QBE Insurance (International) Ltd v Pegasus Group Ltd [2011] NZCA 268; (2011) 16 ANZ Insurance Cases 61-894 (13 June 2011)

Last Updated: 22 June 2011

IN THE COURT OF APPEAL OF NEW ZEALAND

CA830/2009 [2011] NZCA 268

BETWEEN QBE INSURANCE (INTERNATIONAL) LIMITED First Appellant

AND AMERICAN HOME ASSURANCE COMPANY Second Appellant

AND PEGASUS GROUP LIMITED Respondent

CA58/2010

AND BETWEEN AMERICAN HOME ASSURANCE COMPANY Appellant

AND PEGASUS GROUP LIMITED Respondent

Hearing: 3 and 4 May 2011

Court: Ellen France, Randerson and Harrison JJ

Counsel: M G Ring QC, M O Robertson and H P Twomey for First Appellant in CA830/2009 A Challis and K Harkess for Second Appellant in CA830/2009 and Appellant in CA58/2010 M J Tingey and B M M McKinlay for Respondent

Judgment: 13 June 2011 at 3 pm

JUDGMENT OF THE COURT

1. Leave granted to the appellants in CA830/2009 and CA58/2010 to file amended notices of appeal.
   
2. The appeals in CA830/2009 and CA58/2010 are dismissed.
   

100. The cross-appeal in CA830/2009 is dismissed.
     

4. QBE and AHA must each pay Pegasus an amount equal to 85 per cent of costs for a standard appeal on a band A basis and usual disbursements.
   

REASONS OF THE COURT

(Given by Harrison J)

Contents

Introduction

[1] Pegasus Group Ltd, the respondent and cross-appellant, imports and distributes branded sporting and leisure goods. Between 2001 and 2005 the company stored its stock in warehouses operated by New Zealand Express Ltd (NZE). As a result of a stocktake in February 2005, Pegasus discovered that large quantities of its goods were missing. It claimed that the missing stock was stolen by NZE employees. But before Pegasus could take any formal steps NZE was placed in liquidation.
[2] Pegasus had a policy of material damage insurance with QBE, the first appellant. The company made a claim for indemnity for its goods lost while in NZE’s custody. QBE denied liability.
[3] NZE had what was called a crime insurance policy with American Home Assurance Company (AHA), the second appellant. That policy indemnified NZE against liability to third parties for goods lost while in its custody. Pegasus made a claim for its loss directly under the AHA contract following NZE’s liquidation. AHA denied liability.
[4] Pegasus issued proceedings in the High Court at Auckland against both insurers. Following a two week trial, Winkelmann J found that Pegasus had proved its claim against each insurer. In both cases the company relied substantially on its stock records to establish the thefts and the amount of its loss. Winkelmann J entered judgment for Pegasus against QBE for these sums:[1]

• (a) $354,369 for material damage less:
  • (i) $2,500 by way of policy excess;
  • (ii) any amount which Pegasus recovered from AHA pursuant to the judgment (excluding any award for interest and costs);
  • (iii) $197,637.68 to reflect a set-off;
• (b) $412,588 for loss of profit due to business interruption;
• (c) $48,747 for expenses;
• (d) compound interest calculated according to a fixed formula.

[5] Winkelmann J separately entered judgment for Pegasus against AHA for:

• (a) $354,369 less:
  • (i) $25,000 by way of policy excess;
  • (ii) $197,637.68 by way of set-off;
• (b) interest on the judgment sum from December 2006.

[6] All three parties are dissatisfied with the judgment. QBE and AHA appeal against the Judge’s findings on liability. Each says that the effect of an inventory exclusion clause in its policy was to bar Pegasus from using stock records to prove its loss or the amount; and that, in any event, while there was evidence that NZE’s employees stole Pegasus’ stock frequently and over an extended period, the Judge had no proper factual foundation for inferring that it was on a large scale. However, the insurers do not challenge other significant findings by the Judge including the assessment of Pegasus’ business interruption claim and consequential losses. It is common ground that the fate of these two awards depends on QBE’s primary challenge to the material damage award. Both will be set aside if that appeal succeeds. In this respect, without opposition from Pegasus, we grant QBE and AHA leave to file amended notices of appeal. Mr Tingey for Pegasus expressed concerns about his ability to respond to one aspect of the proposed amendments but as the hearing progressed he did not seek leave to file further submissions on the point.
[7] Pegasus cross-appeals against the allowance of a set-off against QBE’s liability but does not challenge the dismissal of its claim of bad faith against that insurer.

Indemnity

(a) Approach

[8] Both Mr Ring QC and Ms Challis for QBE and AHA respectively criticise Winkelmann J’s approach to determining Pegasus’ claim. The primary stage in her analysis was to determine whether Pegasus had suffered loss and if so to what extent. The Judge acknowledged that the company had no direct evidence of actual thefts of the scale alleged but based its case on what she described as two threads or strands of circumstantial evidence.[2] One thread was proof of the accuracy and reliability of Pegasus’ stock records in disclosing the level and value of the lost goods. The other was the nature and scale of NZE’s lack of security and its ingrained culture of employee defalcations.
[9] Winkelmann J was satisfied that there was a sufficient factual foundation from which to infer that sustained staff thefts were the likely cause of the goods’ disappearance and that Pegasus’ claim was properly quantified. After making these findings, she proceeded to the next stage of determining the company’s rights of indemnity under each policy. Mr Ring and Ms Challis say the Judge should have started by construing the policy documents, in particular the inventory exclusion provisions, to settle Pegasus’ rights to indemnity from the outset.
[10] We are not satisfied that Winkelmann J’s approach was in error. Pegasus had to prove both the cause and amount of its loss to bring its claim within the operative clause of either policy. While the relevant provisions under both policies had similarities, there were also differences. The Judge was entitled to conclude that issues of indemnity, and in particular the application of the inventory exclusion clauses, were best determined once primary factual findings were made.[3]
[11] We accept, nevertheless, that on appeal Mr Ring and Ms Challis have given priority to the insurers’ defences based on the inventory exclusions. Accepting their arguments will result in Pegasus’ inability to rely on its stock records to prove either the cause or amount of its loss, with a decisively adverse effect on its claim. Instead, as both counsel submit, the company will be restricted to other evidence. Given the way the insurers’ case has been presented on appeal, we shall approach it in the order submitted by Mr Ring and Ms Challis. However, Winkelmann J’s brief references to the exclusion provisions suggest that this line of defence did not enjoy the same degree of prominence at trial.

(b) QBE

(i) Policy

[12] Pegasus was insured with QBE for material damage. The contract was described as a “Combined Corppak” policy, noting Pegasus’ business as importing, assembling and distributing sporting goods and toys at NZE’s premises. Stock to a maximum value of $2 million was the largest component of the company’s insured property. An excess of $2,500 was to apply for losses due to theft. The annual premium was $17,555.
[13] By cl 21.1.1 of the policy, QBE agreed that:

If, during the Period of Insurance, physical loss or damage – unintended or unforeseen by the Insured – happens to any of the Insured Property, then, subject to the terms, conditions and exclusions of this section of this Policy and the General Policy Exclusions and General Policy Conditions and Clauses, the Company will indemnify the Insured for the loss or damage and expenses. The Insured will be indemnified by payment or, at the Company’s option, by repair or by replacement of the lost or damaged property.

(Emphasis added.)

[14] QBE accepts that Pegasus’ claim fell within this insuring or operative clause. However, the insurer relies upon an inventory exclusion provision, cl 21.3.5, which, Mr Ring says, modifies the insuring promise. The clause provided:

This section does not insure unexplained disappearance, loss, or shortages revealed at any stocktaking or shortages due to accounting or clerical errors.

[15] Mr Ring challenges Winkelmann J’s conclusion that QBE’s inventory exclusion did not apply for this reason:

[156] Pegasus has proved that stock has been lost by reason of staff theft at NZE. It has brought itself within the terms of clause 21.1.1. I am also satisfied that exclusion clause 21.3.5 has no application, as this is not an unexplained disappearance, nor is it a shortage due to accounting or clerical errors. QBE suggested an interpretation of the exclusion clause whereby the obligation to indemnify is also excluded if the shortage is revealed at any stocktaking. But I am satisfied that the only coherent reading of the clause is that the expression “shortages revealed at any stocktaking” follows on from and is qualified by “unexplained disappearance”. To be clear, the clause excludes liability to indemnify for unexplained disappearances revealed at any stocktaking. It follows that Pegasus is entitled to indemnity for the value of stock which it has proved is missing by reason of theft.

(ii) QBE’s case

[16] Mr Ring subjected cl 21.3.5 to detailed analysis. It includes, he says, at least two basic and discrete alternatives: either “unexplained disappearance, loss or shortages revealed at any stocktaking”; or “shortages due to accounting or clerical errors”, which plainly does not apply. Mr Ring says the phrase “unexplained disappearance” was intended to stand alone because any “unexplained disappearance” must also be a “loss” but not vice versa. So, if both were intended as alternatives, each separately qualifying “revealed by any stocktaking”, the former would be redundant as subsumed in the latter.
[17] Mr Ring says it must follow that there is a separate exclusion for “shortages revealed at any stocktaking” – that is, the difference between actual and expected quantities; and that this alternative exclusion extends to the preceding “loss”. On this basis, “unexplained” does not qualify either phrase but is limited to “disappearance”. In Mr Ring’s submission, the phrase “revealed by any stocktaking” refers to making known or visible what was previously unknown or unable to be seen. Thus, he says, the exclusion proscribes the insured’s use of a loss or shortage revealed by a stocktaking for the purpose of satisfying the essential elements of the operative clause of the policy relating to the existence and/or the quantity of a physical loss of insured property.
[18] In amplifying this submission, Mr Ring says:

• (a) clause 21.3.5 reflects a mutual recognition by the parties of a much higher risk of error – as to both existence and quantity of loss – in relying on an inventory. That is because its overall accuracy depends not on empirical evidence of physical loss but on the individual accuracy of a number of processes by a number of people over two or more stocktakes and in the intervening period. Included in this group are the counters, recorders and data entry operators for the first stocktake, the personnel generating the stock movement documents and the data entry operators for those documents, the personnel receiving and dispatching deliveries, and the counters, recorders and data entry operators for the second stocktake;
• (b) as a result, if shortages revealed by stocktakes were sufficient to prove physical loss, the insurer would actually be insuring the reliability and accuracy of the stocktakes, of the related stock movement processes and records and of all the personnel concerned, rather than insuring the physical risk in the goods themselves;
• (c) a common intention may be inferred from the terms of the exclusion that instances may arise where even though the insured has suffered a general and physical loss it will not be entitled to indemnity because the only evidence of the existence or quantity of this physical loss falls foul of the exclusion. Someone has to take the risk where there is the recognised potential for uncertainty and unreliability, and it must also be inferred that the mutual basis on which the premium was calculated and paid was that the insured took this risk.

