THE SUPREME COURT OF APPEAL OF SOUTH AFRICA
JUDGMENT
Case No: 642/2010
In the matter between:
VENFIN INVESTMENTS (PTY) LTD Appellant
v
KZN RESINS (PTY) LIMITED t/a KZN RESINS Respondent
Neutral citation: Venfin Investments v KZN Resins (642/2010) [2011] ZASCA 128 (15 September 2011).
Coram: BRAND JA, PONNAN JA, SNYDERS JA, MALAN JA AND THERON JA
Heard: 16 August 2011
Delivered: 15 September 2011
Summary: Appellant’s claim in convention based on cession of rights by third party in terms of alleged oral agreement – factual issue whether agreement entered into decided against appellant – counterclaim based on the debts owing to respondent by same third party – proposition that appellant liable for these debts in terms of s 156 of Insolvency Act 24 of 1936, alternatively, pursuant to the terms of the cession – both issues decided against respondent.
________________________________________________________________
ORDER
On appeal from: KwaZulu-Natal High Court (Durban) (Van der Reyden J sitting as court of first instance):
1. The appeal is upheld. The order of the court a quo is set aside and replaced with the following:
“(a) The plaintiff’s claim is dismissed with costs, including the costs of two counsel.
(b) The defendant’s counterclaim is dismissed with costs, including the costs of two counsel.”
2. The respondent is ordered to pay the appellant’s costs of appeal including the costs of two counsel, but the costs pertaining to the record is restricted to 10 per cent thereof.
3. The cross-appeal is dismissed with costs.
JUDGMENT
________________________________________________________________
BRAND JA (PONNAN JA, SNYDERS JA, MALAN JA AND THERON JA concurring):
[1] This appeal has its origin in a business relationship between the respondent, KZN Resins (Pty) Ltd (KZN) and another company, Fibalogic (Pty) Ltd, that went awry. Fibalogic is now in liquidation, but during its corporate lifetime it was a manufacturer of electric water heaters, commonly known as geysers, that were made of fibreglass. Between 2000 and 2002 KZN supplied Fibalogic with resin used in this manufacturing process. Litigation started on 12 September 2002 when Fibalogic instituted an action in the KwaZulu High Court (Durban) against KZN for an amount in excess of R26 million. Broadly stated, Fibalogic’s cause of action was for its loss resulting from geysers that were returned to it because of defects, for which KZN allegedly undertook to pay compensation in terms of an oral agreement.
[2] KZN delivered a plea in which it disputed the oral agreement. It also instituted a counterclaim for about R2 million which it alleged was the balance of the purchase price of resin sold and delivered to Fibalogic. After the close of pleadings, Fibalogic ceded its claim against KZN to the appellant, Venfin Investments (Pty) Ltd (Venfin). As part of the cession agreement Venfin indemnified Fibalogic against any claim by KZN for goods sold and delivered. Thereafter, Fibalogic was placed under liquidation and Venfin was formally substituted as the plaintiff in the ongoing proceedings.
[3] When the matter eventually came before Van der Reyden J he was asked, by agreement between the parties, to order a separation of issues. In terms of the separation order he consequently granted, three specified areas of dispute were to be determined at the outset while all other issues, including those relating to the quantum of the claim and the counterclaim, stood over for later determination. The three specified areas of dispute were formulated thus:
(a) The issues relating to the oral compensation agreement relied upon by Venfin for its claim in convention, as defined in paragraph 9 of Venfin’s particulars of claim read with paragraph 8 of KZN’s plea (the first issue);
(b) The issue arising from KZN’s counterclaim as to whether the indemnity furnished by Venfin to Fibalogic – in terms of their cession agreement – is governed by s 156 of the Insolvency Act 24 of 1936, so as to render Venfin liable for KZN’s claim against Fibalogic for resin sold and delivered to the latter (the second issue); and
(c) The issues raised by KZN’s counterclaim as to whether the cession agreement by itself had the effect of transferring and imposing on Venfin, Fibalogic’s liability for the resin sold and delivered to it (the third issue).
[4] Evidence in the preliminary proceedings was led in two tranches during February and August 2008. Judgment was delivered about 15 months later on 24 March 2010. In terms of the judgment the court a quo decided the first and third issues in favour of KZN and the second issue in favour of Venfin. In consequence, the claim in convention was dismissed while the counterclaim succeeded pursuant to the court’s findings on the third issue. In both instances, costs were awarded in favour of KZN, save for the costs resulting from the second issue, which were ordered against it. Venfin’s appeal to this court against the judgment on the first and third issues is with the leave of the court a quo. So is KZN’s cross-appeal against the costs order on the second issue. I propose to deal with the three issues individually.
The first issue – did KZN agree to compensate Fibalogic for damages resulting from defective geysers?
