches perpetrators who have affected, or who are likely to have affected, “the price for trading”.

Conclusion

  1. In considering this limb, I proceed on the same meaning of “likely” I have accepted above.
  2. As the first limb has failed, so too must this limb.
  3. This is because, firstly, on the evidence presented, I am not persuaded on the balance of probabilities that the Buy Orders and the Amended Sell Orders were not placed with any genuine intention that they should be traded but, rather, with the knowledge that, ultimately, they would be cancelled or substantially reduced in volume (as per the Cancellations and Reductions) so as to manipulate the Opening Price for each of the XJO Securities in their respective Rotations in the OSPA.
  4. In addition, ASIC has not established a sufficient causal connection between, on the one hand, the Buy Orders and the Cancellations, or the Amended Sell Orders and Reductions, and, on the other hand, the Opening Prices of the relevant XJO Securities other than, perhaps, in respect of the price spike in the Group 1 securities on 18 October 2012. Given my rejection of Mr Graves’ analyses and the other evidence relied on by ASIC in this regard, and absent proof on the balance of probabilities of the manipulative purpose which ASIC has alleged, I am not able to determine, as a matter of objective fact, what likely effect, if any, these orders had on the Opening Prices of the securities concerned other than in respect of Group 1 on 18 October 2012.

THE “ARTIFICIAL PRICE FOR TRADING” CASE

Introduction

  1. Relevantly, s 1041A(1)(c) of the Act provides that a person must not take part in, or carry out, a transaction that has or is likely to have the effect of creating an artificial price for trading in financial products on a financial market operated in this jurisdiction. The “transactions” on which ASIC relies are the placement of the Buy Orders and Cancellations (19 April 2012) and the placement of the Amended Sell Orders and Reductions (other Serial Expiry Days). These are the same orders, cancellations and reductions relied on in the “price for trading” case.
  2. Although s 1041A refers to “transactions” rather than “acts” (as used in s 1041B), ASIC submitted that there is no relevant distinction between the two terms. In fact, ASIC treated “transactions”, as used in s 1041A, as a hypernym that encompasses “acts”. In advancing this submission, ASIC called in aid:
(a) the breadth of the word “transaction” in its ordinary signification: see, in that connection, the observation of the New South Wales Court of Criminal Appeal in R v Manasseh and Austin [2002] NSWCCA 27 (Manasseh) at [57] that “transaction” is a comprehensive word which covers an act, doing, negotiation or dealing;
(b) the contention that the legislative purpose of s 1041A supports such a broad meaning; and
(c) the fact that United States case law in respect of s 9 of the Securities Exchange Act 1934 (US) has accepted that the placement of auction orders may constitute “transactions” that give rise to the manipulation of prices, regardless of whether those orders result in consummated transactions: In the matter of Kidder Peabody & Co 18 SEC 559 at 8; SEC v LEK Securities Corporation 276 F Supp 3d 49 (2017) at 62.
  1. As to the meaning of “artificial price”, ASIC submitted that the considerations which inform whether there is a false or misleading appearance as to price for the purposes of s 1041B(1)(b) of the Act also inform whether there is an “artificial price” for the purposes of s 1041A. ASIC submitted that, consistently with the submissions it advanced in respect of its “price for trading” case, an “artificial price” will be created where there is a price that has been influenced by factors other than the forces of “genuine supply and demand”. Seen in this light, ASIC’s “artificial price for trading” case is no different to its “price for trading” case. It is also implicit in ASIC’s formulation of the “artificial price for trading” case that the “price” is one formed through the OSPA or in open or “live” trading.
  2. As to the meaning of “likely”, ASIC repeated the submissions it made on the meaning of “likely” in the context of s 1041B(1)(b) of the Act.