[19] In summary, Mr Ring says that, to avoid the application of cl 21.3.5 while still satisfying the operative clause, Pegasus had to prove the existence and quantity of physical loss for which it claimed to be entitled to indemnity by means of evidence other than loss or shortages revealed by various stocktakes – such as by eye witness accounts, confessions or circumstantial evidence. Mr Ring’s primary proposition is that Pegasus is barred from using the stock records, regardless of their evidential quality or value, for any purpose whatsoever in support of a claim. His fallback proposition is that, if an insured party has independent evidence from which loss can be inferred on a large scale and the stocktake reveals the exact amount of lost goods, then the loss is not being revealed by the stocktake and the exclusion does not apply. He submits, however, that there was no such independent evidence in this case.

(iii) Analysis

[20] Clause 21.3.5 is not happily worded. It is open to a number of interpretations. Mr Ring’s first alternative includes at least these possibilities:

• (a) “unexplained disappearance or loss” can effectively be treated as synonomous so that both are qualified by the term “unexplained”; or
• (b) “a shortage revealed at any stocktaking” is also qualified by the term “unexplained”; or
• (c) the words “revealed at any stocktaking” are referable not just to “shortages” but also “unexplained disappearance or loss”.

[21] QBE’s policy must be construed as a whole and in accordance with its purpose. In our judgment, the operative or insuring clause, cl 21.1.1, is the starting point for considering Mr Ring’s submission. It provides cover for material damage or physical loss of stock to a value of $2 million. To qualify for indemnity, the insured has to prove that physical loss – unintended or unforeseen – has happened to its property. The insured perils are not nominated. However, it is common ground that theft of Pegasus’ stock from NZE’s warehouse is one such peril unless excluded by cl 21.3.5. In this case, Pegasus had to establish that theft was the proximate cause of its loss, and its amount.
[22] We are satisfied that the purpose of the exclusion, construed in the context of its overall terms, is to exclude cover for disappearance or loss, or shortages revealed at any stocktaking only in circumstances where there is no independent evidence to explain the disappearance, loss or shortage. We accept that cl 21.3.5 would exclude liability where there is no evidence of loss other than a shortage revealed at stocktaking. Such a loss could be due to a variety of reasons which do not amount to physical loss or damage.
[23] Inaccuracies in the stocktaking are an obvious example. In that case the shortage revealed could not amount to physical loss or damage within the operative provision. As Mr Ring submits, the insurer would effectively be guaranteeing the accuracy of the stocktaking. By referring to the manner of discovery, the parties recognise the theoretical nature of the shortage. That is because it comes into existence solely as the result of an accounting process, not as a result of an insured peril such as theft, burglary or fire.
[24] Thus, to satisfy the operative provision Pegasus had to establish that it suffered physical loss of its stock caused by an insured peril. The fact that a stocktake revealed a shortage of items was insufficient of itself for Pegasus to discharge its burden of proof. That fact could not without more prove the loss or justify an inference that the event was caused by theft. Evidence external or additional to the stocktake was necessary, to prove that the physical loss had happened and explain its cause.
[25] However where that independent evidence is available and the insured is able to prove that the shortage has occurred through an insured peril then cl 21.3.5 does not operate to exclude the insurer’s liability. By contrast, where there is no explanation for the disappearance or loss, liability is excluded. Even if the word “unexplained” does not extend to qualify the phrase “shortage revealed at any stocktaking”, its presence colours the clause as a whole. In summary, the inventory exclusion when read with the insuring clause, supports the conclusions that, first, an unexplained disappearance or loss is not an insured peril and, second, a shortage revealed at a stocktaking is not without more a physical loss. These two elements are complementary both within the exclusion clause and with the operative clause.
[26] In Betty v Liverpool and London and Globe Insurance Company Ltd[4] the policy excluded “unexplained loss or mysterious disappearance of property ... or loss or shortage of property disclosed on any inventory ...”. The United States Court of Appeals, Fourth Circuit, described the purpose of the exclusion as being to enable:[5]

... an insurer to except itself from a loss or shortage reflected solely on the insured’s books and not substantiated by independent external proof – a mere theoretical inventory loss. Such an interpretation is compatible with the other provisions of the exceptive clause which excludes ‘unexplained loss or mysterious disappearance’.

[27] However, on Mr Ring’s argument, cl 21.3.5 goes significantly further. It renders the evidence of the stocktaking inadmissible when tendered for the purpose of proving physical loss. In support of this argument Mr Ring cites a number of North American decisions. The decisions, from different jurisdictions in Canada and the United States, are not easily reconcilable, in part because the subjects exclusions vary materially. We are unaware of any New Zealand, Australian or English authorities on point.
[28] Of the United States decisions Betty is of most assistance, not only because of its authority but also because the exclusion clause, cited above, is the most analogous to QBE’s. In that case an insured tyre company suffered an unexplained quarterly financial loss. Previously it had been profitable and had enjoyed increased sales in the particular quarter. A physical inventory count disclosed over 1,000 missing tyres. The loss was unexplained. In reliance on the exclusion the insurer declined the company’s material damage claim. In upholding the insurer, the District Court declined to take into account the evidence of a witness who was convicted of stealing seven of the tyres (they were later returned by the police) and who said he had seen a truck loaded with tyres driving away from the yard. The Judge relied on the ground that the tyre company had not known of its losses until the inventory count and did not learn of the witness’s theft until later.
[29] In reversing the District Court, in words which implicitly answer Mr Ring’s submission, the Court of Appeals said this in Betty:

In order to affirm the judgment below it is necessary to construe the exceptive clause of the policy to mean that no loss is covered if it is first discovered upon taking inventory, no matter what proof may be subsequently brought to light showing the loss to be clearly within the risks for which the policy was written. We feel that such a construction would be unrealistic. It does not seem reasonable to us that business men would enter into an agreement to insure against a loss discovered in one way and not insure against the same loss if it should be discovered in another way ...

[30] The Court noted that exclusion clauses of this nature were originally developed for use in theft policies, to protect the insurer against frauds by the insured party for what were called ‘inside jobs’. (There was no suggestion in Betty that its genesis lay in what Mr Ring submits was the parties’ mutual recognition that a stocktake was an inherently unreliable process because of its high risk of error.) Later the Court said this:[6]

In the setting in which the phrase ‘unexplained loss or mysterious disappearance’ is used in the policy ... that is to say as an exclusion to an all risks policy, we think it means a disappearance of property where the surrounding circumstances offer no logical explanation of what happened to it ...

[31] As we read Betty, its effect is to recognise a distinction between uninsured loss and uninsured perils or causes. Under the policy the loss was uninsured if the only proof of its existence was the inventory. But even if a physical loss was proven, it was uninsured if its cause remained unexplained or mysterious.
[32] Mr Ring relies on two other United States authorities. In Betco Scaffolds Company Inc v Houston United Casualty Insurance Company[7] the Texas Court of Appeals split by a five – four majority in favour of an insurer’s construction of a similar exclusion clause. The majority distinguished (arguably declined to follow) Betty, concluding that:[8]

Based on what we perceive to be the overall intent of the policy and the plain meaning and rationale of the inventory exclusion clause, we conclude that this provision expressly allocates to the insured the risk of a loss or shortage which comes to the attention of the insured solely by reason of taking a regularly scheduled, ie periodic, physical inventory. We recognise that a regularly scheduled inventory could coincide with the investigation of a casualty in such a way that the inventory is intended by the insured as a means to quantify the loss. In that event, the inventory exclusion provision would not exclude the loss because the loss would not have been disclosed upon taking inventory.

The second part of this passage echoes Mr Ring’s fallback position on the interpretation of QBE’s exclusion.

[33] The minority in Betco, in a strongly worded dissent delivered by the Chief Justice, Murphy CJ, pointed out that the majority’s decision was contrary to a number of other American authorities, which had declined to apply similar exclusion clauses where the loss was linked to an identifiable external cause. In the minority’s opinion, it was the amount of the loss, not the loss itself, that was discovered during the inventory. The minority observed further that:[9]

Under the majority view, any loss disclosed by an inventory, whether or not the loss was due to an event covered under the policy, would not be covered under the policy. The moral of the story told by the majority is that an undetected thief is an insurer’s best friend.

[34] Mr Ring also relies on Banner Lumber Company Inc v Indiana Lumbermen’s Mutual Insurance Company.[10] In that case the District Court found that there was no independent evidence from which it could be inferred there was a loss on the large scale established by the inventory. The only independent evidence was of one small scale theft. Thus the Court found for an insurance company on an exemption which excluded cover for:

... missing property where the only proof of loss is unexplained or mysterious disappearance or shortage discovered on taking inventory, or other instance where there is no physical evidence to show what happened to the property.

(Emphasis added.)

[35] The Banner provision is relevant because its terms explicitly exclude liability for loss where its “only proof” is one of the events identified in QBE’s exclusion. Moreover, the concluding phrase (“there is no physical evidence”) reinforces an intention to exclude a claim where the cause of loss (“what happened to the property”) is not explained. Banner’s wording is closer to the construction which Mr Ring advances. His problem is that cl 21.3.5 omits these critical references.
[36] Counsel cited a number of first instance Canadian decisions. The exclusion clauses in Finlayson Enterprises Ltd v United States Fidelity and Guaranty Company[11] and Flett Motors Ltd v Royal Insurance Company of Canada Ltd,[12] were materially identical. Like the provision in Banner, they excluded liability for loss where:

Proof of which either as to its factual existence or its amount, is dependent upon an inventory computation or a profit in loss computation; provided, however, that [this exclusion] shall not apply to loss [of property] which the insured can prove, through evidence wholly apart from such computations, is sustained by the insured through any fraudulent or dishonest act or acts committed by any one or more of the employees.

