NZLII Home | Databases | WorldLII | Search | Feedback

Supreme Court of New Zealand
You are here:  NZLII >> Databases >> Supreme Court of New Zealand >> 2015 >> [2015] NZSC 109

Database Search | Name Search | Recent Decisions | Noteup | LawCite | Download | Help

Graham, Reeves, Jeffries and Bryant v The Queen [2015] NZSC 109 (22 July 2015)

Last Updated: 22 July 2015

     
IN THE SUPREME COURT OF NEW ZEALAND

BETWEEN
 Applicants
AND
 Respondent
Court:
Elias CJ, William Young and Glazebrook JJ
Counsel:
D H O’Leary for Applicants C R Carruthers QC and D R La Hood for Respondent
Judgment:
22 July 2015


JUDGMENT OF THE COURT

  1. The application for recall is dismissed.

  1. Costs are reserved.

____________________________________________________________________

REASONS

A recall application

[1] The applicants seek a recall of the judgment of this Court delivered on 25 October 2013[1] refusing them leave to appeal against their convictions.[2] In the same judgment, the applicants were granted leave to appeal against sentence. Their sentence appeal was subsequently heard and was later allowed in a judgment delivered on 7 May 2014.[3]
[2] The recall application is premised on the assertion that the Court would not have declined the applications for leave to appeal against conviction had it known that: (1) the trial Judge founded the convictions on evidence that had not formally been proved (namely the Peden schedule); (2) defence counsel never entered into an agreement under s 9 Evidence Act 2006 to admit the schedule; (3) the ruling of the trial Judge was that it was for the Crown to apply to recall Ms Peden to prove the schedule if it wished to put it to further witnesses; and (4) appeal counsel did continue to press in the Court of Appeal hearing that there had been procedural flaws at trial, and that the Court of Appeal judgment had misrecorded this. The submissions in support of the recall application advance further complaints, namely that the Peden schedule involved a deviation from the Crown case as particularised and was wrong and lacking in logic as involving double counting.

The Peden schedule

[3] The Peden schedule consists of a number of spreadsheets, the most material of which is in two parts, one headed “FORE CAST LOAN REPAYMENTS & DRAWDOWNS [Board papers]” and the other “COMPARISON OF FORECAST LOAN CASHFLOWS AGAINST ACTUAL”. It was prepared by Michelle Peden, who was a Crown expert witness. It was not, however, produced by her in evidence. Rather it was prepared, it seems, after the Crown case closed and was made available to the prosecutor, Mr Carruthers QC, for use in his closing address. The spreadsheet is attached to this judgment as appendix 1. We have labelled various rows for ease of reference.
[4] Because the period with which we are concerned extended only until December 2007, our analysis focuses on the first five columns of the spreadsheet. We have, however, included the spreadsheet in full. We note that Lombard engaged Ferrier Hodgson (who were subsequently rebranded as KordaMentha) to review their books and liquidity and, as is indicated by the boxed text between the two parts of the spreadsheet, Ferrier Hodgson’s report was based on the same forecast figures as those which were before the Board at the 26 September 2007 meeting.[4]
[5] The 26 September forecast is represented by what we have labelled row B. As the row B figures record, the forecast was for loan repayments to be made in the months of September, October, November and December 2007 totalling $63.211m (this being the sum of $14.251m, $18.765m, $16.875m and $13.320m).[5]
[6] The actual repayments are set out in row F and amount to $33.033m (this being the sum of $1.978m, $6.830m, $13.655m and $10.570m), representing 52.26% of the Ferrier Hodgson forecasts.
[7] Rows F and G compare the actual repayments in each of the months of September, October, November and December 2007 to the repayments as last forecast for those months (“the most current forecasts”). For September and October these are the forecasts contained in the Ferrier Hodgson report. For November and December there were updated forecasts which appear in rows C and E. The recoveries for each of these months as against the most current forecasts were 13.9% for September, 36.4% for October, 69.7% for November and 55.0% for December. The cumulative total of the most current forecasts for those months was $71.822m.
[8] The calculations set out in [6]–[7] are very simple and of a kind which counsel and the trial Judge were perfectly capable of carrying out for themselves, albeit that they probably would not have done so using a spreadsheet or in the particular format of the Peden schedule. The material contents of the schedules might have been set out as in the schedule attached as appendix 2.
[9] Row H of appendix 1 proceeds on the same basis as row G save that the forecasts to which recoveries are compared are those in the Ferrier Hodgson report. Row H does not feature significantly in the arguments in respect of the recall application and it therefore does not require comment save to say that, as explained in [6], the recoveries made represent 52.26% of those forecast in the Ferrier Hodgson report.
[10] Rows I, J and K represent a cumulative analysis of the accuracy of the forecasts for each of the months of September, October, November and December 2007. This was carried out against the most current forecasts. It will be observed that in row K, the figure in the December column is the same (leaving aside a rounding of 45.99 to 46) as the last figure in the “TOTALS” column of the schedule in appendix 2. As accurately noted in the High Court judgment,[6] 46% represents total recoveries against most current forecasts. The significance of the cumulative total of $71.822m in row I is that it represents a step in the calculation of that percentage.
[11] The 46% figure is less that the 52% referred to in [6] and [8] because the forecasts figures are different, being $63.211m in the case of the Ferrier Hodgson report forecasts and $71.822m for the most current forecasts. The most likely reason for this is that some loans which were forecast to be, but were not, paid in September or October were also forecast to be repaid in November and/or December and in this sense were “double counted” (or conceivably even “triple counted” if included in both the November and December forecasts). This point was advanced, as we will explain, by a defence witness, Mr Gray in his challenge to the Peden schedule.
[12] We note in passing that there was an issue at trial whether the loan recoveries were understated because they did not allow for an apparent loan repayment of approximately $11m in relation to a Mahia development which was part of a transaction by which the same amount of money was re-advanced. This issue was resolved against the inclusion of this figure by both the trial Judge[7] and Court of Appeal[8] and we need say no more about it.