Conclusion

  1. In considering this limb, I proceed on the same meaning of “likely” I have accepted above.
  2. The “artificial price for trading” case fails for the same reasons that the “price for trading” case fails. I am not persuaded on the balance of probabilities that, in placing the impugned orders, the defendants acted with the manipulative purpose that ASIC has alleged. Further, I am not satisfied that ASIC has established a sufficient causal connection between the impugned orders and the Opening Prices of the relevant XJO securities to show the likely effect alleged.
  3. Given these findings, it is not necessary for me to decide whether the placement of the Buy Orders and Cancellations (19 April 2012), and the placement of the Amended Sell Orders and Reductions (other Serial Expiry Days), constitute “transactions” for the purposes of s 1041A. Uninstructed by JM, I would have been inclined to the view that the placement of an order, or its cancellation or amendment, was a “transaction” within the meaning of the section. However, in JM, when discussing the forces of genuine supply and demand, the High Court distinguished between offers made (which would comprehend all bids and asks) and “transactions” that result from such offers. Thus, at [72], the High Court said:
The price that results from a transaction in which one party has the sole or dominant purpose of setting or maintaining the price at a particular level is not a price which reflects the forces of genuine supply and demand in an open, informed and efficient market. It is, within the meaning of s 1041A, an “artificial price”. The offer to supply or acquire of the kind described is made at a price which is determined by the offeror’s purpose of setting or maintaining the price. It is not determined by the offeror’s purpose, if buying, to minimise, or, if selling, to maximise, the price paid, and it is not determined by the competition between other buyers whose purpose is to minimise the price and other sellers whose purpose is to maximise the price. If the offer results in a transaction, that is a transaction which can be characterised as at least likely to have the effect of creating or maintaining an artificial price for trading in the shares.
  1. Later, at [74], the High Court said:
... Participants in the market can be (and are) informed of the transactions which occur. Participants in the market are entitled to assume that the transactions which are made are made between genuine buyers and sellers and are not made for the purpose of setting or maintaining a particular price...
  1. These passages suggest that, in the context of s 1041A, the word “transaction” is not a hypernym that encompasses unilateral acts, such as making an on-market offer (i.e., placing a bid or an ask) but, rather, acts that have been consummated by acceptance. If an on-market offer is not, relevantly, a “transaction”, there is no reason to think that the word has a different or extended statutory meaning in the context of the Pre-Open Phase to include an order placed, amended or cancelled. This suggests that the textual differences between s 1041B(1) (“act”) and s 1041A (“transaction”) are not accidental and that, contrary to ASIC’s submission, there is a relevant distinction between the two terms. Therefore, had it been necessary for me to decide the question, I would have felt constrained to conclude that the placement of the Buy Orders and Cancellations (19 April 2012), and the placement of the Amended Sell Orders and Reductions (other Serial Expiry Days), do not constitute “transactions” for the purposes of s 1041A.

BREACH OF DUTY CASE

  1. ASIC accepted that its breach of duty case under s 180 of the Act against Mr Boshoff could only succeed if its cases under s 1041B(1)(b) and s 1041A were established. As those cases have not been established, the case under s 180 fails.

DISPOSITION

  1. The Originating Process will be dismissed. Although in the normal course costs would follow the event, the defendants have nevertheless requested the opportunity to be heard on that question. Unless the question of costs can be agreed, the defendants are to file and serve written submissions on that question by 4.00 pm on 19 June 2019. ASIC is to file and serve written submissions in response by 4.00 pm on 28 June 2019. In each case, submissions are to be limited to three pages. If it is necessary for the defendants to reply, the submissions are to be filed and served by 4.00 pm on 5 July 2019, limited to two pages. I will then determine the question of costs on the papers, unless a party wishes to be heard orally.
I certify that the preceding five hundred and sixty eight (568) numbered paragraphs are a true copy of the Reasons for Judgment herein of the Honourable Justice Yates.



Associate:

Dated: 7 June 2019

SCHEDULE

Introduction
1. This primer includes the following:
(a) a glossary of terms;
(b) a basic description of index arbitrage;
(c) basic commentary concerning the ASX and ASX24.
Glossary of Terms
  1. Attached is a glossary of terms likely to be used in the proceedings.
Basic description of index arbitrage
  1. Arbitrage is the practice of buying an asset in one market while selling it in another market in order to exploit, for profit, any difference in price. Index arbitrage (expanded on further below) is a form of arbitrage and is a strategy that seeks to profit from the difference between actual and theoretical prices of a stock market index (“securities index”) and related futures contract.1 It involves the practice of buying (or selling) the constituent parts (or some of the constituent parts as a proxy portfolio) of a securities index and selling (or buying) futures contracts which are structured to follow or track the level of the relevant securities index. The objective is to try to profit from the divergence between the trading price and the estimated “Fair Price” (explained below) of a futures contract associated with or derived from a securities index.

_______________________________

1 A futures contract is a standardised agreement for the purchase or sale of a commodity or financial product (such as a security) on a given date in the future.

Securities index/market index
  1. A securities index (or market index) is a number, typically published by a stock exchange, which represents the aggregated weighted price level of securities (otherwise described as “stocks”, “shares” or “equities”) that are constituent members of the index. The index is used as a measure to track changes in the aggregate value of the underlying securities over time.
  2. In Australia, there are a number of market indices, including the S&P/ASX 200 index (“ASX 200” or “XJO”, which is the ASX unique code for that index).