[37] In Finlayson the Ontario High Court of Justice admitted evidence of an inventory calculation undertaken by a textile company. It followed a former employee’s admission that he had stolen stock (sheets and pillow cases) to a value of about $1,000. The insurer under a fidelity policy declined liability. Osler J accepted that the existence and cause of the loss was independently established by the ex-employee’s dishonesty. He took the inventory assessment into account solely as proof of the company’s loss of stock worth $48,000. The Judge also discounted Finlayson’s claim by 5 per cent to allow for the risk of error in the stocktaking process. Mr Ring said that a similar discounting approach should not be applied here.
[38] The decision in Finlayson is relevant because in a case where the exclusion was explicit as to the admissibility of inventory evidence, in contrast to cl 21.3.5, the Court accepted that evidence – as did Winkelmann J in this case – as proof of the amount of loss even though it greatly exceeded the amount proven by the admission of theft by the dishonest ex-employee. Osler J must have inferred that all the loss revealed by the inventory count was attributable to the dishonest ex-employee. The only factual distinction from Pegasus’ case was that the ex-employee’s admission of guilt in Finlayson preceded the verifying inventory.
[39] In Flett a stocktaking revealed a significant shortage of parts belonging to a car dealer. The company engaged a security firm to investigate. An employee admitted theft of goods to a value of between $5,000 and $10,000. The dealer made a claim under its fidelity policy. The insurer declined to indemnify. In the New Brunswick Court of Queens Bench, Riordan J held that the inventory evidence was admissible both to prove the fact of theft and, significantly, to verify its amount of $26,000. While the stocktaking revealed a shortage, the existence of the company’s loss did not depend for proof on the inventory computation given the undisputed evidence of employee theft. Again, the facts are closely analogous to this case.
[40] In Trenville Farms Ltd Partnership v Wawanesa Mutual Insurance Co[13] a hog farming partnership claimed against its insurer for the loss of 1,029 of its 3,000 hogs. Their disappearance was disclosed on a stocktake. The partnership had no direct evidence of theft. But the partnership’s position was that there was no other reasonable explanation because, as Kenny J noted, “that number of hogs do not simply disappear into thin air”. Its insurer declined the claim relying on a provision, materially identical to QBE’s, that excluded:

Loss due to unexplained or mysterious disappearance of property, or shortage of property disclosed on taking inventory.

[41] Kenny J reviewed the earlier Canadian authorities including Finlayson and Flett. She was satisfied that the disappearance of 33 per cent of the partnership’s stock was not mysterious but could only be explained as theft. Hence the first part of the exclusion did not apply. Additionally, the shortage was not disclosed on taking inventory because a member of the partnership had commissioned that course after he suspected some of the hogs were missing. The inventory computation was admissible for that reason.
[42] On our reading, these Canadian decisions directly undermine Mr Ring’s argument. In each, the Court admitted stock records in circumstances close to Pegasus’ claim. In particular, contrary to Mr Ring’s submission, the Canadian courts have consistently recognised the evidential value of inventory assessments in verifying the amount of loss where the direct evidence of theft did not itself establish that figure; and none of the decisions could be construed as supporting the point of absolute or unconditional exclusion of stocktake records inherent in Mr Ring’s principal proposition.
[43] In this case, Pegasus’ obligation was to bring its claim within the operative or insuring provision of QBE’s material damage policy. Its best evidence of physical loss was of thefts by NZE employees and stock records. In combination, these two lines of evidence were probative of loss in a defined amount. Clear and unequivocal words would be required to exclude either or both because such a result would largely deprive Pegasus’ policy of commercial value.
[44] In effect, Mr Ring’s argument seeks to import into QBE’s inventory exclusion the specific provisions excluding the inventory records which were incorporated into the policies considered in Banner, Finlayson and Flett. But even those provisions do not go as far as Mr Ring urges. It is unnecessary to invoke the principle that exclusion clauses are to be construed narrowly[14] because, whatever rule of construction is applied, cl 21.3.5 cannot be construed, in our judgment, to support Mr Ring’s primary proposition that evidence which may otherwise be relevant and probative in support of a material damage claim must be excluded whatever the circumstances. And, as Mr Tingey submits in answer to Mr Ring’s fallback proposition, there is nothing in cl 21.3.5 which allows the stock records to be used only to establish quantum within a range established by other evidence. On our view, the undefined scope of Mr Ring’s fallback proposition, unrelated to the wording of the inventory exclusion, exposes the inconsistency in and unsustainability of his primary proposition. We are satisfied that QBE cannot rely on cl 21.3.5 in answer to Pegasus’ claim.

(c) AHA

[45] It is common ground that Pegasus was entitled to pursue AHA in its capacity as NZE’s insurer. By s 9 of the Law Reform Act 1936, a claimant is entitled to receive the benefit of an insured party’s claim to indemnity from its insurer in circumstances such as these. As Winkelmann J noted, this constitutes a statutory charge in favour of Pegasus, and s 9 puts the insurer in the same position as its insured for the purposes of considering Pegasus’ claim.[15] Winkelmann J set out the relevant provisions of AHA’s policy as follows:

[12] Prior to liquidation NZE had entered into the crime insurance policy with AHA to provide insurance for certain risks including theft. The policy provided that NZE was entitled to indemnity as follows:

The Insurer shall indemnify the Insured for their direct financial Loss sustained at any time consequent upon a single act or series of related acts of Theft, fraud, dishonesty or criminal damage committed by any Employee (acting alone or in collusion with others) or Theft or criminal damage committed by any other person, which is:-

(i) committed with the clear intent to cause the Insured a Loss, and

(ii) Discovered by the Insured during the Policy Period or the Discovery Period, and

(iii) committed within the Geographical Limits, and

(iv) not excluded under the terms and conditions of this policy.

...

[13] The definitions clause, clause 2, contained the following definitions of relevance:

2.11 “Loss” means the direct financial loss (other than salaries, commissions, fees, bonuses, promotions, awards, profit sharing, pensions or other employee benefits paid by the Insured, which are not deemed direct financial loss), sustained by the Insured in connection with any single act or series of related, continuous or repeated acts (which shall be treated as a single act) of Theft, fraud, dishonesty or criminal damage committed by any Employee (acting alone or in collusion with any other Employee or others) or of Theft or criminal damage committed by any other person, or in which such Employee or other person is involved or implicated. Loss includes any liability the Insured may have to restore or make good Money, Securities, or other property of another person.

2.22 “Theft” means the unlawful taking, including by violence or threat of violence, of Money, Securities, or other property to the intended permanent deprivation of the Insured.

[14] Significantly, however, since NZE’s business involved it storing the property of others, the policy contained a clause that extended cover to the property of third parties which provided:

Cover is extended to include direct financial Loss sustained by any other person or organisation following Loss of Money, Security, or other property under the care, custody or control of the Insured or for which the Insured is liable.

[15] Like QBE, AHA relies on its inventory exclusion clause. Exclusion 4.9 provides that AHA is not liable for:

Loss, the proof of which is dependent solely upon a:

(a) profit and loss computation or comparison; or

(b) comparison of inventory records with an actual physical count.

However, where an Employee is involved in [or] suspected of causing Loss and has been identified, inventory records and actual physical count of inventory can be submitted as partial evidence in support of proof of Loss.

[46] The Judge dismissed AHA’s reliance on clause 4.9, the inventory exclusion provision, for these reasons:

[291] AHA argues that this clause is successfully invoked where both proof of the existence of the loss and its amount are dependent on an inventory computation. It says that proof of Pegasus’ claim against AHA, other than for the theft which Securitek captured on the surveillance tapes, is solely dependent on a comparison of inventory records to an actual physical count. And that this is an example of just the circumstances in which the exclusion was intended to apply.

[292] Pegasus does not rely solely upon the comparison of inventory records with physical count to prove either the factual existence of loss or its amount; rather it relies on that evidence to bolster its claim that theft occurred and to quantify the level of theft. There is ample other evidence already discussed that establishes that widespread theft of Pegasus’ stock was occurring. It follows that the proviso to the inventory computation exclusion clause applies. The inventory records are relied on as partial evidence only. Indemnity is therefore available.

[47] Ms Challis’s argument followed similar lines to Mr Ring, recognising of course the differences between the two policies. However, we agree with Winkelmann J’s reasoning. The plain wording of AHA’s inventory exclusion is decisive against its application.
[48] In this context, by virtue of s 9 of the Law Reform Act Pegasus is treated as the insured party (thus it is unnecessary to address the affect of the bailee’s liability extension in AHA’s policy). The operative clause indemnified Pegasus for its direct financial loss. Theft, fraud or dishonesty by an employee was the nominated peril or risk. AHA’s inventory exclusion simply excluded from cover loss which depended for its proof “solely” upon stock records. While AHA’s provision had the same effect as QBE’s policy in excluding a theoretical loss, it was explicit within the context of a policy designed to cover the peril of theft in excluding the inventory records if they constituted the only proof of loss. But here, as we have explained, Pegasus did not depend solely upon its stock records – it also relied on independent evidence of theft.
[49] Moreover, the effect of AHA’s inventory exclusion is further limited by its proviso. Inventory records are expressly admissible as “partial evidence” where an identified employee is involved in causing the loss. The words “partial evidence” simply mean that the inventory records are admissible if other evidence is available. The phrase does not imply, as Ms Challis appears to argue, a hierarchy requiring a Court to consider that other evidence first when determining a claim. It follows that we reject Ms Challis’s proposition that AHA’s proviso was not intended to provide an indemnity where the inventory comparison was submitted as proof of theft, but only where it was submitted as proof as the measure of loss – that is another distinction which cannot be sustained by the policy wording.
[50] In our judgment, Winkelmann J correctly found that AHA’s inventory exclusion did not apply.

(d) Consequences

[51] Our conclusion that Pegasus’ stock records are admissible to prove the fact of its loss and the amount affects the balance of our evaluation of the insurers’ appeal. That is because, as Winkelmann J found, only two possible explanations for the company’s stock shortages are available. One is inventory error, however caused; the other is physical loss through theft.
[52] Importantly, neither Mr Ring nor Mr Challis challenges this finding. Both accept that, if the Judge’s findings on the August 2004 and February 2005 stocktakes are upheld, with the result that Pegasus proves its loss of $354,369, then employee theft is the only possible explanation. Both say, however, that the Judge would not have embarked upon an analysis of the stock records if she had first enquired whether there was independent evidence from which she could be satisfied that Pegasus had suffered thefts on that scale. The essence of their complaint is that the Judge placed undue reliance upon the loss revealed by the stock records. In their submission, she impermissibly adopted that figure as proving all the items of stock stolen instead of considering separately whether the evidence separately established the theft of all the 93,457 items.
[53] In this respect, Mr Ring and Ms Challis challenge Winkelmann J’s findings on both the reliability and accuracy of Pegasus’ stock records and the nature and extent of employee theft, but in the reverse order. Mr Ring submits that the Judge started her inquiry in the wrong place, which led her, once she had found favourably on the stock records, to approach the theft question with a fixed mindset that 93,457 items were missing. The Judge should have begun, he submits, with the evidence of actual theft – and concluded that the opportunity for large scale theft did not prove its assessed actuality – before considering the stocktakes.
[54] This argument has a certain appeal in logic. That is because the normal order of inquiry is into the fact or existence of the insured peril, in this case loss by theft, before fixing its amount by reference to documentary material if available. However, the Judge’s approach was, we think, correct where the loss disclosed by the record – that is its timing and magnitude – is directly relevant to determining whether theft was its cause. The Judge gave appropriate context to this issue when observing that:

[70] It is critical to Pegasus’ case that the stocktakes in August [2004] and February 2005 and that undertaken by Monarch [in July 2005] were accurate. This is so because they are relied upon to quantify loss at the three points in time that they were undertaken. It is also true because if the stocktakes were significantly inaccurate, the inclusion of the data into Pegasus’ system could, as Mr Graham’s evidence for QBE emphasised, have destroyed the integrity of the information recorded there.