What happened at trial

[13] Mr Carruthers QC put the Peden schedule to Sir Douglas Graham in cross-examination. He explained that he intended to incorporate the schedule in his closing and he was putting it to Sir Douglas to pre-empt complaint of a kind to which he had previously been subject as to the fairness of relying on such an analysis without having first put it to defence witnesses. It was not suggested that the schedule had any evidential value. Rather it was said to be an analysis of what was already in evidence. In permitting the document to be put to Sir Douglas, the Judge made it clear that any errors in it could be addressed by the defendants.
[14] What subsequently happened in relation to the schedule is briefly recorded in [15] of the leave judgment.[9] Although what is said in [15] of that judgment is accurate, the recall application necessitates some amplification.
[15] We have looked at the transcript of the ruling of Dobson J in relation to the stopping of the cross-examination of the defence witness referred to in [15] of the leave judgment. The Judge referred to the defence contention that there were inaccuracies in the Peden schedule, an indication that evidence would be called as to those inaccuracies and a defence argument that in this context, cross-examination “could not be helpful to the Court”.[10] The Judge upheld this argument “in the circumstances as they were”.[11] He referred to the possibility of Ms Peden being recalled to give evidence but this was simply mentioned with the Judge making no comment as to it.[12] On the material before us, the Crown was prepared to recall Ms Peden but this offer was not taken up by defence counsel who instead intimated that her recall would be opposed. The upshot was that she was not recalled.
[16] In their recall application, the applicants contend that the Judge at this time ruled that the Peden schedule:

... could not be put to defence witnesses in cross examination unless and until the Crown applied to recall the author of the document to prove it and to be subjected to cross-examination as to its accuracy[.]

As will be apparent, this is not a complete summary of the Judge’s ruling.

[17] As the leave judgment noted,[13] and as we have already indicated, a defence witness, Mr Gray, later gave evidence in which he analysed and criticised the schedule. In the course of this evidence he raised the double counting issue which we have mentioned. In cross-examining Mr Gray, Mr Carruthers did not challenge Mr Gray’s evidence as to double counting in relation to rows I, J and K but did put to him the calculations in row G, the accuracy of which Mr Gray acknowledged save for what we take to be his reservations in relation to the transaction mentioned in [12]. In the course of this cross-examination, Mr Carruthers added together the row G percentages for September, October, November and December, divided their total by four and came up with an average of 43.75%. While arithmetically correct, this figure (which does not reflect the quantum of the recoveries in the different months) is not particularly helpful. The better and more useful figure is the 46% which appears in row K and as the un-rounded 45.99% in the “TOTALS” column of appendix 2.