The S&P/ASX 200 index (aka the XJO)
  1. The constituent securities of the XJO are securities drawn from approximately 200 (but not always 200) eligible companies listed on the ASX. This index is administered by Standard & Poor’s.
  2. The XJO is market capitalisation-weighted, meaning the company’s contribution to the index is relative to its total market value, i.e., its share price multiplied by the number of its tradable shares. The tradable value of the shares is assessed by excluding from the calculation of market capitalisation any strategic holdings (which are classified as either private, corporate or government holdings).
  3. The XJO is also subject to the application of a price divisor to have it reflect changes in share price levels rather than changes in market capitalisation arising from corporate actions such as share issues.
  4. The S&P index committee has sole responsibility for determining the methodology, maintenance, weighting, selection of constituent securities and index procedures of the XJO.
  5. The constituent securities and their weighting in the index is generally rebalanced every quarter and the changes take effect after close of trading on the third Friday of March, June, September and December. Non-quarterly changes can also occur, including both as to the weighting and the constituent securities.
  6. The constituent securities have different weightings in the index. Generally, the top 30 or so constituent securities make up over 70% of the weighting of the index.
  7. The level of the XJO is recalculated by S&P during Normal Trading on the ASX (i.e., during the period that the ASX is open for trading) and is published periodically throughout Normal Trading by the ASX under the code “XJO”.
  8. The ASX publishes the changes in the level of the XJO every 30 seconds via a broadcast message from ASX Trade, known as a "BD2" broadcast message.
  9. The ASX does not publish the XJO level at times outside Normal Trading.
  10. Some participants2 have proprietary systems (or subscribe to industry services that have proprietary systems), that undertake their own calculation of the XJO levels more frequently than is published by the ASX. Prior to trading of securities on the ASX, calculations of “indicative XJO” (a reference to the indicative level of the XJO) can also be made by an individual participant using “indicative” prices of securities periodically published by ASX Limited or calculated by the participant itself. This is explained further below.

_________________________________

2 Where the expression “participant” is used in this Primer, it is a shorthand reference to a person or entity, seeking to place orders or trade on the ASX or ASX24, and is not intended to be used in the technical sense of “Market Participant” – unless specifically stated.

ASX

The equities market

  1. The ASX operates on an electronic trading platform which, since 29 November 2010, has been known as “ASX Trade”. It can be accessed by “trading participants” (also commonly known as stockbrokers or brokers) (“Market Participants”).
  2. ASX Trade is designed to receive messages from ASX trading participants, including buy orders for securities (also known as “bids”) and sell orders for securities (also known as “asks” or “offers”). It compiles an electronic list of bids or asks of securities to build an order book for each security (“Order Book” or “Central Order Book”).
  3. Trade is a priority-based system in which orders for a particular security in the order book are sorted and displayed in price/time priority order, provided for under the ASX Operating Rules.
  4. Bids (Buy Orders) or Asks (Sell Orders or Offers) can be placed in the order book as follows:
• “Enter” being a new order;
• “Delete”, or “Cancel”, removing an order from the order book.
  1. Most orders on ASX Trade are “Limit Orders”, with a maximum chosen price for a bid, or a minimum chosen price, for an ask. ASX Trade does not facilitate “Market Orders”, i.e., an order to be traded at whatever is the prevailing market price, which are found on some other international exchanges.
  2. Trades occur when bids and asks are matched within ASX Trade, and are subsequently reported by the ASX.
ASX phases
  1. The ASX operates in a series of phases. A general listing and description of the ASX phases likely to be referred to in the proceedings is set out below.
7am - 10am* Pre-open Phase
10am* Commencement of Open Phase
10am* - 4pm Normal Trading
4pm - 4:10pm Pre CSPA
4:10pm - 4:11/124pm4 Closing Single Price Auction
* random +/- 15 secs and staggered opening
  1. A brief description of the phases is as follows.

___________________________

3 The Defendants will be referring to ASX Operating Rules [4030] and [4032].

4 There are competing ASX publications as to whether the closing time of the CSPA is at 4:11 or 4:12. For the purposes of these proceedings, those differences are of no significance.


The Pre-opening Phase (or Pre-Open)
  1. The Pre-Open takes place from 7:00 am to approximately 10:00 am, Sydney time. During the Pre-Open:
• ASX Trade does not trade orders in the Pre-Open Phase.

Open Phase (or Opening)
  1. Market opening commences at approx.10:00 am Sydney time. It is a staggered opening lasting for approximately 9 minutes5 in which groups of securities (grouped alphabetically by their ticker code) start trading, with an auction, (being the opening price auction or Opening Single Price Auction or “OSPA”), at five intervals known as rotations (one for each group). In the Opening Phase (i.e. at the opening of each rotation or group), ASX Trade calculates the opening prices (or as they are sometimes called the “auction prices” or the “match prices”) at which the securities open for trading by way of the OSPA (the calculation is explained further below). The five groups and the times at which the opening auction for those groups takes place are as follows:
Group/rotation 1 10:00:00am +/-15secs for A-B securities, e.g. ANZ, BHP
Group/rotation 2 10:02:15am +/-15secs for C-F securities, e.g. CPU, FXJ
Group/rotation 3 10:04:30am +/-15secs for G-M securities, e.g. GPT
Group/rotation 4 10:06:45am +/-15secs for N-R securities, e.g. QAN
Group/rotation 5 10:09:00am +/-15secs for S-Z securities e.g. TLS
  1. The opening time of each group is randomly generated by ASX Trade and occurs up to 15 seconds on either side of the times given above, e.g. Group 1 may open at any time between 9:59:45 am and 10:00:15 am.