[5] The outcome of this issue turns exclusively on the correctness of the court a quo’s findings with regard to a dispute of fact between the parties. Venfin’s version of the facts was summarised thus in paragraph 9 of its particulars of claim, to which reference is pertinently made in the formulation of this issue:
‘On or about 23 November 2001, at Paarl, Fibalogic, represented by its Managing Director Mr Dawie Thirion, and [KZN], represented by its Chairman Mr Salim Kajee, both of who were duly authorized thereto, concluded an oral agreement, the material terms of which included the following:
KZN would compensate Fibalogic for the costs incurred by Fibalogic in respect of all returns in excess of the Fibalogic’s norm of returns, which norm was 2% of the number of units produced by it;
such costs would include the cost of the replacement water heater (geyser) and Fibalogic’s labour and travelling costs.
The agreement was confirmed on 26 November 2001 by Fibalogic’s Mr Thirion in a letter to KZN’s Mr Kajee, a copy of which is annexed marked “A”.’
[6] Annexure A, which was destined to play a prominent role at the hearing before Van der Reyden J, reads in relevant part:
‘Dear Salim
We refer to our meetings of 16, 22 and 23 November 2001.
After analysing all the variables we are all in agreement that there is a difference in performance between the resins supplied by you and the resins previously used.
We have discussed and shown you from our analyses the norm expected from a resin as used in our back-up layer on our tanks. We believe that the main question remains unanswered as to the difference between your resin and the one previously used.
. . .
Summary of the meeting, dated Friday, 23 November 2001:
. . . ;
KZN Resins will compensate Fibalogic the difference between the agreed norm (of two per cent of production) and the actual rate experienced. This will include the cost of the geyser as well as the labour/travelling cost. (Warrantee costs);
KZN Resins will continue to subsidise the additional lay-up costs on the 150 litre tank. . . .
Whilst we appreciate and accept your offer as outlined above, we however are still of the opinion that the reason(s) in variation of performance must be found and we will run independent tests to answer this.
Regards
D Thirion
CC F G Rupert
D Reid’
[7] KZN did not respond in writing to Annexure A until 8 July 2002, in circumstances that will presently transpire. On that date Mr Salim Kajee conveyed a letter to Mr Dawie Thirion by telefax in which he essentially denied the agreement relied upon by Venfin. Paragraph 8 of Venfin’s plea (which is also referred to in the formulation of this issue) echoed that denial. In relevant part it reads:
‘Ad paragraph 9
Save that the Defendant admits that on or about 23rd November 2001, and at Paarl, Mr Thirion and Mr Kajee had a discussion and that the Plaintiff wrote a letter dated 26th November 2001 (annexure “A” to the Particulars of Claim) to KZN, KZN denies each allegation in paragraph 9 and:
specifically denies that it concluded an oral agreement with Fibalogic either in the terms alleged or at all;
specifically denies that annexure “A” correctly reflects or records the matters discussed at the meeting between Mr Thirion and Mr Kajee on the 23rd November 2001.’
[8] At the hearing, Venfin’s version was supported by the evidence of Fibalogic’s erstwhile managing director, Thirion. For its denial of that version, KZN in turn, relied on the testimony of two witnesses. Kajee, the former chairman of KZN, and Mr Donald Reid, who was the technical director of Fibalogic at the time. The factual dispute about the conclusion and the terms of the alleged oral agreement that emerged from the pleadings, maintained its course throughout the hearing. Yet, there were large areas of common ground, as appears from the following background, which I find most convenient to narrate in chronological fashion.
[9] Reid was involved with Fibalogic from its inception and he remained a shareholder and the technical director of the company throughout its existence, until it was eventually wound up in July 2003. His testimony therefore provides a convenient starting point to the chronological narrative. As a marine engineer by training, Reid came up with the idea of manufacturing corrosion free fibreglass geysers involving a chemical compound called vinyl-ester resin. After he patented his invention, he incorporated the company, Fibalogic, together with a business partner and then ceded the patent to the company. At the outset, the tanks of the geysers, which were manufatured in three capacities of 100 litres, 150 litres and 200 litres, were made of vinyl-ester only. This compound was obtained from a supplier, NCS, under the trade name Derakane. During about 1996, Reid and his partner sold 50 per cent of their shares in Fibalogic to Venfin, a company in the Rembrandt stable. In the course of time, their shareholding was, however, ‘watered down’ to about one per cent, while Venfin owned the rest.
[10] After Venfin became involved – or, as Reid put it, ‘when the accountants came in’ – it was decided, as part of a cost cutting exercise, to change the tanks of the geysers from a single composition to a dual lay-up system involving different kinds of resin. While the inner layers, which came into direct contact with water, still consisted of vinyl-ester, the outer layers were manufactured out of the substantially cheaper isopthalic resin. At that time both the vinyl-ester and the isopthalic resins were supplied by NCS. During January 2000 Fibalogic decided on a further cost cutting exercise by reducing the number of outer layers by one on its 100 and 150 litre cylinders, well-knowing that this would reduce the mechanical strength of these tanks. The number of layers on the 200 litre cylinders remained the same.