Stocktaking

(a) Process

[55] Pegasus began storing its stock at NZE’s warehouse in Mono Place, Ellerslie in 2001. Later, from about September 2004, the stock was stored at NZE’s warehouse in Donner Place, Mt Wellington. The goods were predominantly high volume and low value such as lunch boxes and drink bottles and were palletised in boxes or cartons. At the higher end of the value range, Pegasus had bicycles branded with fictional characters and sporting goods such as golf club sets and branded rugby and basketballs.
[56] Winkelmann J described the stock recording systems employed by each party as follows:

[18] The sole director and shareholder of Pegasus, Mr Kerrin Harrison gave evidence at trial as to how the stock and stock records of Pegasus were managed by NZE and Pegasus, including the processes used between Pegasus and NZE for taking, reconciling and recording the results of manual stock counts. Mr Harrison was the only witness of fact called to give evidence on the management of Pegasus’ stock records. He said that the process operating for storage of Pegasus goods by NZE was as follows: NZE would take delivery of Pegasus’ stock direct from Pegasus’ suppliers. NZE was provided with a copy of the purchase order, and checked the deliveries it received against those purchase orders. It then notified Pegasus of goods received. Pegasus also tracked inwards goods through shipping schedules and payments. By these means it attempted to ensure that any discrepancies between goods ordered, shipped, delivered and received would be found at the time of the inwards goods, and would be discussed with NZE and resolved. Goods purchased by Pegasus were not entered into Pegasus’ perpetual stock records (run with a software package called Prophet) until they had been received by NZE and any discrepancy between the amount received and the order and shipping documentation resolved, thereby ensuring the accuracy of Pegasus’ stock records at this point. The goods would then also be entered into NZE’s records.

[19] Goods were shipped direct from NZE to Pegasus’ customers. Pegasus would key the orders it received from customers into its Prophet system, export that information from Prophet as an electronic file and then send the information by email to NZE. NZE would import the electronic file into its stock system and produce the picking slips for the stock. Orders were then picked and packed by NZE staff and delivered to customers by a third party trucking company. On occasion Mr Harrison visited NZE and viewed its picking and packing process.

[20] NZE supplied Pegasus with a daily list of product sent to Pegasus’ customers. Pegasus marked the order as complete in its Prophet system, and generated an invoice for the customer. Stock levels recorded at both NZE and Pegasus were reduced by the amount of that order. If any customer did not receive the amount ordered or invoiced for, Mr Harrison said the customer would raise it with Pegasus and the issue would be taken up with NZE. That rarely happened before July 2004.

[57] Each month NZE sent Pegasus a list of the stock recorded in NZE’s system. Pegasus would match that list by product code to its own records and identify any discrepancies which, if they existed, were identified and settled by the time of the following months’ reconciliation. Any negative variances were invoiced to NZE.
[58] Mr John Cregten, a forensic accountant, was called by Pegasus to verify the reliability of its perpetual records and assess the level and value of its loss. The Judge accepted Mr Cregten’s methodology.[16] And she found that Pegasus’ records upon which his conclusion on the stocktakes was based were reliable and correct. She rejected in forthright terms opposing evidence from two QBE witnesses to the effect that Mr Cregten was wrong and that little if any stock was missing.[17]
[59] On appeal, the insurers do not now challenge Mr Cregten’s methodology. Nor do they rely on QBE’s witnesses who sought unsuccessfully to discredit the integrity of Mr Cregten’s opinion. Instead, they dispute the reliability of Pegasus’ perpetual inventory record; at some stage, they say, it became corrupted by the inclusion of inaccurate information as to stock levels supplied by NZE. On this ground, they say, Mr Cregten’s conclusion was wrong.
[60] In this respect, Mr Ring emphasises that:

• (a) it is fundamental to the reliability of each physical stocktake (accepted by Winkelmann J at [70]), and every successive stocktake based upon it, that the reason for any variances between the physical stocktake and the perpetual records are identified and reconciled so both agree; and it is axiomatic that agreement and reconciliation are not simply achieved by adjusting the perpetual records to match the physical stocktake – because this assumes that only the perpetual records are inaccurate and that the extent of this inaccuracy is the exact amount of the required adjustment;
• (b) the key prerequisites to a reliable system are accurately recording stock movements in and out and accurate periodic counts, with the result that, if NZE’s physical count was inaccurate and its perpetual system was then adjusted to match, NZE’s perpetual system would also become inaccurate; and, vice versa, if Pegasus’ perpetual system was then reconciled to match NZE’s perpetual system.

[61] In Mr Cregten’s opinion, July 2003 was a reliable starting point for the stocktaking reconciliation process. At that time NZE’s stocktake was reconciled to Pegasus’ perpetual system to a variance of less than $1,000 involving a mismatch of only four inventory items.[18] However, there were a growing number of variances in the ensuing year. Both parties investigated and were largely able to reconcile them.[19]

(b) The August 2004 stocktake

(i) High Court

[62] The first major discrepancies were noted in stocktakes in mid-2004. As Winkelmann J described, they followed a reconciliation process:

[27] ... in a stocktake undertaken by NZE on 9 June 2004. That showed a shortfall in stock with a total value, calculated by Pegasus, of $69,720.33. Pegasus queried this result because of the size of the discrepancy. NZE responded that they were very busy and that therefore the count might not be accurate. Pegasus initially processed stock adjustments in relation to the June 2004 stocktake so that the stock in its inventory system and what had been counted, balanced. It withheld payment from NZE equal to the value of the missing stock, and requested a second count. When NZE said that it would do another stocktake, Pegasus reversed the initial adjustments so that there would be no double adjustment and consequent distortion of its perpetual record. Again, it is at issue whether NZE also reversed out the entries to its stock records that were consequent on this stocktake.

[28] About two months later, around August 2004, NZE completed its second stocktake, and a similar level of discrepancies to those in the June 2004 stock count were found. Again NZE said that the count might not have been accurate, but it reassured Pegasus that it had rented a new building, that the stock was to be shifted to that building and that a new stock count would be completed after the shift.

[63] Mr Cregten’s conclusion on the August 2004 stocktake was summarised by the Judge as follows:

[46] Mr Cregten reviewed the June and August 2004 stocktakes, noting that the discrepancy identified by Pegasus following the August stocktake was valued by it at $66,024.53 plus GST, for which it invoiced NZE. He said that to the extent that stock may have been missing earlier than August 2004, that was captured at that point and through the invoicing system the systems were reconciled so that the physical stock level and Pegasus’ and NZE’s perpetual systems were all in balance. Mr Cregten’s opinion was that the August invoices issued by Pegasus after the August 2004 stocktake were a reasonable basis for establishing the initial quantification of inventory missing at that time. However, he identified a need to make adjustments to those. He considered an adjustment necessary to the unit value recorded in Pegasus’ system that had been utilised in the invoices, and also to the number of items claimed. He said that the number of items included as missing in these invoices had to be reduced by those that were found in subsequent counts, and those that were identified by him in an audit he undertook.

...

[57] a) August 2004: Pegasus claimed 18,346 items missing. Mr Cregten’s investigations subsequently identified that 5,367 of those items had been found. Value attributed to those items by Mr Cregten: $37,993 ex-GST (to be contrasted to a system generated value of $43,093 ex- GST).

[64] The Judge described the nature of the insurers’ challenge to Mr Cregten’s opinion at trial in this way:

[59] QBE and AHA argue that the key flaw in Mr Cregten’s analysis is that he assumes Pegasus’ perpetual record is accurate. They argue that reliance is misplaced in light of evidence that the perpetual record was distorted by the inclusion of defective data collected by NZE on completion by it of grossly inadequate physical stocktakes, sloppy inputting and a “glitch” in NZE’s system. QBE and AHA’s criticisms of the accuracy of the information recorded were summarised by counsel for AHA as follows:

1. NZE had inadequate and deficient stock control procedures generally, with significant potential for error in all areas of NZE’s operations.
2. There was a glitch in the NZE computer system which meant that some information was not captured.

100. When the monthly stock lists were supplied by NZE to Pegasus, Pegasus simply imported that information into its own records, without any attempt to reconcile it.

4. The stocktake process was fundamentally flawed, and the movement of stock from Mono to Donner Place exacerbated existing problems.
5. Stock was left behind at Mono Place, and is not missing.
6. Although there were problems with short deliveries, it is more likely these arose through stock being incorrectly picked/packed than through theft, given evidence that stock was found at the warehouse following short shipments being reported. Pegasus’ records are not sufficient to prove that the items were missing rather than simply not delivered or not recorded by Pegasus as delivered.
7. Poorly controlled stock movements following liquidation mean that it is not now possible to rely on a comparison of Pegasus’ and Monarch’s records of stock on hand to calculate what if any stock is missing following the February 2005 stocktake.

(ii) QBE’s case

[65] Winkelmann J rejected those challenges.[20] The insurers have not attempted to repeat most them on appeal. Instead, Mr Ring focuses on the first and third (set out at [64] (a) and (c) of the judgment) in support of a submission that the Judge erred on the August 2004 stocktake in two specific respects. First, he submits that NZE never properly reconciled the variances between the June 2004 stocktake results and its perpetual records; rather, he says, NZE simply adjusted its perpetual records to match the physical count.
[66] Mr Ring relies on these passages from Mr Harrison’s evidence-in-chief:

41. A major problem first appeared in mid 2004. NZE did a full stocktake on 9 June 2004. The result showed that there was $69,720.33 worth of stock that was recorded as being in NZE’s warehouse but that was not there. After the first stocktake was sent to us by NZE we obviously were not happy about it and questioned the result. NZE advised us that they were very busy and therefore the count may not be accurate. As a precaution we withheld payment from NZE equal to the value of missing stock and requested a 2nd count.

42. In order to ensure that Pegasus’ records matched to NZE’s, NZE made a stock adjustment so that the stock balanced to what they had counted. We also changed the amount of stock recorded in Pegasus’ system at that time to match the adjustment made by NZE. We effectively removed the missing stock from Pegasus’ system after the stocktake was done.

43. Initially NZE thought that the first stock count was short because they had not been able to locate all of the stock in the warehouse. They advised they were going to do another stocktake. We wanted to ensure that no double adjustments were made so we reversed the adjustment that had been made after the 9 June 2004 stocktake before the start of the 2nd stocktake.