____________________

5 The staggered opening time is 16 minutes on Quarterly Expiry Days.

Normal Trading
  1. Normal Trading takes place from approximately 10:00 am (noting the random +/-15 seconds and the staggered opening) to 4:00 pm, Sydney time. Market Participants enter orders into ASX Trade which are matched against each other in price/time priority order on a continuous basis. Trading of securities occurs in this phase.

Pre CSPA
  1. Between 4:00 pm and 4:10 pm, Sydney time, the market is placed in Pre-CSPA. Trading stops and Market Participants enter, amend and cancel orders in preparation for the market closing.

Closing Single Price Auction (CSPA)
  1. The Closing Single Price Auction takes place at a random time between 4:10 pm and 4:11 pm, Sydney time.
  2. ASX Trade calculates closing prices during this phase.

Further description of the Opening Phase and the calculation of “Opening Prices”
  1. As stated above, at approximately 10:00am, the market commences opening for “Normal Trading”, but on a staggered basis.
  2. Because orders can be entered, amended or cancelled during the Pre-Open Phase (i.e. before market opening) but trades do not take place, orders may “overlap”. This means bids may be sitting in the order book at a higher price than the lowest asks sitting in the order book. During the Pre-open Phase “indicative” opening prices may be calculated at any moment in time by applying the ASX Algorithm (explained below) based on the orders sitting in the order book at that moment. These calculations are hypothetical, as trades cannot take place at this time.
  3. At the end of the Pre-open Phase, upon the Opening of each Group, ASX Trade uses the ASX Algorithm to set the Opening (or auction or match) Price for each listed security available for trading (i.e. not suspended).

OSPA
  1. The “Opening Single Price Auction” (“OSPA”) establishes the Opening Price for each security by operation of the ASX Algorithm (described below).
  2. There are five opening single price auctions, one for each Group as it opens. During the Opening Phase the ASX Algorithm (or, as it is sometimes called, “the ASX opening algorithm” or the “auction algorithm”) sets the Opening Price for each listed security. All bids in the Order Book at the Opening that are at or above that opening price will trade and similarly all asks in the Order Book at Opening that are at or below that opening price will trade, except that, if there is an excess of bid volume over ask volume, or vice versa, at the Opening Price, the excess volume (determined by time priority) will not trade.
  3. The OSPA is sometimes also used to describe the “Open Phase”, this being the period over which the five Groups have staggered openings. However, during the Open Phase some securities will already be open and trading in Normal Trading. For example, securities in Groups 1 – 4 are already trading in Normal Trading prior to the opening of Group 5 (S-Z) at 10:09am +/- 15 seconds.

Opening Prices
  1. The ASX employs the ASX Algorithm to determine the Opening Prices for all securities listed on the ASX (if the securities are not otherwise suspended from trading at that time). The same ASX Algorithm is also deployed to determine closing prices in the CSPA phase.

Explanation of the ASX Algorithm
  1. The ASX Algorithm is a four-step approach to determine the opening price for each security as it opens for trading. The operation of and use of the ASX Algorithm is explained in Chapter 9 of the ASX Trade Workstation Manual.6 In summary terms, the four-step algorithm works as follows.
  2. Based on the orders in the Order Book at the time, the ASX algorithm determines the Opening Price as follows:
(d) The First Step is based on the principle that the price should be the one at which the maximum volume of a security would be traded.
By matching bids to asks, the algorithm identifies what volume of securities will trade at particular prices and will identify at what price step the maximum executable volume of trades in the security will occur (by comparing the volume of matching bids and asks at each price step). If the first step of the algorithm determines that the maximum executable volume of trades could be executed at more than one price, then the algorithm moves to the second step to determine which of those price steps should be the Opening Price.
(e) The Second Step is based on the principle that the price should be the one that leaves the least quantity of shares in unfilled orders.
The second step ascertains the volume of unfilled orders at the eligible Opening Prices determined under step 1 and will determine which of those prices leaves the least quantity of shares in unfilled orders. If the same quantity of shares in unfilled orders would arise at more than one of those prices, then the algorithm moves to the third step to determine the Opening Price.

(f) The Third Step is based on the principle that the highest potential price should be used if pressure is only on the buy side (i.e. there is a greater volume in bids than asks) or the lowest if the pressure is only on the sell side (i.e. there is a greater volume in asks than bids).

The third step identifies where the pressure is, either on the buy side or the sell side. If it is only on the buy side, the Opening Price will be the highest of the eligible Opening Prices identified under the second step and, if it is only on the sell side, it will be the lowest of the prices identified in the second step. If the pressure is on both the buy and sell side, then the algorithm moves to the fourth step to determine the Opening Price.

(g) The Fourth Step is based on the principle that the price should be set with reference to the last traded price.

If an Opening Price has not been determined under the first three steps of the algorithm, an Opening Price is then set by reference to the last traded price of the security, which is most likely to be the security’s last closing price.
The ASX Algorithm determines the closing price in the CSPA in the same way.
_____________________________
6 This chapter is re-produced at Appendix E of Mr Graves’ Report and Schedule E of Professor Aitken’s Report.