[11] Thirion holds a university degree in commerce. He joined Fibalogic as its managing director in April 2000. He was head-hunted for that position by Venfin, because Fibalogic was consistently running at a loss and it was hoped that Thirion could turn that situation around. During the second half of 2000, NCS proposed a substantial increase in its price for isopthalic resin. In consequence, Fibalogic started looking for an alternative supplier. Eventually it decided on KZN. From 13 October 2000 Fibalogic thus procured its isopthalic resin from KZN while NCS continued to supply its vinyl-ester. KZN manufactured the isopthalic resin in accordance with Fibalogic’s specifications. Every batch destined for delivery to it was analysed by the South African Bureau of Standards (SABS) and the analysis recorded in a certificate which accompanied the delivery. On occasion when Fibalogic was not satisfied by the certificate that the batch conformed to its specifications, it was returned to KZN.
[12] There had always been returns of hot water cylinders to Fibalogic because of defects. In the past these returns averaged about 1.36 per cent of cylinders manufactured. After October 2000 there was, however, a marked increase in this rate of return. It rose from the previous average of 1.36 per cent to well in excess of the industry norm of 2 per cent. The increase was particularly pronounced with regard to the 150 litre cylinders. The consequences of the increase were serious and Fibalogic’s very existence was threatened, unless the problem could be resolved. Thirion in particular was under severe pressure from the Venfin representatives on the Fibalogic board of directors to resolve the problem of increased returns.
[13] At that stage no-one knew what the cause of the failures was or who was responsible for these failures. The reason for the uncertainty was that the hot water cylinders consisted of about 220 components, including different kinds of resin, glass fibre matting, heating elements, thermostats, etc, which were procured from a number of different sources. Moreover, there were various stages of manufacturing where things could go wrong, for instance when the resin in liquid form was mixed with the chemical catalysts. At the trial it was common cause between all witnesses, including Thirion, that with the benefit of hindsight, the problem could be ascribed to the change in design from a single lay-up consisting of vinyl-ester only to a dual lay-up of vinyl-ester on the inside and isopthalic resin on the outside of the cylinder. This was empirically established by Thirion himself. After the liquidation of Fibalogic, he became the manager of the company that took over the business of manufacturing geysers. According to his evidence that company solved the problem of excessive returns by reverting to the original, single lay-up design. But this much was not appreciated by those involved in 2001.
[14] During 2001 Thirion and Reid suspected that the increased rate of returns was attributable to the use of KZN’s resin, purely because the increase coincided with the change in their supplier from NCS to KZN. KZN, on the other hand, regarded Fibalogic as a substantial client, with potential to grow even further. Hence it was keen to assist Fibalogic in resolving its compelling difficulties. As part of its attempts to do so, KZN offered, in August 2001 to provide additional resin without cost for an extra outer layer on the 150 litre geysers. It will be remembered that during January 2000 Fibalogic reduced the number of outer layers on its 100 and 150 litre geysers as part of a cost saving exercise. With regard to the 200 litre cylinders, the number of layers remained the same. Since the returns were more pronounced with regard to the 150 litre cylinders than with reference to the larger 200 litre ones, it was thought that the problem might have been caused by the reduction of one outer layer. Consequently, Fibalogic decided to re-introduce the additional outer layer on the 150 litre cylinders. Agreement was then reached that Fibalogic would supply the glass fibre matting for the extra layer while KZN would provide the resin for that layer free of charge. According to Kajee this resulted in a discount of about 5 per cent in the price of resin sold. The extra layer was apparently introduced from about 14 August 2001.
[15] The hope was that the additional layer would resolve the problem in due course. Yet it did not immediately reduce the pressure on Thirion. This is borne out by the minutes of the Fibalogic board meeting of 31 October 2001 which recorded that ‘the chairman expressed his concern over [the costs of product failure] and demanded that it be brought under control’. The product failure, so the minutes stated, ‘includes the failure of suppliers’ products, valves, thermostats and elements’.
[16] This led to the crucial meeting which was held on 22 and 23 November 2001. It was attended by Thirion and Reid on behalf of Fibalogic and by Kajee on behalf of KZN, though Reid was not present on the 23rd. At the meeting Thirion again stressed that the return rate had increased from an average of 1.36 per cent to over 2 per cent, from about the time that Fibalogic changed to KZN resin in October 2000. He also said that in the circumstances his perception was that the resin supplied by KZN was responsible for the problem and that, if KZN should refuse to shoulder that responsibility, he would have to change the supplier of isopthalic resin. At the same time, everybody concerned believed that the re-introduction of the additional outer layer to the 150 litre cylinders on 14 August 2001 would stem the tide of increased returns, though the exact source of the problem remained unknown.