The judgment of Dobson J

[18] It is clear that Dobson J was influenced by the row G calculations. At [104] of his judgment he set out the percentages calculated in that row (that is 13.9% for September, 36.4% for October, 69.7% for November and 55% for December).[14] He then went on to say:[15]

Cumulatively, the repayments over those four months were 46 per cent of the amounts projected.

It is clear that the “46 per cent” came from row K. As is apparent, we consider that this is the appropriate percentage figure to represent the row G calculations for the period September – December.

[19] We note that later, when assessing the significance for the future of the shortfalls, the Judge appears to have worked on the basis that the figure was about 50%.[16] The Judge also went on to say:[17]

Since September 2007 the loan managers’ projections had been astray by somewhat more or less than half the predicted repayments in each month.

The Court of Appeal judgment

[20] The Peden schedule was addressed by the Court of Appeal at some length and in different places but for present purposes what is primarily relevant is what was said first at [114]–[115] and then at [157]–[160] of its judgment:[18]

[114] The Crown naturally relied on Ms Peden’s evidence to support its case. The schedule showed forecast loan repayments and drawdowns at various stages over the period August to December 2007 and compared those forecasts with actual repayments and drawdowns over the period. As a percentage of those forecasts, the months average repayments were:

September 13.9 per cent

October 36.4 per cent

November 69.7 per cent

December 55.0 per cent

[115] The cumulative average discrepancy in these figures was said to be 46 per cent. These figures were criticised at trial for accuracy and their relevance disputed as we later discuss. But the Crown’s case was that the significant point to emerge was the trend and scale of the inaccuracies. By the end of November 2007, only $22.4 million had been actually received compared with the company’s forecast for loan repayments of $49 million at the time of the Ferrier Hodgson report in September. This discrepancy was reflected in the continuing deterioration of the company’s cash reserves of approximately $33 million at the date of the Ferrier Hodgson report down to $7.2 million at the date of the 19 December Board meeting.

...

[157] The Crown did not accept there was any double counting but was content to leave to one side the cumulative figures and simply work from a comparison of the average monthly forecasts versus the actual payments received. On that footing, Mr Gray accepted in cross-examination that Ms Peden’s schedule showed that the forecast loan cash flows compared with the payments actually received showed that the average achieved for the months of September, October and November was 40 per cent (being 13.9 per cent, 36.4 per cent and 69.7 per cent divided by three) and, to the end of December, was 43.75 per cent. The average across the whole period to the end of March 2008 was 46 per cent. In agreeing that this was what the figures showed, Mr Gray did not accept their correctness. Nevertheless, they were accepted by the Judge.

[158] In response to Mr Appleby’s criticisms, the Crown submitted that Ms Peden’s approach in analysing the forecast over the period September to December 2007 took into account the adjustments to the Ferrier Hodgson forecast made by the company on a monthly basis. As such, her analysis did not suffer from any defect in that respect. The Crown also pointed out that the analysis adopted by Ms Peden did not rely upon the tracking of actual performance against forecast in respect of particular loan repayments. Rather, it focused on the overall cash position of the company by reference to the total loan repayments forecast (from whatever source) against the actual loan repayments received (irrespective of their source). This gave the true picture of the deterioration of the company’s cash position caused by a failure to meet loan repayments. On this basis, the criticism of double counting was said not to be justified.

[159] Ms Peden’s analysis was also relied upon to support the Judge’s conclusion that the directors ought to have questioned the accuracy of management’s cash flow projections. But we accept that it was prepared in hindsight and was not the sort of detailed analysis that the appellants could have been expected to undertake before issuing the amended prospectus.