“Indicative” opening prices
  1. Prior to Normal Trading, there are no traded or market prices for securities on the ASX.
  2. As orders can be entered, amended and cancelled in the order book by Market Participants in the Pre-open Phase, the ASX uses the ASX Algorithm during the Pre-open Phase, to calculate at each time a new order, amendment or cancellation is placed in the order book a hypothetical or “Indicative” opening/auction/match price by reference to the orders in the order book at the time of that calculation.
  3. These “indicative” prices are calculated in real time as each order is placed, amended or cancelled from the order book. These “indicative” prices are published in the Pre-open Phase by the ASX through its BO10 message or broadcast. The broadcast provides information (amongst other information) on changes in the Indicative prices. The BO10 message is subject to a 500 millisecond “holdback” period7 which means it may include the cumulative effect of several orders occurring within that half second period.
  4. The “indicative” prices are not prices at which any security can trade as, at the time they are calculated and published, the ASX is not open for any trading in those securities.
  5. As noted earlier, some participants employ proprietary systems (or subscribe to industry services that provide proprietary systems) that calculate “indicative” opening prices during the Pre-open Phase more frequently then every half second. Some participants will also use these proprietary systems to then calculate from those “indicative” opening prices a calculation of an “indicative” XJO. The ASX does not publish any XJO levels until the commencement of Normal Trading.
  6. It is the ability of participants to calculate an “indicative” XJO level from their own systems or subscription services that may allow them to consider potential arbitrage opportunities prior to the ASX opening for Normal Trading. This is explained further below.
_________________________
7 There appear to be instances where it may have been published earlier than a 500 millisecond period.

ASX24
The futures market
  1. ASX24 (formerly known as the Sydney Futures Exchange) is a market for the trading of derivative financial products (approved by its operator, Australian Securities Exchange Limited), including futures and options contracts.
  2. A futures contract is a standardised agreement for the purchase or sale of a commodity or financial instrument (such as a security) on a given date in the future (“Expiry Date”) at a price fixed at the time the contract is made (“Contract Value”).
  3. The SPI future is a “cash-settled futures contract” which requires the seller or buyer to pay to its counterparty, on the Expiry Date, the difference between the Contract Value and the value of the contracted commodity or financial instrument on the Expiry Date (“Settlement Value”).

SPI Futures
  1. A common cash-settled futures contract traded on ASX24 is the ASX SPI 200 Index Futures Contract (“SPI Future” or “SPI”). Some of the popularity of the SPI Future arises from its offering a simple and economical mechanism to buy or sell exposure to the XJO, instead of trading all 200 stocks that make up the XJO individually on ASX Trade. The underlying value of a SPI Future is derived from the level of the XJO.
  2. The SPI Futures contracts initially written were quarterly contracts with Expiry Dates in March, June, September and December. Since 2010, monthly serial SPI Futures contracts have also been listed which expire on the third Thursday in January, February, April, May, July, August, October and November.
  3. Accordingly, there are two types of SPI Futures contracts:
(h) “Quarterly SPI Futures” that expire on the third Thursday of March, June, September or December; and
(i) “Serial SPI Futures” that expire on the third Thursday of the intervening months, i.e., January, February, April, May, July, August, October or November.
  1. The SPI Futures contracts have a listing code “AP” with coding for the year and month of expiry. Attached is a schedule that sets out the Expiry Date and type (serial or quarterly) of each set of SPI Futures that expired in 2012, in or after April 2012, with the ASX24 and other industry provider codes that refer to those contracts.
  2. The contract specifications of the SPI Futures contract are extracted below:
Contract Specifications of the ASX SPI 200 Futures Contract
ASX SPI 200®1 Index Futures
Contract Unit:
Valued at A$25 per index point (e.g. A$150,000 at 6,000 index points).
Contract Months:
March / June / September / December up to six quarter months ahead and the nearest two non-quarterly expiry months.
Commodity Code:
AP
Listing Date:
02/05/2000
Minimum Price Movement:
One index point (A$25)
Last Trading Day:
All trading in expiring contracts ceases at 12.00pm on the Third Thursday of the settlement month. Non-expiring contracts will continue to trade as per the stated trading hours.3
Cash Settlement Price:
The Special Opening Quotation of the underlying S&P/ASX 2002 Index on the Last Trading Day. The Special Opening Quotation is calculated using the first traded price of each component stock in the S&P/ASX 2002 Index on the Last Trading Day, irrespective of when those stocks first trade in the ASX trading day. This means that the first traded price of each component stock may occur at any time between ASX market open and ASX market close (including the Closing Single Price Auction) on the Last Trading Day.
Should any component stock not have traded by ASX market close on the Last Trading Day, the last traded price of that stock will be used to calculate the Special Opening Quotation.
Trading Hours:
5.10pm – 7.00am and 9.50am – 4.30pm3 (For period from second Sunday in March to first Sunday in November)
5.10pm – 8.00am and 9.50am – 4.30pm3 (For period from first Sunday in November to second Sunday in March)
Settlement Day:
The first business day after expiry, ASX Clear (Futures) publishes the final settlement price of the contract. On the second business day after expiry, ASX Clear (Futures) settles cash flows as a result of the settlement price.