[17] The narrative thus far is essentially common cause in the sense that it was put forward by one side and either admitted or not denied by the other. The sharp dispute, which goes to the heart of the case, turns on Kajee’s reaction at the meeting. Thirion’s side of the controversy echoed the contents of his letter of 26 November 2001 to which I have referred extensively (in para 6 above). According to that version Kajee agreed at the resumed meeting on 23 November 2001 that KZN would compensate Fibalogic for its loss incurred, calculated on the basis of the difference between the agreed industry norm of 2 per cent and the actual rate of return experienced, including labour and travelling costs. Kajee’s version, on the other hand, was that it became clear to him during the course of the meeting that Thirion had expected KZN to pay for the returns above 2 per cent of production. Though he did not share the belief that KZN’s resin was to blame, he suggested that Fibalogic should look to KZN’s insurer for compensation. He explained to Thirion, so Kajee testified, that Fibalogic had product liability cover for R2.5 million and that if KZN’s resin should prove to be the cause of the problem, the insurer would have to pay. In fact, Kajee testified, he offered to assist Fibalogic in initiating a claim against KZN’s insurer.
[18] According to Kajee one of the main reasons why he would not admit liability on behalf of KZN, as alleged by Thirion, originates from the very terms of that insurance policy. He had been cautioned by KZN’s insurance broker, so Kajee said, that in terms of the policy, any admission of liability by him would entitle the insurer to repudiate the claim. Thirion denied that Kajee ever suggested that Fibalogic should initiate a claim against KZN’s insurer. Reid on the other hand remembered that he did. Because Reid was not present on 23 November 2001 he was not able to comment on the direct conflict between Thirion and Kajee as to whether the latter undertook responsibility for Fibalogic’s problem. Yet he testified that, from his expert knowledge of the industry, he would not have expected the supplier of the resin to effectively guarantee the entire geyser against failure. His reason for this view was that there were too many variables beyond the supplier’s control that could cause the failure.
[19] After the meeting Thirion then wrote the by now familiar letter of 26 November 2001 (quoted in para 6 above) in which he recorded his version of Kajee’s undertaking. According to Thirion he wrote the letter with the assistance of Reid who is more fluent in English than him. This is denied by Reid. What is common cause is that Thirion did not receive any written response to this letter from Kajee until nearly one year later. Kajee’s version is that when he received the letter he telephoned Reid to find out what was going on. From Reid’s reaction it was obvious to him that Reid knew nothing about the letter, but that he agreed with Kajee that the alleged undertaking on behalf of KZN was never given. According to Kajee, he then telephoned Thirion who told him, in essence, to ignore the contents of the letter. He therefore found it unnecessary to respond in writing. In his testimony Reid confirmed, not only that Kajee had telephoned him about the letter, but also that he was present when Thirion received a telephone call from Kajee. On his part, Thirion emphatically denied that he ever received the alleged telephone call from Kajee.
[20] Thirion’s version as to Kajee’s undertaking is supported to some extent by the minutes of a meeting of the Fibalogic board on 29 November 2001. According to these minutes Thirion reported, inter alia, that ‘failures on geysers due to the resin being used led to negotiations with KZN Resins. They undertook to pay the amount of plus minus R300 000 to Fibalogic and in future to compensate Fibalogic for the difference between the agreed failure norm (2 per cent of production) and the actual failure rate. This will include the costs of the geyser as well as the labour/travelling costs’. Of further significance is the fact that Reid attended that board meeting and that he did not repudiate or query Thirion’s account. The support that Thirion’s version derives from the minutes is somewhat marred by his reference to an amount of R300 000 which he did not mention in his letter or in his evidence about the terms of the agreement.
[21] Kajee’s version, on the other hand, that he suggested a claim against KZN’s insurer, derives support from at least three documents. First there is a note in Thirion’s own handwriting about a conversation between him and Kajee on 7 December 2001 which clearly related to a compensation claim by Fibalogic which had been referred to KZN’s insurer. The second document is a letter by KZN’s insurance broker to its insurer dated 28 January 2002. The letter confirms ‘advice of a potential product liability claim’ against the insured, KZN, by Fibalogic and that ‘the insured have done their best themselves without admitting liability, to negotiate with Fibalogic’. Thirdly, there are the minutes of the Fibalogic board meeting of 27 February 2002 which recorded a report by the management of Fibalogic that KZN ‘had various meetings with their insurer and that we have to lodge our claim directly against [KZN]. We are now processing all the existing information and will lodge our claim at month end’.
[22] It is common cause that Kajee then tried to persuade KZN’s insurer to pay Fibalogic’s claim; that the insurance assessor investigated the claim and brought out a report; and that the claim was eventually repudiated by the insurer. While giving evidence Thirion was very upset by a comment in the assessor’s report that at one stage during the contract period there was a problem with KZN’s test equipment which affected the quality of its resin delivered to Fibalogic, about which the latter was never informed.