[160] On the other hand, despite the possibility of some inaccuracies, the analysis was useful as illustrative of the extent to which loan repayments were not being met as forecast despite close monitoring and frequent adjustment of the forecasts to take into account changing circumstances as they arose. Importantly, the analysis identified in summary form what the directors already knew from the other information they had at hand at critical times. Independently of Ms Peden’s analysis, the directors knew there were major discrepancies between the forecast repayments and the actual payments received. They also knew about the marked deterioration in the company’s cash position ... .

The four grounds upon which the recall application is based

[21] We are satisfied that the four grounds upon which the recall application is based are all misconceived.
[22] In refusing leave to appeal against conviction, the Court was well aware that the Peden schedule had not been produced as a Crown exhibit or proved in evidence. This disposes of the first two grounds relied on in support of the recall application. There is no injustice to the applicants in the fact that the schedule was not produced as an exhibit. All the base figures relied on in the Peden schedule (that is the forecasts and the actual receipts) were otherwise proved in evidence. The calculations in row G were very simple and it was well within the competence of counsel and the Judge to have performed them. It would also be an easy matter for counsel and the Judge to arrive at the 46% figure which appears in row K.
[23] As [15] of the leave judgment makes clear, this Court was well aware that the Judge had stopped cross-examination of a defence witness on the Peden schedule.[19] The applicants’ assertion as to the reasons why the Judge did so is, as noted at [15][16] of this judgment, not complete. More importantly, however, the Court was also well aware that the Peden schedule was not independent evidence of its contents or premises but rather merely purported to be a summary of, and arithmetical calculations based on, what was otherwise established by the evidence. The third recall ground is thus also without substance.
[24] In the brief discussion on the Peden schedule in the leave judgment, reference was made to the corresponding discussion in the Court of Appeal at [109]–[115] and [155]–160].[20] At [112], the Court of Appeal indicated that Mr Farmer QC, counsel in those proceedings for the parties who are now the applicants, had not pressed the argument that “there had been a breach of natural justice or a procedural flaw” in respect of the production of the schedule. This comment was not material to the decision not to grant leave to appeal against conviction. We note in passing that we have read the transcript of that part of the hearing before the Court of Appeal where the members of the Court discussed the status of the Peden schedule with Mr Farmer. Having done so we agree that Mr Farmer did not abandon the procedural point. On the other hand, in the face of what was plainly some scepticism from the bench as to its underlying merits, he did not press it firmly.
[25] The fourth ground relied on in support of the recall application is therefore also without substance.

Other complaints

A preliminary comment

[26] As we have noted the applicants have advanced further complaints, namely that the Peden schedule involved a departure from the particulars in the indictment and as well that the calculations were wrong and lacking in logic as involving double counting.

Did the Peden schedule involve a departure from the Crown case as particularised?

[27] The indictment alleged that the amended prospectus was untrue in respect of liquidity risk, an allegation which was particularised in the indictment in this way:

The presentation of ‘liquidity risk’ omitted material particulars namely actual and significant adverse liquidity issues including; the deterioration in the cash position from the balance of 30 September 2007, the failure to achieve forecast (internal and Ferrier Hodgson) cash receipts and loan repayments.

Further particulars were sought and provided by reference to certain paragraphs in: (a) a summary of facts; and (b) the briefs of evidence of three witnesses, Ms Peden, Mr Cable and Ms McGreevy. In recording this, the trial Judge noted:[21]

... the Crown submissions acknowledged the prospect that the detail of its allegations may alter, prejudice being a [touchstone] as to whether that will be appropriate.

[28] In the summary of facts, the only forecast discussed in detail was in relation to recoveries of “approximately $49 million” in respect of certain major loans between September and December 2007 (the “major loans forecast”) (the precise figure was $49.075m). There were references to failures to “[receive] loan repayments as forecasted by [KordaMentha]” which, on a literal interpretation, could be treated as a reference back to the major loans forecast but need not necessarily be so confined. There were also more general references to cash on hand being less than forecast, references which, of necessity, bring in the forecasts which gave rise to that figure; these including forecasts other than the major loans forecast. The brief of evidence of Ms McGreevy referred, in the paragraphs identified by the Judge, to other forecasts and the briefs of evidence of Ms Peden and Mr Cable also referred to other forecasts. Although this was not in the paragraphs specified by the Judge there was no suggestion that those paragraphs should be struck out.
[29] We do not accept that Crown reliance on the forecasts other than the major loans forecast lay outside the case as particularised. This would involve taking an entirely unreal and artificial view of the situation. Unsurprisingly it was not the way the case was run (as all forecasts were produced in evidence and discussed) and there was plainly no suggestion that anything other than the major loans forecast was out of bounds. If this suggestion had been advanced in the High Court, the Judge would have amended the indictment/particulars to make explicit what was in any event obvious. This point was not taken in the Court of Appeal and was not raised in the application for leave to appeal to this Court. It is far too late to raise it now.