1 ASX SPI 200® is a trademark of the Australian Securities Exchange.

2 "S&P/ASX 200" is a trademark of Standard & Poor’s. The trademark is used under licence by the Australian Securities Exchange.
3 Trading hours: Australian Eastern Standard Time / Australian Eastern Daylight Time.

Source: http://www.asx.com.au/documents/products/asx24-contract-specifications.pdf

  1. In summary:
(j) The price of a SPI Future is quoted in whole XJO index points. Each index point is equal to $25.00. The Contract Value of a SPI Future is therefore calculated as its price in index points multiplied by $25. Thus, for example, if the XJO level was 5000 points then each SPI future would have a Contract Value of $125,000 ($25 x 5000).
(k) Trading in a SPI Future ceases at noon (12:00:00) on its Expiry Date, and cash settlement occurs on the second business day after the Expiry Date.
(l) The Settlement Value of a SPI Future is equivalent to the value of the prices of the XJO securities as they first traded on the relevant Expiry Date. If they traded at market opening it will be their Opening Price. For any security that was not traded at the market opening, it will be the first traded price of the security on that Expiry Date, or if the security did not trade on the Expiry Date, it will be its last traded price.
Trading platform – ASX Trade24
  1. Trading of SPI Futures is conducted on ASX24’s electronic trading platform known as “ASX Trade24”. Like the ASX, the trading can only be undertaken by a Market Participant.

Price/time priority
  1. Like ASX Trade, ASX Trade24 is a priority-based trading platform in which orders are sorted in price/time priority. Thus:
(a) a bid with a higher order price is traded before a bid with a lower price. Conversely, an ask with a lower order price is traded before an ask with a higher price; and
(b) any bids or asks at the same price as other bids or asks respectively are traded in the order in which they were entered in ASX Trade24, so that those entered earlier in time are matched and traded before those entered later in time.
  1. All bids and asks for SPI Futures are entered on ASX Trade24’s electronic order book for SPI Futures (“SPI Order Book”).
  2. Market Participants trading futures can at any point in time only see the Bids and Asks in the SPI Order Book at the 5 highest Price Steps for the Bids and 5 lowest Price Steps for the Asks.

Futures trading phases
  1. There are two daily normal trading sessions for the trading of SPI on the ASX24, a “night session” followed by a “day session”. The night session is linked to US daylight saving time. In summary, the details of the sessions are as follows:
• 5.10pm to 7.00am and 9.50am to 4.30pm (during US daylight saving time)
• 5.10pm to 8.00am and 9.50am to 4.30pm (during US non-daylight saving time)
the times are Sydney times. US daylight saving begins first Sunday in April and ends last Sunday in October.
Before the opening of each session there is pre-open phase, lasting approximately 10 minutes, which is immediately followed by an opening auction at 5:09:30pm (prior to the commencement of the night session) and 9:49:30 am (prior to the commencement of the day session).
  1. Prior to the ASX24 the trading platform for the trading of futures including the SPI was called “SYCOM”. Accordingly, the industry sometimes refers to the closing price at the end of the morning session (7:00am/8:00am) for a SPI Futures contract as its “last SYCOM price”.

Index arbitrage
Overview
  1. Securities index arbitrage is a trading strategy by which the securities (or a proxy portfolio of securities) comprising an index, are traded against futures (or similar derivative products) that derive their value from that index.
  2. Index arbitrage trading in Australia which involves the XJO, involves buying (selling) securities (or a proxy portfolio of securities) comprising the XJO, and selling (buying) SPI Futures, with a view to profiting from any price disparity from “Fair Price” of the SPI (explained further below). The selling elements and buying elements of establishing an arbitrage position are respectively known as the two “legs” of an arbitrage position.
  3. The two legs of an index arbitrage are sometimes called “Cash” when referring to the securities side and “futures” when referring to the futures side, i.e. “buying Cash selling futures”, “selling Cash buying futures”.

Fair Value, Fair Price and Mispricing
  1. To assess whether there are arbitrage opportunities, an arbitrageur will calculate “Fair Value” of the SPI. This is generally the value as calculated by the particular arbitrageur (measured in index points) of the funding costs (and possibly stock borrow costs) and estimated dividend flows between the trade settlement date of the Cash and the futures Expiry Date. This “Fair Value” is added to the level of the XJO to calculate the “Fair Price” for the SPI. Each arbitrageur is subject to varying costs of funding and stock borrow costs, and will calculate their own estimates of expected dividends flowing from the underlying stocks to the holder of the stocks. This means every arbitrageur will have their own calculation of Fair Value, and hence their own calculation of the Fair Price of the SPI.
  2. Mispricing is the measure of any divergence between the calculated Fair Price of the SPI and the market price of the SPI. It is described in “basis points” (bps), each being 1/100th of one percent. So, for example, if the SPI is trading at a level one percent above Fair Price, this is described as 100 bps of mispricing. This mispricing is known as “Basis”.
  3. The glossary of terms further expands on the definitions of “Fair Value”, “Fair Price” and “Basis”.
  4. Arbitrageurs that identify mispricing, or Basis, are incentivised to capture the available potential profit by executing trades between the two markets, i.e. buying (selling) the SPI Future and selling (buying) individual stocks that comprise (in whole or sometimes in part as a proxy portfolio) the XJO in their appropriate quantities/weightings. The level of mispricing required will vary between arbitrageurs as it will reflect their different Fair Price calculations and also the transaction costs involved in executing the trades.