[23] After November 2001, Thirion started to deduct from amounts payable to KZN for resin sold and delivered the compensation to which Fibalogic was in his view entitled in terms of their alleged agreement. This gave rise to a demand by KZN’s accountant for payment of the outstanding amounts. In response, Thirion wrote a letter to Kajee on 28 June 2008 in which he referred to their alleged compensation agreement of November 2001. In this letter he added that he could understand why Kajee would not admit the agreement in writing for ‘fear of committing to an unknown quantum or to jeopardise any future insurance claims you may incur’ but that this did not detract from the fact of the agreement. In reply to this letter Kajee, for the first time in writing denied the compensation agreement on 8 July 2002.
[24] On the crucial dispute of fact as to whether the compensation agreement was ever entered into, the court a quo essentially accepted the version of Kajee and Reid in preference to that of Thirion. On appeal Venfin contended that the court had misdirected itself in doing so. Its first argument in support of this contention relied on Kajee’s failure to respond to Thirion’s letter of 26 November 2001, until nearly one year later. In the light of the decisions in McWilliams v First Consolidated Holdings (Pty) Ltd 1982 (2) SA 1 (A) at 10E-G and Hamilton v Van Zyl 1983 (4) SA 379 (E) at 388F-G, so Venfin’s argument went, the court a quo should have held that, as a matter of law, Kajee’s failure to respond amounted to an admission of the allegations in Thirion’s letter. As I see it, the flaw in this argument is that it rests on a misunderstanding of the cases upon which it seeks to rely. Neither McWilliams nor Hamilton sought to lay down any principles of law. They reflected conclusions based on the application of logical reasoning to the facts of those cases.
[25] The flaw in Venfin’s argument can be illustrated, I think, by supposing an admission on the part of Thirion that Kajee had telephoned him to deny the allegations of an agreement when he received the letter. In these circumstances any attempt to construe Kajee’s failure to respond in writing as an admission of the letter, would clearly be untenable. The fact that Thirion denied the telephone conversation deposed to by Kajee makes no difference in principle. The denial brings about the enquiry whether Kajee’s version should be accepted. It does not, as Venfin’s argument would have it, exclude that enquiry as a matter of law. The mere fact that the letter was written does, of course, lend some support to Thirion’s version. That support is, however, somewhat detracted from by his assertion that he, as a university graduate, required the assistance of Reid to write a letter in simple English. The assertion became even more suspicious when it was denied by Reid, who had no apparent reason to do so, if it were true.
[26] Venfin’s second line of argument in support of the professed misdirection by the court a quo, focussed on the court’s alleged unwarranted criticism of Thirion and the proposition that, on the contrary, Thirion was a better witness than Kajee and Reid. In the course of this argument the testimony of all three witnesses was subjected to a detailed analysis; alleged unsatisfactory aspects in the evidence of Kajee and Reid, particularly in cross-examination, were accentuated; alleged strong points in Thirion’s evidence were underscored; and the flaws in his evidence explained. Not unexpectedly, this gave rise to what essentially amounted to a mirror image of the same argument on behalf of KZN. According to this argument it was Thirion who was the unsatisfactory witness while the flaws in the evidence of Reid and Kajee were explained, and so forth. I find it unnecessary to give an account of this rather painstaking exercise. I am prepared to accept that there is some merit in the criticism against the evidence of Kajee and Reid on the one hand and that of Thirion on the other. It also seems to me that the contents of the documentary evidence referred to, including the correspondence, were at best inconclusive. Yet, despite all this, I find Thirion’s version to militate so strongly against the inherent probability that the contrary version of Kajee, as supported by Reid, should in my view be preferred. In what follows I propose to motivate this finding.
[27] At the time of the November meeting the cause of Fibalogic’s problems had not been identified. That much is indeed underscored by the comment in the last paragraph of Thirion’s letter of 26 November 2001 that ‘the reason for the variation of performance must [still] be found’. Though both Thirion and Reid suspected KZN’s resin, their suspicions remained unsubstantiated by their own extensive search for the cause. Reid, who has no apparent reason to side with KZN, expressed the considered view as an expert in the field, that he would not expect the manufacturer of resin to take responsibility for all failures of geysers for which there was a myriad of potential causes. Yet, on Thirion’s version, this is exactly what Kajee undertook to do despite his awareness of the myriad of potential causes. Moreover, he undertook to do so out of the blue and without any demure. According to Thirion, Kajee did not even seek to qualify or limit his undertaking in any respect. Even if the increased failure resulted from some cause entirely unrelated to the quality of KZN’s resin, the latter would, according to Thirion, be liable in full after returns had reached the 2 per cent level.