44. NZE completed a 2nd physical count about 2 months later and a similar number of discrepancies were found. A substantial amount of stock was identified as missing. ... Pegasus did another adjustment for the 2nd physical count in its system and invoiced NZE for the missing stock. ...

[67] We do not read Mr Harrison’s evidence as supporting Mr Ring’s submission that NZE failed to reconcile its stock figures with its own perpetual records, making them unreliable. Mr Harrison’s account was to a different effect. It was that NZE carried out a physical stocktake in June 2004 which disclosed a significant discrepancy. Pegasus questioned the result but as an interim measure adjusted its records to reconcile with NZE’s stocktake. That adjustment was reversed out prior to the second or verifying stocktake in August.
[68] Second, Mr Ring submits that the Judge erred because there was no evidence from any NZE staff member based upon direct observations as to the quality of the August 2004 stocktake.[21] He says the Judge gave insufficient weight to evidence of Suzanne Hill, NZE’s former Auckland business manager. Ms Hill commenced employment with the company in October 2004. She was engaged to oversee both the Mono Place and Donner Place warehouses. By then, Pegasus’ stock had been transferred from Mono Place to Donner Place. On arrival at NZE Ms Hill found “lots of potential for error in all areas, inwards, outwards and daily operations” and no proper systems in place for undertaking accurate stocktakes. Her impression was that little attempt had been made by NZE to carry out reconciliations of stock following stocktakes.
[69] In support of Mr Ring’s submission, Ms Challis refers to Ms Hill’s specific discoveries that in October 2004 NZE lacked systems for taking accurate stocktakes, did not freeze stock during stocktakes, did not reconcile or investigate variances and simply reported variances back to customers. In summary, according to Ms Hill, NZE’s stocktake situation at the end of 2004 was chaotic, with poor stock controls.
[70] However, we are satisfied that Winkelmann J had a proper foundation for finding that Ms Hill’s evidence did not assist on the processes followed in the August 2004 stocktake, some two months before her arrival.[22] By October, the Judge noted, NZE’s warehousing operations were spread over two warehouses following the shift in September. Both operations had been subjected to substantial disruption caused by the shift of stock and staff between two centres. While Ms Hill’s evidence of systemic deficiencies at the Donner Place warehouse in October carries weight, her impression of what may or may not have happened at a stocktake carried out two months earlier, at a different warehouse, and for the specific purpose of ascertaining whether goods were missing is of limited evidential value. The Judge’s finding was also supported by Mr Harrison’s evidence of reconciling the 2004 stocktakes.
[71] The absence of an ex-NZE employee who physically observed the August 2004 stocktake is not fatal, even material. The exercise was conducted against a background of dispute and Pegasus’ dissatisfaction with the June results. It is open to inference that both sides were anxious to ensure its accuracy; the financial consequence for NZE of confirming the June results would be a compensatory adjustment to Pegasus and, for its part, Pegasus would be concerned to ensure that it did not incur loss through theft of its stock. As Mr Tingey submits, NZE was contractually bound to pay for any missing stock.

(iii) AHA’s case

[72] Ms Challis makes additional challenges to the Judge’s finding on the August 2004 stocktake. On analysis, they also fall into two grounds, that the Judge:

• (a) wrongly found that the earlier stocktakes had required only a slight adjustment to Pegasus’ perpetual record, amounting to good evidence that the record was largely accurate,[23] when there was evidence of problems with NZE’s count and stock control systems before that date. In particular, Ms Challis relies on Mr Harrison’s acknowledgment that in March 2004 Pegasus sent a report showing miscounting of significant numbers of product lines;
• (b) erred in relying on Mr Harrison’s evidence that the June and August 2004 stocktake discrepancies were of a similar order. Ms Challis relies on Mr Harrison’s account of conducting random checks of Pegasus’ stock at NZE’s warehouse after June 2004 when he found most of the items he was seeking. But he was told by a NZE employee that the company was recording errors in its stock count. As a result, the parties agreed to reverse it and repeat the count.

[73] We agree with Mr Tingey that these grounds are unsustainable. Mr Harrison’s evidence was that the June 2004 stocktake was reversed because neither party was satisfied of its accuracy. Pegasus’ claim was based upon the August 2004 stocktake. The fact that discrepancies revealed in the June and August 2004 stocktakes were of a similar order is not inconsistent with a finding that NZE’s stocktakes were accurate. The Judge had concluded that stocktakes before 2004 only required a slight adjustment to Pegasus’ perpetual record. The reversal of the June interim reconciliation does not undermine this finding.
[74] More importantly, there is no direct evidence that NZE’s stocktake in August 2004 was wrong or erroneous. Mr Harrison’s acceptance that, if NZE’s processes in August 2004 were wrong then there would be discrepancies in that month’s stocktake, did not assist AHA. Agreement with a hypothesis does not prove the hypothesis itself.

(c) The February 2005 stocktake

(i) High Court

[75] The next significant stocktake was in February 2005. Winkelmann J outlined events leading up to it, and the stocktake itself, as follows:

[29] NZE moved into its new building in Donner Place in September 2004. Mr Harrison anticipated that the stock would be stored correctly in NZE’s new building, it would then be counted, and discrepancies resolved. But he said that between September and December 2004 there were a large number of short deliveries to customers where both NZE’s and Pegasus’ perpetual records indicated the items were in stock. Pegasus progressively issued invoices for the short deliveries to NZE up to April with a total value of $48,295 plus GST. Corresponding reductions in the stock levels for the invoiced items were processed in Pegasus’ stock inventory.

[30] In February 2005 NZE undertook another stocktake. On this occasion a very substantial negative variance was identified. Pegasus valued the amount of missing stock identified on this stocktake at $304,525.43. Mr Harrison said that discussions took place between NZE and Pegasus in an attempt to resolve the discrepancies on a product by product basis. The discrepancies were not resolved. Ultimately NZE promised that it would reimburse Pegasus for the shortfall in the stock, but went into liquidation in early May without having done so.

[76] Later, the Judge referred in more detail to Mr Cregten’s evidence, which she accepted:

[57] The results of these investigations and his subsequent analysis can be summarised as follows:

1. August 2004: Pegasus claimed 18,346 items missing. Mr Cregten’s investigations subsequently identified that 5,367 of those items had been found. Value attributed to those items by Mr Cregten: $37,993 ex-GST (to be contrasted to a system generated value of $43,093 ex-GST).
2. February 2005: Pegasus claimed 45,111 items missing. Mr Cregten’s investigations identified that 13,265 of those items had been found. Value attributed to that stock: $136,559 ex-GST (to be contrasted to a system generated value of $168,106)

100. Short supply claim: Pegasus claimed 4,235 items missing. Mr Cregten’s investigations identified that 862 of those items had been found. Value attributed to the missing stock: $24,239 ex-GST (to be contrasted to a system generated value of $38,820 ex-GST).

4. July 2005: Pegasus claims 50,214 missing items. Mr Cregten’s investigations identified that 4,955 of those items had been found. Value attributed to those items: $155,578 ex-GST (to be contrasted to the system generated value of $179,458 ex-GST).

[77] At this point we note the Judge’s reference to a stocktake in July 2005. That was undertaken progressively from May by Monarch Ltd which had taken over NZE’s operations at Donner Place. At trial and on appeal the insurers have not taken issue with the accuracy of that stocktake, which revealed further losses, or Mr Cregten’s conclusion upon it and we do not need to refer to the issue further.
[78] Ms Hill was Pegasus’ critical witness on the February 2005 stocktake. The Judge summarised her evidence-in-chief in this way:

[78] Ms Hill agreed that the records of the stock transfer were not accurate and that some stock was left behind at Mono Place. But Ms Hill’s evidence was that when she started at NZE she implemented systems to ensure the accuracy of stocktakes undertaken. In particular she took the following steps:

1. Ensured that rolling stocktakes were carried out on a frequent basis at regular intervals. Those stocktakes were done on a daily or perhaps weekly basis, depending on the situation. For example, if a particular customer raised a query about their stock, then NZE might carry out a stocktake on a particular line of that customer’s stock, or all of it.
2. Put in place policies to ensure that proper reconciliations of stock were carried out as part of any full stocktake process, rather than any discrepancies simply being reported back to NZE’s customers. As part of that, she ensured that stock was frozen during the stocktake process so that proper re-checks could be carried out.

100. Ensured that NZE’s employees used proper count sheets when they were carrying out stocktakes.

[79] In relation to the February stock count, her evidence in chief was that the count and subsequent reconciliation were undertaken using the new policies and procedures she had implemented. Those processes took approximately five days to complete and Pegasus’ stock was frozen during that time. She said that it was only after the count and reconciliation was done that adjustments were made to NZE’s system. It was also at that point in time, having eliminated other explanations for the discrepancies, that she considered the possibility that stock may have been stolen.

(ii) Insurers’ case

[79] Mr Ring and Ms Challis submit that Winkelmann J erred in relying on Ms Hill’s evidence. But Mr Ring accepts three important points. First, QBE’s challenge to the Judge’s finding on the February 2005 stocktake stands or falls on Ms Hill’s evidence. Second, Ms Hill’s account, as summarised by the Judge, is of conducting a properly reconciled stocktake. Third, Ms Hill’s evidence on oath as to the detail and manner of the stocktaking process was not challenged by QBE’s counsel (not Mr Ring) at trial. Rather, as we shall explore, QBE’s focus at trial was on establishing that an earlier unsworn statement was correct.
[80] Ms Hill’s previous written statement was made to QBE’s investigator, Mr John McCallum, on 10 July 2006. Mr Ring seeks to exploit what he submits are inconsistencies between Ms Hill’s statement and what she said at trial. In her statement to Mr McCallum, Ms Hill said, among other things, that:

There was never a reconciled stocktake done on Pegasus goods during my time at NZE. Those stock figures were just never reconciled or investigated. What was on paper and what was on the shelf was never verified. Every week NZE supplied an electronic list to Pegasus and [Mr Harrison] just altered the stock figures accordingly. He altered his physical inventory every week on his computer system to match NZE.

[81] Later in the same statement Ms Hill said this:

[Mr Harrison] didn’t want an independent in there. He wanted NZE to do the stockcount and he didn’t want to stop trading long enough to have it done. However, we did convince him to stop trading and the stocktake did take place in about February [2005] but once again it was done by NZE and not fully reconciled.

[82] At trial Ms Hill was taxed by counsel for both insurers on the inconsistencies between her evidence-in-chief and her statement to Mr McCallum. Her credibility was squarely in issue. The Judge found that:

[85] There is an apparent inconsistency between the statements recorded as made by Ms Hill to the private investigator and the evidence that she gave in Court in relation to whether a fully reconciled stocktake took place in February 2005 and whether theft was a possible cause of the stock losses. As a prior inconsistent statement, the statement to the private investigator is relevant in terms of assessing the reliability of Ms Hill’s evidence and her credibility, and also as part of the narrative behind this issue.