Index arbitrage activity
  1. If there is mispricing when both the ASX and Futures markets are open for trading arbitrageurs generally will seek to profit from that mispricing by executing trades as close as simultaneously as possible on both exchanges, and to repeat that process throughout intra-day trading. This type of index arbitrage activity is sometimes described as “vanilla” index arbitrage. Many arbitrageurs have automated systems or deploy algorithms to trade on both legs during intraday trading as an index arbitrage strategy.
  2. Index arbitrage opportunities, however, can also exist at times when both markets are not open, as can be the case when ASX Trade is in the Pre-Open Phase (i.e. not open for Normal Trading) but ASX 24 has commenced trading SPI Futures in the morning session from approx. 9:50 am. These opportunities arise because the arbitrageur may be able to make calculations or observe calculations of the indicative opening prices for each security during the Pre-open Phase between 7am and the Opening Phase. The arbitrageur then may be able to calculate a theoretical opening XJO index level and compare that calculation to the last trade of the overnight SPI Futures trading session (the “last SYCOM price”). If the disparity is larger than Fair Value, this would suggest that there exists a potentially profitable index arbitrage opportunity. The arbitrageur may then enter orders for “baskets” of XJO securities into the ASX Trade order book, anticipating that the SPI futures leg of the arbitrage can be executed at a time after 9:50 am, should the potential profit (or Basis) remain apparent at that point. Such activity involves a timing mismatch between executing the futures and executing the Cash legs. As such, unhedged risk is necessarily accepted in this style of arbitrage.
  3. The likelihood of a potentially profitable arbitrage existing (i.e. mispricing existing) tends to be higher on futures contract expiry days each month. This is because index arbitrageurs (and other market participants) who have held an arbitrage position to expiry will potentially be closing out their securities (Cash) positions at expiry. As the expiring SPI Futures contract is settled by reference to the first traded price of the XJO securities on the expiry day (which generally will be for most securities its Opening Price), those arbitrageurs are generally agnostic as to the Opening Price for each individual security. To explain, any “loss” experienced by the arbitrageur in buying or selling the component stocks at the Opening Price is made good by the concomitant adjustment to the expiring futures’ settlement price. Such arbitrageurs are only likely to be concerned that their Cash position is able to be fully traded at the Opening prices. As the ASX does not allow “Market Orders”, arbitrageurs who are closing out their Cash positions on expiry will therefore, to ensure execution, tend to place their orders in the Pre-open Phase in what is described as “deep in the book”. That is, an arbitrageur selling stock will place a limit order at a price considered to be well below the likely opening price for the stock at Opening; conversely, if he or she is buying stock, he or she will place a limit order at a price assessed to be well above the likely opening price of the stock at Opening.
  4. The extent of such activity on a futures expiry day can, amongst other causes, lead to there being (or expected to be) mispricing at or around the time of ASX Opening. That mispricing (or expectation of mispricing) allows other index arbitrageurs to seek to profit from it by seeking to establish Index arbitrage positions against the next quarterly expiring futures contract.
  5. An arbitrageur with the IT resources to conduct calculations of “indicative” XJO prior to ASX opening is in a position to assess whether (and to what potential extent) index arbitrage opportunities may exist at or around the time of ASX Opening and is better placed to then make decisions to enter, amend or cancel orders in the order book during the Pre-open Phase, placing him into a position to profit from any mispricing that may persist at or around the ASX Opening.
  6. Because of the nature of the Pre-open Phase, assessments of mispricing will change throughout the morning as more order activity comes into the order book of ASX Trade. Arbitrageurs can monitor, calculate and re-calculate Basis, i.e. the mispricing, throughout the Pre-open Phase. Some are also able to calculate hypothetical Basis absent their own orders. Such calculations are sometimes known as calculations of “Underlying Basis”.