[28] What must also be borne in mind, is that KZN’s resin was made up according to specifications provided by Fibalogic and certified to be in accordance with these specifications by the SABS. In the circumstances one would expect Kajee to argue that, even if KZN’s resin was found to be the cause of the increased failure rate, the problem could lie with the specifications. Yet, according to Thirion, Kajee did not even try to raise this argument. In fact, he raised no argument at all. He simply undertook to pay. Added to all this, it flies in the face of probabilities, I think, that Kajee would not even try to establish the amount of the liability KZN would have to pay. He simply undertook to pay an indeterminate amount. In addition, according to Thirion, he undertook to do so for an unfixed period in the future.
[29] With regard to the duration of the undertaking, Fibalogic alleged in its pleadings that KZN agreed to pay compensation in respect of all geysers manufactured between 13 October 2000 and 14 August 2001. As I understand it, the first mentioned date was when Fibalogic started to use KZN’s resin while 14 August 2001 is the approximate date when it decided to revert to an extra outer layer on its 150 litre geysers. Yet, whatever the origin of these limitations, they are not set out in the letter of 26 November 2001. Moreover, they do not accord with Thirion’s evidence that, according to his understanding, Kajee’s undertaking would apply for as long as Fibalogic purchased resin from KZN. On the contrary, in cross-examination he expressly disavowed the limitation alleged in Venfin’s pleadings.
[30] In argument counsel for Venfin seemed to accept that an undertaking by Kajee for an indefinite period in the future was inherently improbable. They therefore reverted to the limitations alleged in Venfin’s pleadings. When they were reminded that these allegations were pertinently disavowed by Thirion in evidence, their answer was that the limitations should be regarded as incorporated into the agreement by way of a tacit term. However, as explained by Corbett AJA in Alfred McAlpine & Son (Pty) Ltd v Transvaal Provincial Administration 1974 (3) SA 506 (A) at 531-532, a tacit term is an unexpressed provision of a contract derived from an inference as to what both parties must have intended. In this light, it is simply not open to a party who expressly denied that he ever intended a particular term to form part of a contract, to contend that the term must be inferred. Put somewhat differently; according to the celebrated officious bystander test, a tacit term can only be incorporated if it can confidently be said that, if at the time of the contract the officious bystander were to ask the parties about the existence of that term, they would both have said ‘of course it forms part of our contract; it is so obvious we did not even trouble to say that; it is too clear’ (see eg Botha v Coopers & Lybrand 2002 (5) SA 347 (SCA) para 23). That being so, a party who had expressly denied the existence of a particular term can hardly suggest that the same term was so obvious he did not even find the need to express it.
[31] The submission to the contrary by Venfin’s counsel, that the inherent probabilities in fact favoured Thirion’s version, rested on the following three arguments:
(a) Kajee regarded Fibalogic as a substantial client with the potential to grow even further and he knew that if he should refuse to give the undertaking, KZN would probably lose that client.
(b) In August 2001, as part of the attempts to address the problem of increased returns, KZN agreed to supply Fibalogic, free of charge, with resin for an additional outer layer to its 150 litre cylinders. It was anticipated by everybody, including Kajee, that the additional layer would solve the problem of the increased rate of returns in respect of cylinders manufactured after that date. This would substantially reduce KZN’s potential liability in terms of the undertaking.
(c) Kajee was aware of the fact that at some stage in the past KZN’s test equipment was faulty and that it consequently delivered substandard resin for which it was responsible.
[32] I shall deal with these three arguments individually. First, there is the argument based on the fact that Fibalogic was a substantial client of KZN. I believe the answer to the argument is that, despite this fact, it remains improbable that Kajee would risk the potential commercial suicide of KZN in order to retain the business of one client, albeit a substantial one. Moreover, Kajee knew KZN was insured against product liability. If KZN was responsible, the insurer would probably pay. In the event, KZN would retain Fibalogic as a client without incurring any financial risk. In this light it is far more likely that Kajee would refer Fibalogic to his insurer. Conversely, these circumstances render it improbable in the extreme that Kajee would shoulder liability on behalf of KZN without even referring the matter to the insurer. This is particularly so where Kajee appreciated that his admission of liability may provide the insurer with a ground for repudiation which would frustrate his obvious way out.
[33] As to the argument based on Kajee’s anticipation that the problem had been solved in respect of geysers manufactured after August 2001, I believe the answer is at least twofold:
(a) First, this argument would fly in the face of the appellant’s further argument that Kajee’s undertaking was somehow limited to geysers that were manufactured prior to the introduction of the additional outer layer in August 2001. If Kajee only undertook liability for geysers that were manufactured prior to that date, his anticipation that, in respect of geysers manufactured after that date there would be no more returns, would have no effect on the extent of KZN’s liability.