A related argument

[30] In what was an unnecessary and unsatisfactory complication to the material which we have had to address, two of the applicants in their affidavits compare recoveries for September – December of $33m (the actual figure was $33.033m) with what they say were anticipated receipts in that period of $49m. For instance in his affidavit Sir Douglas asserts that one of the critical omitted matters relied on by the Crown was:

[T]hat September 2007 loan repayment forecasts were inaccurate in that of the $49m expected in loan repayments between September and December 2007 only $33m [67%] had been received.

Mr Jeffries also assumed that the loan repayments as forecast for this period were $49m. This point is related to, but not exactly the same as, the particulars issue which we have just discussed.

[31] Confusingly, there are three figures of approximately $49m which appear in the material we have seen:
[32] On the basis of the material we had seen we could not see how it was relevant to compare recoveries for September – December of $33.033m with a figure of $49m. The [31](a) figure is a forecast for September – November and it would be illogical to compare it to recoveries for the September – December period. The [31](b) figure is the major loans forecast but the $33.033m includes recoveries in respect of other loans. The [31](c) figure was not substantially relied on at trial as it predated the Ferrier Hodgson report. Accordingly, we invited further submissions.
[33] The response of the applicants was that:

The figure of $49m referred to in the affidavits of applicants Graham and Jeffries ... constituted the Crown case of expected loan receipts for the period September to December 2007.

Counsel for the applicants sought to justify this proposition by reference to evidence as well as on the basis of the particulars argument we have discussed. But the reality that the 26 September forecast of loan repayments for the four months came to $63.211m means that this proposition represents a substantial mischaracterisation of the Crown case.

A double counting problem?

[34] As has been seen, the actual recoveries in September – December were $33.033m, or 45.99% of the $71.822m that was the cumulative total of the most current forecasts for each month. The double counting complaint relates to the use of the $71.822m figure. The nub of the complaint, as alluded to earlier, is that aggregating the most current forecasts allows the possibility that certain recoveries expected, but not received, in one month could then be expected for a second time in an updated forecast for the next month. When the most recent forecasts for each month were added certain forecast recoveries could, in short, be counted more than once. The applicants contend that the $71.822m figure overstates the extent of total recoveries expected in the relevant period, with the result that the 45.99% figure is misleading.
[35] As will be apparent, the applicants maintain that the appropriate figure against which recoveries of $33.033m should be compared is $49m. For reasons already given, however, it is clear that the total the recoveries forecast in September for the relevant four month period was $63.211m. The double counting complaint is therefore confined to $8.611m (being the difference between $71.822m and $63.211m). The complaint is, in any event, misconceived because it misconstrues the significance of the $71.822m and 45.99% figures. As has been mentioned, the indictment put in issue failure to achieve not only Ferrier Hodgson forecasts but also internal forecasts, which were updated from time to time. Relevantly to this, row K measures the extent to which actual recoveries were falling short of forecasts as they were updated. It does so by expressing total recoveries as a percentage of aggregated updated monthly forecasts. It is one way – of course not the only way – of looking at the substantial underperformance of actual as against forecast recoveries over the September to December period.
[36] The whole issue is in any event of no moment. As discussed, actual recoveries in the relevant period were 52.26% of the Ferrier Hodgson forecasts. These forecasts were not susceptible to double counting criticism. As was noted at [19], Dobson J spoke of loan managers’ predictions as being astray by “more or less than half”, an approach which he was entitled to take irrespective of what appeared in rows I, J and K of the Peden schedule.