Management of an Index Arbitrage position
  1. An arbitrageur who has bought securities and sold futures is said to be “long stock / short futures” or to hold a “positive arbitrage” position. An arbitrageur who has sold securities and bought stock is said to be “short stock / long futures” or to hold a “negative arbitrage” position. In either case, there are essentially three ways for the arbitrage position to evolve, as follows:
(a) The arbitrageur may await a favourable move in Basis so that the stock and futures legs can both be reversed (unwound) in the market, thus capturing part or all of the anticipated profit, or
(b) The arbitrageur may ‘roll’ the arbitrage by buying back (selling back) the futures leg prior to expiry and selling (buying) futures that expire in a subsequent month, all the while retaining their position in the stocks underlying the index, or
(c) The arbitrageur may hold the arbitrage to the futures expiry day, whereupon the stocks are likely to be bought (sold) at Opening Prices, and the price realised will be reflected in the held expiring futures position, which is subsequently cash settled.
  1. A wide and varied group of participants engage in Index Arbitrage in various forms and can apply differing trading practices, depending on their trading mandate and risk appetite and perception of risk.

Risks
  1. Depending on how and when Index Arbitrage is undertaken, there are a variety of risks associated with it. These can include:

(a) interest rate risk (i.e., in very simple terms, the risk that interest rates that transpire will differ from the assumed interest rates used to calculate the Fair Value of futures);

(b) unexpected changes in dividend size or timing;

(c) risk of index constituent stocks being suspended from trade;

(d) stock borrow risk i.e. that stock previously borrowed to settle short sales is unexpectedly recalled by the lender;

(e) Unhedged exposure risk (i.e. in very simple terms, the futures position and Cash position are not fully hedged so they may be exposed to price movements).

  1. Different participants may have different approaches and adopt different strategies to identify, manage and mitigate risks. Among other things, this may be influenced by what risk the arbitrageur has a mandate to accept and their own appetite for risk.

Glossary of Terms
• “Ask” (also see “Offer”) is a sell order for (relevantly) securities or futures.
• “Auction Price” – see “Match Price”.
BASIS=
( Futures Price-Index Level-Fair Value ) X 10,000
Index Level
And if an arbitrageur is using a Basis Mark, then the Fair Value figure may be adjusted by the Basis Mark.
_____________________
8 “Index Level” in the formula is what an index arbitrageur calculates as the relevant Index level (XJO level or indicative XJO level) depending on the methodology employed for the calculation and also the purpose or use of the calculation.
• “Bid” is a buy order for (relevantly) securities or futures.
Position
Long Stock
Short Stock
Funding Cost
Have to Pay Cost of Funding Position
Receive Interest on Proceeds of Sale, less stock borrow cost
Dividends
Receive Dividend from holding stock
Have to pay Dividends to party that lent me
the stock scrip to short
Long Fair Value = funding costs – dividends received
Short Fair Value = interest received – stock borrowing cost – dividends paid
Long Fair Value = funding costs — dividends received
• “Negative Index Arbitrage” is a reference to buying futures and selling Cash.
• “Normal Trading” – the Open phase starts immediately after the Pre-open Phase and lasts until the start of the Pre-CSPA at 4pm. The Open phase commences with the OSPA in the first rotation. During the Open phase (also known as Normal Trading) orders may be entered, amended, cancelled and traded. During Normal Trading, ASX Trade matches bids and asks automatically and immediately whenever the price for a bid and an ask for a security “overlap”, that is, when the price of the bid is equal to or greater than the price of the ask.
• “Offer” – see “Ask”.
• “Open Interest” refers to open positions in a SPI Future.
• “Opening Price” – see “Match Price.
Rotation
Range of possible rotation
times
• A-B securities (first
rotation):
• Between 09:59:45am and
10:00:15am
• C-F securities (second
rotation):
• Between 10:02:00am and
10:02:30am
• G-M securities (third
rotation):
• Between 10:04:15am and
10:04:45am
• N-R securities (fourth
rotation):
• Between 10:06:30am and
10:07:00am
• S-Z securities (fifth
rotation):
• Between 10:08:45am and
10:09:15am
Short Fair Value = interest received – stock borrowing cost – dividend paid
• “Tick Size” – see “Price Step”.

___________________________
9 There is a dispute between the parties as to whether the term “SPI Equivalent”, in the manner as used by the Plaintiff, is an industry recognised term.

Schedule
____________________________________________________________________
SPI Futures
contract
Contract
type
Expiry Date
ASX 24 coding
Bloomberg
coding
TRTH coding
April 2012 SPI Futures

Serial

19 April 2012

APF2

XPJ12

YAPJ2
May 2012 SPI Futures

Serial

17 May 2012

APK2

XPK12

YAPK2
June 2012 SPI Futures

Quarterly

21 June 2012

APM2

XPM12

YAPM2
July 2012 SPI Futures

Serial

19 July 2012

APN2

XPN12

YAPN2
August 2012 SPI Futures

Serial

16 August 2012

APQ2

XPQ12

YAPQ2
September 2012 SPI Futures

Quarterly
20 September 2012

APU2

XPU12

YAPU2
October 2012 SPI Futures

Serial

18 October 2012

APV2

XPV12

YAPV2
______________________________________________________________________


AustLII: Copyright Policy | Disclaimers | Privacy Policy | Feedback
URL: http://www.austlii.edu.au/au/cases/cth/FCA/2019/849.html