(b) Second, and more significantly, if the problem had indeed been solved by the extra layer, it would mean that the increased failure rate in geysers manufactured prior to August 2001 had nothing to do with KZN’s resin. I say this because we know that the extra layer had been removed by Fibalogic in January 2000 as part of a cost saving exercise. If the reinstatement of the extra layer thus resolved the problem, it would follow that the problem was caused by a deliberate design change by Fibalogic which had nothing to do with KZN. Taken to its logical conclusion, acceptance of Venfin’s argument under consideration would therefore mean that Kajee undertook liability for failures of geysers while anticipating an event that would conclusively absolve KZN from responsibility for those failures. If anything, this renders Thirion’s version even more untenable.
[34] As to the third argument based on Kajee’s knowledge that at some stage in the past the resin delivered to Fibalogic was not up to the required standard, the answer is, in my view, quite obvious. At best for Fibalogic one could in those circumstances expect Kajee to accept liability for the batch or batches of resin delivered that were not up to standard. It would still raise the rhetorical question why he would accept liability for failed geysers which could not have been manufactured with those batches of defective resin.
[35] Finally, it seems to me that even if Venfin had been successful in establishing the compensation agreement, it would have been confronted with another obstacle which appears to be insurmountable. It is this. On Thirion’s version it was a tacit term of the agreement that KZN would not be liable once it was positively confirmed that KZN’s resin was not to blame. At the same time Thirion agreed with Reid’s expert opinion that, with hindsight, the problem of increased returns was caused by a design change and entirely unrelated to KZN’s resin. The answer to this problem proffered by Venfin’s counsel was that causation was not in issue at the preliminary stage. I am not persuaded by that answer. Once the essential element of causation had been eliminated on the undisputed facts at the preliminary stage, as I believe it was, it must follow that there is nothing left to proceed to a next stage. It is the end of the matter.
[36] For these reasons I believe this court should endorse the court a quo’s ultimate finding on the first issue, namely, that Venfin had failed to establish the compensation agreement upon which it relied for its claim. Moreover, I believe that even if the agreement had been established, the claim could not succeed. It follows that in my view the claim in convention was rightly dismissed. This brings me to the court a quo’s finding on the third issue which was the basis upon which KZN’s counterclaim was upheld.
The third issue – did the cession agreement between Venfin and Fibalogic have the effect of transferring KZN’s claims against the latter to the former?
[37] It will be remembered that Venfin never incurred any direct liability towards KZN. The counterclaim against Venfin for resin sold and delivered to Fibalogic was founded, in the main, on the provisions of s 156 of the Insolvency Act 24 of 1936, to which I shall presently return. In the alternative, the counterclaim against Venfin relied on the provisions of a cession agreement between Fibalogic and Venfin which was entered into after the close of pleadings in the action between KZN and Fibalogic and shortly before the liquidation of the latter. It was on this basis that the counterclaim succeeded in the court a quo.
[38] The pertinent clauses of the cession agreement provided:
‘12.1 It is recorded that the company [Fibalogic] is currently involved in a legal dispute with KZN, one of its creditors. The company has instituted action against KZN in the amount of . . .
The company hereby cedes . . . its claim against KZN to the seller [Venfin] which cession the seller hereby accepts.
12.5 The seller hereby indemnifies the company against any loss, liability, damage (excluding consequential damage), cost or expense of any nature whatsoever which the company may suffer or incur as a result of any claim made against the company by KZN for goods and/or services provided to the company . . . (“indemnified loss”).
12.8 The seller shall be obliged to pay to the company the amount of any indemnified loss suffered or incurred by the company as soon as the company is obliged to pay the amount thereof.’
[39] The reasons given by the court a quo for upholding KZN’s counterclaim on the basis of this cession are rather terse. From these terse reasons it would appear, however, that the court was swayed by three considerations. First, that a cession cannot impose a greater burden on the debtor or weaken the debtor’s position, the debtor, in this context, being KZN. Second, that at the time of the cession the directors of Fibalogic were aware of its precarious financial position and that the cession was entered into with the purpose of frustrating KZN’s counterclaim against Fibalogic. Third, that in terms of clause 12.5 of the cession agreement, Venfin effectively stepped into the shoes of Fibalogic by undertaking responsibility for KZN’s claim.
[40] I do not believe that any of these three considerations can be sustained. As to the first, it is indeed a trite principle that a cession cannot weaken the debtor’s position. However, any attempt to do so would affect the validity of the cession. It would not in itself afford the debtor any rights against the cessionary. But, be that as it may, I cannot see how the cession under consideration can be said to have weakened KZN’s position in any respect. KZN retained whatever claim it had against Fibalogic. The fact that after liquidation it only had a concurrent claim did not result from the cession. It resulted from the liquidation. In the absence of the cession agreement, KZN’s position would have been no better. Lastly, in any event, the court a quo found, under the rubric of the first issue, rightly in my view, that Fibalogic had no claim against KZN, which means that KZN did not even qualify as a debtor of the cedent.