[86] When this interview was put to her during the course of cross-examination Ms Hill said that the evidence that she gave in Court was correct. She explained the inconsistency on the basis that she put little thought into the answers she gave to the investigator’s questions. She said that she was under a lot of pressure at that time; she was starting a new job and still had a lot of overhang from the liquidation, including pressure from previous staff and from various different agencies wanting her to talk about it. She said that she simply wanted to “get him out of there”. When the investigator came back with the typed up answers she assumed that he had recorded things correctly and did not read the answers with care, although acknowledging that she did make handwritten corrections to spelling mistakes.

[87] I accept Ms Hill’s evidence that a reconciled stocktake did take place in February 2005. In preferring the evidence she gave in Court I have taken into account the fact that she has no particular interest in the outcome of this litigation. There is no suggestion that she is associated with Mr Harrison in any way. There is therefore no motivation for her to give evidence favourable to Pegasus. Nor is there any suggestion that she has a grievance against the defendants, NZE, or NZE’s liquidators.

[88] I have also taken into account that the private investigator’s statement is not a verbatim account of what Ms Hill said. It is based on handwritten notes prepared by the private investigator and then transcribed by him and signed by Ms Hill. Ms Hill was being asked by the investigator to address and explore a complex series of events; events she likely felt defensive about at that time. She had after all only recently surrendered responsibility for the running of the warehouse from where the stock was said to have been stolen. Moreover, she dealt with the query at her work, where she could be expected to (and said she did) feel under time pressure, and be distracted by competing demands on her.

[89] I accept Ms Hill’s account that she did not approach the preparation of this statement with care and that some of what she said was not accurately recorded. Although spelling was easy to correct, it would have taken much longer to ensure the true statement recorded a full and accurate record of events affecting Pegasus’ stock, and she did not take that time. I contrast the approach she said she took in the preparation of that statement to the way in which she gave her evidence. In her evidence she gave thoughtful, thorough answers, and impressed me as a fair and careful witness.

[83] Mr Tingey’s submission that an appellant must establish compelling grounds to upset the Judge’s findings on Ms Hill’s credibility overstates the position.[24] This Court must review the evidence anew and form its own opinion on disputed facts independently of the trial Judge exercising however the customary caution where a credibility finding is in issue. We are conscious that within the setting of a tightly contested trial the Judge is constantly assimilating and evaluating evidence as it unfolds gradually under examination. That factor does not of course immunise a Judge’s finding from challenge. But the deliberative process, and the well accepted benefit enjoyed by a trial Judge in seeing and hearing a witness, cannot be replicated within the confines of an appellate hearing. We must give appropriate weight to the trial Judge’s assessment of a witness whose credibility was under attack.
[84] Mr Ring properly acknowledges this point. But, he says, judicial impression was only one element in the equation of acceptance of Ms Hill’s evidence. The other was Ms Hill’s explanation for the inconsistency between what she said to Mr McCallum and what she said at trial. The question, he says, is whether Ms Hill’s statement to Mr McCallum materially undermined the reliability of her evidence at trial (we disagree with Mr Ring’s categorisation of reliability; we are satisfied that Ms Hill’s credibility was in issue). In this respect, Mr Ring submits, what Ms Hill said on oath is not entitled to primacy for that reason alone.
[85] It is appropriate, in our judgment, to address this question by analysing the nature and degree of the allegedly decisive inconsistency. Ms Hill referred twice to specific stocktakes in a five page statement devoted principally to whether widespread theft had occurred at NZE’s warehouse. The first reference was open ended – to the absence of a reconciled stocktake during Ms Hill’s time at NZE; that is after October 2004. What she said had no relevance to the August 2004 stocktake. She was apparently describing Mr Harrison’s alleged practice of altering his monthly records without inquiry to match NZE’s. However, Mr Harrison was never cross-examined on this assertion. And the fact that the February 2005 stocktake disclosed such a substantial discrepancy between the records of the two companies suggests the practice described by Ms Hill did not occur.
[86] The second reference is specifically to the February 2005 stocktake. It is, in fact, substantially consistent with Ms Hill’s evidence-in-chief.[25] The only difference is a concluding reference to the stocktake done by NZE, that is by Ms Hill herself, not being fully reconciled. Ms Hill was not cross-examined on what she meant by the phrase “fully reconciled”. But some insight is available from this exchange with QBE’s counsel:

Q. If Mr Harrison’s file note is an accurate reflection of what occurred at the meeting would that have reflected poorly on the February 05 stocktake?

A. Not necessarily. The reason for that meeting [which occurred between Mr Harrison and Pegasus representatives in January 2005] was that we were still trying to reconcile to find the reasons for the discrepancies that were showing up in that stocktaking.

(Emphasis added.)

[87] It is apparent that Ms Hill was using the word “reconcile” as synonymous with finding reasons for discrepancies showing up in a stocktake, which is but one of the steps in the entire process of identifying and then balancing the differences or variances between physical stocktakes and perpetual records.[26]
[88] In cross-examination Ms Hill agreed that both references in the statement were “absolutely inconsistent” with her evidence at trial. However, this concession was itself inaccurate and does not assist QBE. On analysis, the second reference was not absolutely inconsistent with Ms Hill’s evidence on oath; it varied only to the extent of conceding there was no full reconciliation at the February 2005 stocktake. Otherwise it does not throw into doubt the accuracy or veracity of her full account of that stocktake given in evidence-in-chief.
[89] We have independently reviewed Ms Hill’s evidence. We are satisfied that, for the reasons she gave, Winkelmann J had a proper basis for preferring Ms Hill’s evidence on oath to a statement made to QBE’s investigator, to the extent that the two were inconsistent and where Ms Hill had given an explanation for that inconsistency. In this respect, we repeat that Ms Hill’s evidence of what Mr Ring accepts was of a properly reconciled stocktake was not directly challenged in cross-examination. In its absence, the Judge was entitled to prefer Ms Hill’s account given on oath to statements made to QBE’s investigator, to the extent that they materially differed.
[90] We add that the February 2005 stocktake was carried out against a background of Mr Harrison’s complaints of stock shortages. NZE’s position, in Ms Hill’s letter dated 20 January, was that the discrepancies were theoretical, not actual, and that a “thorough stocktake” would answer Pegasus’ concerns. A later NZE letter dated 2 February 2005 admitted that earlier stock discrepancies had resulted in Pegasus’ inability to meet supply commitments to customers prior to Christmas 2004. Mr Harrison’s evidence at trial confirmed that the February 2005 stocktake was carried out during which time the company froze its stock.
[91] Moreover, as Mr Tingey submits, NZE apparently admitted that Pegasus’ stock may have gone missing while in its custody. In a letter dated 29 March 2005 signed by Ms Hill, the company acknowledged goods to a value of $304,737 were unaccounted for as a result of the February stocktake, consistently with NZE’s participation in the agreed result of the February reconciliation
[92] In summary, we are not satisfied that the Judge erred in finding that Pegasus’ records of the August 2004 and February 2005 stocktakes were reliable and correct. It follows that she did not err in her consequential finding that 93,457 items of Pegasus’ goods, to a total value of $354,369, were missing by July 2005 when Monarch completed its stocktake.

Theft

[93] On the question of theft, Winkelmann J concluded as follows:

[155] Given the numbers of staff involved in the theft, the apparently open way in which stock was being stolen, and the rate of theft in the small sample of events within the warehouse captured on the Securitek footage as well as that described by NZE staff, I am satisfied that it is more likely than not that theft of Pegasus’ stock occurred frequently, in large volumes and over an extended period of time. That this is so is corroborated to some extent by the theft of Anthony Trading product.

[94] Both Mr Ring and Ms Challis have subjected this finding to intense scrutiny. Mr Ring accepts that the evidence established, on the balance of probabilities, thefts by a number of NZE employees. But he says that the Judge erred in:

• (a) finding that the thefts occurred frequently, in large volumes and over an extended period of time; and
• (b) failing as a consequence to consider whether the evidence established to the requisite standard the theft of all the 93,457 items – instead she impermissibly adopted the loss or shortages revealed by the stocktakes as proving all items were stolen.

[95] Mr Ring relies on statistics to give what he says is a sense of perspective. He says the only direct evidence of theft was during a short period of covert surveillance in early 2005. Based on an average value per item, the goods lost in that period represented about 109 separate items to a value of $413. However, he says, on the Judge’s findings NZE employees stole 93,457 items to a total value of $354,369 between May 2004 and May 2005. This represented an average of 256 items per day on each day for 365 days; or $971 worth of items on each day at an item value stolen of $3.79.
[96] Mr Ring’s purpose is to establish that, while there was evidence of direct theft, it was not and could not have been on the scale or magnitude actually found by the Judge. What the surveillance cameras showed was regular but petty theft; what was occurring, on the Judge’s finding, was regular but major theft. In support, Mr Ring’s submissions traversed a great deal of ground, some of it overlapping, and we have attempted to distil its essence and address what appeared to be his main arguments. They fall into four categories: (1) the video surveillance evidence and the inferences available; (2) statements by former NZE employees; (3) an absence of evidence relating to Mono Place; and (4) short deliveries.
[97] First, Mr Ring submits that the Judge’s key factual findings on theft were based solely on drawing inferences from the surveillance footage of direct stealing by NZE employees in a two and a half week period at Donner Place in March and April 2005. Securitek placed a covert camera in the warehouse which was focused on the roller door. Another camera was placed nearby on the outside to show where the goods then went. Mr Ring emphasises that the roller door at Donner Place was the main entrance where the trucks arrived and departed. The other camera panned towards the staff car park, because employees were most likely to use their vehicles to remove goods.
[98] The surveillance revealed that about half of NZE’s employees engaged in theft on one or more occasions but only four employees took Pegasus’ stock, each on separate days, to a total value of $413. Mr Ring says that the Judge correctly characterised the theft as occurring on a minor scale – on an annual basis it would have only amounted to about $11,000 – and that Pegasus’ stock was principally large quantities of small value items. The necessary consequence of the Judge’s finding was, he says, that a huge volume of product would have been removed from the warehouse without detection – to a value of $1,177 daily for 365 days.
[99] In particular, Mr Ring submits that the Judge’s rationale for the discrepancy – that the camera showed only one of eight exits for just over two weeks out of 52 weeks – was flawed; and there was no logical basis for assuming that, because one camera covering one exit showed a certain level of thefts in a given period, eight cameras covering all exits for the same period would have revealed an exponentially higher level of theft. These alleged errors, Mr Ring says, illustrate the Judge’s underlying error of equating evidence of actual employee thefts of small quantities of goods plus opportunities for stealing on a grand scale with a conclusion that there were defalcations on that same scale.
[100] In support, Ms Challis seeks to impose further limitations on the statistical scope for averaging. She says that the implausibility of Pegasus’ claim is compounded when three periods within the relevant year are excluded. One is the eight weeks between June to August 2004, given that the similar discrepancies between the two stocktakes suggests little or nothing was stolen in this period. Another is in the month prior to 23 March 2005 when Ms Hill conducted random checks on the Donner Place warehouse which revealed nothing suspicious. Finally, the two and a half weeks of Securitek surveillance in March and April 2005 that revealed the theft of only $413 worth of Pegasus’ stock must be added. On this analysis, the relevant period is compressed from 12 to eight months.
[101] We do not accept this argument. The averaging approach adopted by Mr Ring and Ms Challis is artificial and unreal. Winkelmann J acknowledged that Pegasus’ case required acceptance of the proposition that one or more unknown employees unlawfully took large quantities of stock to unknown destinations on a regular, probably daily, basis without detection, necessarily involving supply to another party at a commercial level.[27] She was satisfied that there was an opportunity for theft on that scale,[28] and that, in particular, there was a one off opportunity for large scale theft generated by the shift of stock from Mono Place to Donner Place.[29]
[102] We add, in answer to Ms Challis’s specific submission, that the fact that similar discrepancies were disclosed in the June and August 2004 stocktakes does not eliminate the opportunity for theft to a value of $37,993 from Mono Place before June 2004 (Ms Challis submits elsewhere that the Judge erred in finding that the discrepancies were of a similar order). Furthermore, the accuracy and reliability of the June 2004 stocktake was not in direct issue. And the fact that Ms Hill, when conducting random checks, found nothing suspicious, or that one camera covering one of eight exits revealed theft of $413 worth of stock, does not operate to limit the nature and extent of NZE employees’ thefts in that period.
[103] In our judgment, Winkelmann J had a proper foundation for concluding that, apart from showing “relatively unrestrained and widespread theft”, the camera at Donner Place did not convey the complete picture.[30] The existence of a substantial discrepancy between what was proved by direct evidence to be stolen averaged across one year and what was actually stolen is not decisive.
[104] Second, Mr Ring criticises Winkelmann J’s reliance on statements made by former NZE employees. In particular, Mr Ring challenges this finding:[31]