Other issues

[37] The applicants seek orders addressed to preparation of a case on appeal, orders which are now unnecessary on the approach we have adopted.
[38] In the affidavits filed in support of the recall application there are complaints about the conduct of the prosecuting agency and generally in connection with the processes which resulted in prosecution, the conduct of the Crown case, the reasoning of the Judge and what happened in the Court of Appeal. To the extent that these are not reflected in the recall application nor developed in the submissions filed we do not address them.

Result and costs

[39] For the reasons given, the application for recall is dismissed.
[40] The Crown seeks costs against the applicants.
[41] The grounds upon which the recall application was based were all completely misconceived. The recall application was presented without anyone having made a serious attempt to analyse the grounds which were to be advanced or the figures relied on. As a consequence, there was an extraordinary mixing and matching error as to the repayments forecast and achieved for the September – December period. Even when problems with the figures relied on were pointed out in a minute, counsel for the applicant persisted with the misplaced contention the Crown case was based on forecast recoveries of $49m for the September – December period. There has been substantial mischaracterisation of what happened at trial. More generally, the recall application is very much a second shot, based in large measure on arguments which, if meritorious, could and should have been, but were not, advanced when leave to appeal was first sought. In the result, counsel for the Crown have been required to chase down a series of hares which the applicant ought never to have started.
[42] For the reasons just given, we are giving consideration to whether we ought to order the applicants to pay costs. If the applicants wish to make submissions in opposition to the making of such an order, they should file and serve submissions within 10 working days. In the meantime, costs are reserved.


Solicitors:
Graham & Co, Auckland for Applicants
Luke Cunningham Clere, Wellington for Respondent

Appendix 1: Spreadsheet included in the Peden Schedule with added row identifiers


A
B
C
D
E







F
G
H


I
J
K


2015_10900.jpg







Appendix 2
September
October
November
December
TOTALS
Forecasts
14.251
18.765
19.605
19.201
71.822






Recoveries
1.978
6.830
13.655
10.570
33.033






Percentage
13.9
36.4
69.7
55.0
45.99



NOTES

[1] The forecast figures correspond to those in bold in rows B, C and D and the recovery figures and percentage figures appear in rows F and G.

[2] The figures in the last column (“TOTALS”) correspond to the figures which appear in the December column of rows I, J and K.

[1] R v Graham [2013] NZSC 104 [Supreme Court leave judgment].

[2] R v Graham [2012] NZHC 265, [2012] NZCCLR 6 (Dobson J) [High Court judgment].

[3] Graham v R [2014] NZSC 55.

[4] There is one qualification to this. The $16.875m figure for November in row B differs from the corresponding figure in [4.1] the Ferrier Hodgson report. However, the total for September – November in row B corresponds with the aggregate figure for that period used in the Executive Summary of the Ferrier Hodgson report, suggesting that the Ferrier Hodgson figure for November at [4.1] was wrong. Row B corresponds to figures included as Appendix 2 to the Ferrier Hodgson report.

[5] We note that two of the applicants in their affidavits compare recoveries for September – December of $33.033m with what they say were anticipated receipts of $49m, rather than the actual figure anticipated for the period of $63.211m. This is discussed below at [30]-[33].

[6] High Court judgment, above n 2, at [104].

[7] At [103].

[8] Jeffries v R [2013] NZCA 188 (Randerson, Wild, French JJ) [Court of Appeal judgment] at [156].

[9] Supreme Court leave judgment, above n 1.

[10] R v Graham HC Wellington CRI-2010-085-2538, 24 November 2011 at [3].

[11] At [4].

[12] At [5]–[6].

[13] Supreme Court leave judgment, above n 1, at [15].

[14] High Court judgment, above n 2 (citation omitted).

[15] At [104].

[16] At [114].

[17] At [328].

[18] Court of Appeal judgment, above n 8 (citations omitted).

[19] Supreme Court leave judgment, above n 1.

[20] Supreme Court leave judgment, above n 1, at [16].

[21] R v Graham HC Wellington CRI-2010-085-2538, 8 July 2011 at [129].

[22] At 1484 of the transcript of evidence.

NZLII: Copyright Policy | Disclaimers | Privacy Policy | Feedback
URL: http://www.nzlii.org/nz/cases/NZSC/2015/109.html