One of the supervisors who gave evidence, who also admitted to theft, said that he saw other staff members stealing stock on at least two out of five days every working week, he saw employees walking out with cartons full of stock and saw one employee back his car up to the warehouse and spend 20 minutes filling it with stock. The same employee admitted he was caught selling Pegasus stock at a garage sale, but he said “[e]verybody was guilty of theft”. He said it was a “free for all”, and that he knew if anything was taken he would never get caught.

[105] Mr Ring says that Winkelmann J’s finding does not represent a fair and accurate summary of the evidence given by the unnamed supervisor. In particular that the supervisor did not witness stealing most of the time, but perhaps on two out of five days a week over a period of about eight months; that he did not say anything about volume or value in this context; and that he did not identify the stolen goods as belonging to Pegasus but, more likely by inference, as being Mobil’s stock.
[106] We reject this submission. The supervisor’s account added significantly to the factual foundation laid by the surveillance footage of frequent thefts by NZE employees. The supervisor himself admitted his participation in the offending; he also implicated all the other 13 employees, describing it as a “free for all”.
[107] Nevertheless, Mr Ring says this was not a balanced picture because two of NZE’s managers, who were on site for long periods, were unaware of any thefts. Mr Ring accepts, however, that neither manager was there for two or three hours in the evening, and one was absent from Donner Place for large periods. Also, he says that Ms Hill who before March 2005 suspected thefts on the night shift did not discover any evidence to this effect during random checks on three out of five nights a week over a monthly period. But, we note, the Judge did not place any weight on this evidence given her conclusion that Ms Hill would not have detected anything but the crudest and most blatant attempts at theft.[32]
[108] We are satisfied that the Judge had a proper evidential basis for these findings:

[138] There is a significant volume of evidence to suggest that there was opportunity for theft on the scale that Pegasus alleges. Apart from the two week period during which Securitek operated surveillance cameras, there were no security cameras on site. Nor was there security on the main gates to the Donner Place warehouse to check the identity of vehicles going into and out of the premises. The inwards goods door at the Donner Place warehouse was left open all day. That was the door furthermost from warehouse management and could not be monitored from where management sat. This meant that it would have been possible for a vehicle to drive around to the inwards goods area, load product out of the warehouse and drive away without being observed by management.

[139] Many employees had out of hours access to the warehouse, and the code for the alarm on site. There was evidence that without any proper reason for being there, staff members were frequently at the NZE warehouse outside of work hours. Staff members were found at the warehouse during weekends and before the warehouse opened on weekdays.

[140] There was no supervisor on site at all for the last three hours that staff were working at the warehouse. The principal supervisor at the site, Mr Gentry, divided his time between the Mono Place and Donner Place warehouses so that he was often absent from the Donner Place warehouse. When he was away from the warehouse, he left two others as acting supervisors. The evidence suggests that those two others were involved in the theft that was going on.

...

[144] There was also evidence of deficiencies in the consignment system which provided an ongoing opportunity for large scale theft. Ms Hill said that key gaps in the system at NZE included failures on the part of NZE to reconcile purchase orders from Pegasus’ customers with delivery consignment notes and consignment notes with invoices received from various trucking companies. This meant that it would have been possible for an NZE employee to write out a manual consignment note and have stock from NZE’s warehouse sent to where the employee wanted, but to not enter that consignment into NZE’s system. If there was no purchase order for that stock, then NZE’s system would have shown that the stock was still in a warehouse. It is true that there is no evidence of this occurring on any particular occasion, but it is nevertheless evidence that there were methods by which large volumes of stock could feasibly be removed from the site.

[109] Winkelmann J relied on additional evidence. For example, there was uncontested evidence that $120,000 worth of stock belonging to Anthony Trading was stolen from NZE’s Donner Place warehouse about the same time. The Judge noted that some of Anthony’s goods were large and would have required vehicles for removal.[33] Mr Ring criticises this observation, saying it was entirely speculative to infer that NZE employees used anything more than their vehicles to remove Pegasus’ stock.
[110] However, in our judgment the fact that utilities, vans or trucks were used to remove one customer’s goods was a sufficient evidential foundation for inferring that the same mode of transport was probably used to remove Pegasus’ goods. And there was evidence that staff members adopted a practice of throwing goods over the back fences of property for collection later in the day. Mr Ring attempted to minimise this, saying that it related to seven or eight cartons of toys. But minimising the proven scale of the practice does not diminish its existence.
[111] Viewed in totality, the evidence given at trial and accepted by the Judge provided a sufficient evidential foundation for her conclusion that NZE’s staff were stealing Pegasus’ stock frequently, in large volumes and over an extended period of time.[34]
[112] Third, Mr Ring submits that Pegasus led no specific evidence of thefts of its own stock or of the opportunity for such thefts at NZE’s Mono Place warehouse before the shift to Donner Place in September 2004. He points out that the claim dates back to losses commencing in May 2004 and the judgment must have allowed for theft from Mono Place because it included the loss assessed by Mr Cregten of 12,979 items valued at $37,993 at the August 2004 stocktake. Also, some of Pegasus’ stock was still at Mono Place during part of the period covered by the February 2005 stocktake. Mr Ring says there is insufficient evidence to support a finding of theft from Mono Place, which could only have been based on inference from the later actual thefts at Donner Place.
[113] We disagree. We repeat that, once the Judge found the stock records were accurate, the only available inference was that the missing property was stolen. Here it was augmented by the later surveillance footage of the regular thefts from Donner Place and the former NZE employees who testified at trial. Both sources provided a factual foundation for an inference that Pegasus’ goods were stolen from Mono Place before the scale of thefts intensified at Donner Place.
[114] In this respect, Ms Challis emphasises the evidence of Mr Greg Murphy. He managed the Donner Place warehouse until December 2004 and referred to significant deficiencies in NZE’s transfer of Pegasus’ stock from Mono Place. He conveyed an overwhelming impression of chaos, where in particular Pegasus’ stock arrived before the new warehouse was ready to function, no steps were ever taken to check the stock counted against stock which it left at the Mono Place warehouse, no full stock after the transfer was ever carried out, and stock was stored in make shift aisles on the floor. However, in our judgment all this served to confirm the Judge’s finding of an environment where employee dishonesty was the acceptable norm and where the particular circumstances of the transfer presented an open invitation to large scale theft by diversion of Pegasus’ property.
[115] Fourth, Mr Ring challenges Winkelmann J’s finding that within Pegasus’ claim for losses were 3,373 items valued at $24,239 plus GST which the company said were stolen by NZE employees by short deliveries to its customers during the period between September and December 2004.[35]
[116] Mr Ring says there was no direct evidence that any NZE employee actually under filled any consignment, and, particularly, no direct or surveillance evidence. He says Mr Harrison was the only witness to this effect, stating that “short deliveries to customers were experienced”. At best, he says, this amounted to no more than an inference that complaints had been made of short deliveries on 33 occasions. He says the evidence was of a single if not double hearsay nature, and was clearly inadmissible, even though QBE did not raise this point at trial.
[117] Mr Tingey takes the technical point that this issue was not raised in QBE’s notice of appeal. In any case, this ground of appeal fails because no challenge was taken to Mr Harrison’s evidence at trial. While it may have been of a hearsay nature, it was admissible in the absence of an objection. Furthermore, the existence of short deliveries was corroborated by Pegasus’ invoices to NZE for the short deliveries in response to customer complaints. The company was unlikely to have taken that step unless satisfied there were short deliveries.

Conclusion

[118] We accept that the Judge’s conclusion that Pegasus lost goods to a value of $354,369 as a result of large scale thefts by NZE employees over a one year period is not based upon direct evidence.
[119] However, a solid factual foundation for that inferential finding emerges from the combination of evidence adduced at trial. In particular we refer to the actual loss of 93,457 items (once the stock records were accepted as correct and reliable); the images captured by the surveillance cameras installed in March and April 2005; the pervasive culture of dishonesty among NZE employees; the theft of valuable goods belonging to another NZE customer at the Donner Place warehouse to a value of $120,000; and the opportunity to steal Pegasus’ goods in the disorganisation attending the shifting of stock from Mono Place in September 2004.
[120] All these factors support Pegasus’ claim as quantified. As both Mr Ring and Ms Challis concede, once the stock records are accepted, there is simply no other plausible explanation for Pegasus’ losses. To adopt Kenny J’s observation in Trenville Farms, 93,457 items do not “simply disappear into thin air”. And we add, for the purposes of completeness, that we do not accept that the Judge’s careful and comprehensive evaluation of the facts was distorted by considering the stocktaking records before other evidence. We are not persuaded that her decision was incorrect in any material respect.

Set-off

[121] Pegasus cross-appeals against Winkelmann J’s allowance of a set-off against QBE’s liability of $197,637. This sum represented the amount of storage fees which Pegasus withheld from NZE after discovery of its losses in February 2005.
[122] The Judge held:

[301] Clause 29.2.3 [of QBE’s policy] is relevant in determining this issue. It provides:

29.2.3 OTHER INSURANCE – APPLICABLE TO SECTIONS A, B, C, D & E OF THIS POLICY

If at the time of any loss, destruction or damage happening to any property hereby insured, there by [sic] any other subsisting insurance or any indemnity provided by any Act of Parliament whether effected by the Insured or by any other person, covering the same property or the Insured’s interest therein, the insurance under this Policy shall not apply until the full amount of indemnity under such other property or any indemnity provided by any Act of Parliament has been applied as far as it shall go in satisfaction of the Insured’s loss, destruction or damage.

[302] That clause limits QBE’s obligation to indemnify to the balance remaining after payment in full under the indemnity provided by s 9 Law Reform Act 1936. (State Fire Insurance General Manager v Liverpool and London and Globe Insurance Co [1952] NZLR 5).

[303] As to Pegasus’ arguments in relation to the warehouse fees, as I have held, s 9 operates as a form of statutory subrogation. The liquidators will not be able to pursue Pegasus in respect of the warehouse fees, as NZE’s rights to claim set-off have already been exercised.

[304] Therefore Pegasus’ loss, for which it is entitled to indemnity, is reduced by the amount of the warehouse fees that it withheld and by the amount of indemnity it is entitled to from AHA. These amounts are to be deducted from QBE’s obligation to indemnify Pegasus.

[123] In this respect, the Judge noted that Pegasus protested against AHA’s failure to plead a set-off on the ground that it would have led further evidence to prove other claims against NZE.[36] However, the Judge was satisfied that AHA’s failure had not prejudiced Pegasus. That was because s 310(1) of the Companies Act 1993 applied to this effect:

310 Mutual credit and set-off

(1) Where there have been mutual credits, mutual debts, or other mutual dealings between a company and a person who seeks or, but for the operation of this section, would seek to have a claim admitted in the liquidation of the company,—

(a) An account must be taken of what is due from the one party to the other in respect of those credits, debts, or dealings; and

(b) An amount due from one party must be set off against an amount due from the other party; and

(c) Only the balance of the account may be claimed in the liquidation, or is payable to the company, as the case may be.

[124] The Judge then noted:

[297] ... Insolvency set-off occurs automatically. Liquidation of the debtor company (NZE) operates to extinguish the principal debt as at the date of liquidation, leaving the balance as the amount provable in the liquidation (Stein v Blake [1995] 2 All ER 961; New Zealand Bloodstock Leasing Ltd v Jenkins HC AK CIV 2004-404-5795, 19 April 2007). This is to be contrasted with legal set-off, which must be pleaded and is effective only with the judgment of the Court.

[298] For this reason, it is difficult to see how Pegasus can have been prejudiced by AHA’s failure to plead set-off. The liquidation of NZE had the effect of automatically netting off NZE’s and Pegasus’ cross claims, leaving an amount owing to Pegasus. AHA is entitled to raise defences available to NZE, including set-off. Pegasus is able to prove in the liquidation of NZE for the balance of its claims. I add that NZE’s liquidator would not be able to raise the set-off defence again when dealing with any proof Pegasus may file in the liquidation or sue Pegasus for the warehouse fees, as s 9 is a form of statutory subrogation.

[125] By way of background, QBE applied for leave to amend its statement of defence five days before the trial started. The insurer sought to raise three new and affirmative defences including set-off. Winkelmann J allowed the application to plead set-off but refused leave to plead the two affirmative defences.[37]
[126] The ratio of the Judge’s decision on this point is:[38]

I accept QBE’s submission that the effect of any set-off on Pegasus’ claim is a matter of law. But it should have been pleaded as it is also an affirmative defence. But I also consider that Pegasus should be able to deal with this late amendment without suffering unfair prejudice. As Mr Tingey confirmed, Pegasus’ principal answer to it is that any set-off matter between NZE and Pegasus in relation to payment of storage fees is irrelevant to the indemnity claim under the insurance policy. A second and subsidiary point is that the liquidator has no claim against Pegasus for the storage fees, so Pegasus has not been advantaged by its failure to pay on a disputed claim for those fees. However, even in relation to that second argument, disputes as to whether the sum is actually owing are capable of being dealt with within the context of this trial at the level of principle. I have therefore determined that the application to amend the statement of defence to add in this affirmative defence should be allowed.

[127] Significantly, the Judge was satisfied that the set-off could be dealt with as a issue of principle or law. Further evidence would not be necessary. AHA did not apply to amend its defence to plead set-off.
[128] Mr Tingey submits that Winkelmann J erroneously allowed QBE to claim a set-off because she ignored other claims available to Pegasus against NZE. Those claims, he says, could have been set off against the liquidators’ claim for warehousing fees, thereby reducing the amount available to be set off by QBE at trial; and could not be determined as neither NZE nor the liquidator was a party to the proceeding. If QBE had pleaded its defence earlier, Mr Tingey says, Pegasus would have led further evidence to prove its claims against NZE for failure to meet its contractual obligations.
[129] Mr Tingey accepts that, while AHA stood in NZE’s shoes for the insured claim, part of Pegasus’ claim against NZE was for losses not covered by the AHA policy. Subrogation did not arise for those losses. Accordingly, he says, the Court should not have deducted the amount of the set-off but left it for determination between NZE’s liquidator and Pegasus.
[130] When analysed, it now seems the essence of Pegasus’ appeal is not against the Judge’s finding relating to the company’s rights of indemnity under QBE’s policy, but is directed to her earlier finding about AHA’s liability. Mr Tingey does not challenge the Judge’s conclusion that QBE can only be liable to Pegasus for its loss net of any recovery from NZE or AHA whether by way of indemnity or set-off. Mr Ring points out that Pegasus has not appealed against the finding that AHA is entitled to benefit of the set-off.
[131] Instead, Mr Tingey’s argument is directed towards the proposition that QBE is not entitled to that benefit. But once Pegasus elected not to appeal the threshold finding in AHA’s favour, it cannot succeed against QBE. That is because, as the Judge found, cl 29.2.3 limits QBE’s obligation to indemnify after deducting what is owed by AHA. As Mr Ring submits, QBE can only be liable to Pegasus for its loss net of any recovery by the company from NZE or AHA whether by way of indemnity or set-off. The cross-appeal must fail on that ground alone.
[132] We add that Mr Tingey’s argument must fail for another reason. Mr Tingey refers, in general terms, to Pegasus’ claims for damages sustained by NZE’s failure to deliver stock on time, by NZE’s poor practice of dispatching orders incorrectly and as a result of NZE not providing services to the agreed standard. However, he accepts that no detailed evidence was led from Mr Harrison on this issue at trial and Mr Tingey did not provide us with any particulars of the claims said to be available to Pegasus against NZE’s liquidator, thereby constituting a separate set-off against the liquidators’ right of claim.
[133] Over six years have elapsed since NZE was placed in liquidation. Pegasus has taken no steps to pursue its claims against NZE’s liquidator: NZE’s liquidator has taken no steps to recover the outstanding warehousing fees payable by Pegasus. The limitation period has now expired. The only inference to be drawn from this inactivity is that the liquidator accepts that he does not have an arguable claim against Pegasus or the funds to pursue it; and Pegasus accepts the commercial reality that NZE’s liquidator does not have the funds to meet a claim by an unsecured creditor for its losses.
[134] Mr Tingey separately submits that Pegasus did not receive the full value of its claim for lost stock as the excesses under the QBE and AHA policies of $2,500 and $25,000 respectively, a total of $27,500, were not paid to it. This ground was not raised in Pegasus’ cross-appeal but only emerged for the first time in Mr Tingey’s written synopsis. We are not required to consider it further.
[135] For these reasons, Pegasus’ cross-appeal must be dismissed.

Costs

[136] Costs must follow the event. The appeals by QBE and AHA have been unsuccessful. Pegasus’ cross-appeal has also failed but argument on that issue was relatively confined, justifying a 15 per cent reduction in our primary costs award on the appeals.
[137] Accordingly, QBE and AHA must each pay Pegasus an amount equal to 85 per cent of costs for a standard appeal on a band A basis and usual disbursements.

Solicitors:
Shieff Angland, Auckland for First Appellant
McElroys, Auckland for Second Appellant
Bell Gully, Auckland for Respondent

---

[1] Pegasus Group Limited v QBE Insurance (International) Limited HC Auckland CIV-2006-404-6941, 1 December 2009.

[2] At [38].

[3] At [37].

[4] Betty v Liverpool and London and Globe Insurance Company Ltd [1962] USCA4 340; 310 F 2d 308 (4th Cir 1962).

[5] At 310.

[6] At 311.

[7] Betco Scaffolds Company Inc v Houston United Casualty Insurance Company 29 SW 3d 341 (Tex App Houston 14th district 2000).

[8] At 346–347 (original emphasis).

[9] At 353.

[10] Banner Lumber Company Inc v Indiana Lumbermen’s Mutual Insurance Company United States District Court ED Mich 07-13810, 22 October 2009.

[11] Finlayson Enterprises Ltd v United States Fidelity and Guaranty Company [1980] ILR 1–1245 (ONHCJ).

[12] Flett Motors Ltd v Royal Insurance Company of Canada Ltd (1992) 127 NBR 2d 303 (NBQC).

[13] Trenville Farms Ltd Partnership v Wawanesa Mutual Insurance Co [2003] ABQB 83, [2003] 5 WWR 82.

[14] See Lumley General Insurance (NZ) Ltd v Body Corporate No 205693 [2010] NZCA 316, (2010) 16 ANZ Insurance Cases 61-853 at [27].

[15] At [285] and [288].

[16] At [135].

[17] At [112]–[134].

[18] At [43].

[19] At [44].

[20] At [61]–[111].

[21] At [76].

[22] At [75].

[23] At [76].

[24] Austin, Nichols & Co Inc v Stichting Lodestar [2007] NZSC 103, [2008] 2 NZLR 141 at [13].

[25] See above at [80].

[26] See above at [60](a).

[27] At [137].

[28] At [138]–[148].

[29] At [145].

[30] At [149].

[31] At [150].

[32] At [142].

[33] At [152].

[34] At [155].

[35] At [47], [48], and [57](c).

[36] At [296].

[37] Pegasus Group Ltd v QBE Insurance (International) Limited HC Auckland CIV-2006-404-6941, 27 April 2009.

[38] At [57].