[41] As to the second consideration, there is no evidence that the cession was entered into in order to frustrate KZN’s claim. On the contrary, the cession formed part of a much larger transaction involving, in the main, Venfin’s sale of its shareholding in Fibalogic to a third party. In any event, if that was the purpose of the cession, that purpose had not been achieved. As I have said, KZN’s position with regard to its claim against Fibalogic, remained exactly the same. Assertions of a potentially frustrated reliance on set-off by KZN, raised by its counsel in argument, were inapposite. Set-off is a shield not a sword. Absent a valid claim by Fibalogic against it, KZN required no shield.
[42] The answer to the third consideration based on clause 12.5 of the cession, is that the provisions of the clause are exclusively for the benefit of Fibalogic. They bestowed no right on KZN. As a matter of law, Fibalogic’s obligations to KZN could only have passed to Venfin by way of a delegation, which would require a tripartite agreement between the creditor (KZN), the debtor (Fibalogic) and the assignee (Venfin). No agreement of delegation was either pleaded or established in evidence by KZN. It follows that in my view the court a quo had erred in allowing the counterclaim on the basis that it did. This brings me to the second issue.
The second issue – did the indemnity furnished by Venfin to Fibalogic render it liable to KZN for Fibalogic’s indebtedness under s 156 of the Insolvency Act 24 of 1936?
[43] It will be remembered that the issue arose from KZN’s allegation that Venfin was liable to it for the debt of Fibalogic by virtue of s 156. It remains to be said that, because Fibalogic was wound up for inability to pay its debts, s 339 of the Companies Act 61 of 1973 rendered the provisions of the law relating to insolvency, applicable. The court a quo dismissed KZN’s claim based on s 156 with costs. The adverse costs order, in turn, gave rise to the cross-appeal. The provisions of s 156 read as follows:
‘Insurer obliged to pay third party’s claim against insolvent
156. Whenever any person (hereinafter called the insurer) is obliged to indemnify another person (hereinafter called the insured) in respect of any liability incurred by the insured towards a third party, the latter shall, on the sequestration of the estate of the insured, be entitled to recover from the insurer the amount of the insured’s liability towards the third party but not exceeding the maximum amount for which the insurer has bound himself to indemnify the insured.’
[44] Venfin’s answer to the claim, which found favour with the court a quo, was that s 156 applies only to the liability of an insurer – properly so called – to a third party under a policy of (indemnity) insurance. Since Venfin is not an insurer properly so called, Venfin argued, the section finds no application. KZN’s reposte to that answer was that the wide wording of s 156 does not warrant the restriction on its application that Venfin’s argument seeks to impose. In support of this counter argument, KZN referred to the wide wording of the section – ‘any person’ who is obliged to indemnify ‘another person’ in respect of ‘any liability’. The references to ‘insurer’ and ‘insured’ in the section, so KZN’s argument went, were clearly for ease of reference only. The legislature might as well have used any other term such as ‘the indemnifier’, in its stead. Had the legislature intended to impose the limitation contended for by Venfin, so KZN’s argument proceeded, it would have referred specifically to the situation where ‘an insurer’ was obliged to indemnify ‘an insured’ in terms of ‘any contract of insurance’. In further support of its argument, KZN referred, by way of comparison, to corresponding legislation in England (the Third Party (Rights Against Insurer) Act 1930, s 1) and in Australia (the Bankruptcy Act 1966, s 117) where the statutory measures akin to our s 156 are expressly reserved for contracts of indemnity insurance.
[45] Though these are undoubtedly weighty arguments, difficulties flow from the reference to ‘insurer’ and ‘insured’. It is true that on the face of it, the use of these terms could be understood as merely for ease of reference. But they are well recognised terms of art. What is more, a contract of indemnity is wide enough to cover an indemnity insurance policy. In the parlance of natural science, indemnity contracts can thus be described as the genus of which indemnity insurance is one of the species. The effect, as I see it, is this: by referring to the person liable as ‘the insurer’ instead of, for example, the ‘indemnifier’, or for that matter, a neutral term such as ‘the pumpkin’, the legislature appears to limit the wide meaning of ‘any person’ or ‘any indemnifier’ to the specific form of indemnity provided by an insurance policy. Stated somewhat differently, the definition of the genus by reference to one of the species renders the section capable of the limited interpretation that it only applies to that species. In addition, the heading of the section also directs the focus at the species. By all accounts, the section is therefore ambiguous. The loose and imprecise language used left its meaning uncertain.
[46] History seems to support the limited interpretation contended for by Venfin. Section 156 has been in existence, in unamended form, for 75 years. Over that period the section has been uniformly referred to, both in judicial pronouncements and in insolvency textbooks, in the context of insurance contracts only (see eg Le Roux v Standard General Versekeringsmaatskappy Bpk 2000 (4) SA 1035 (SCA); Coetzee v Attorneys’ Insurance Indemnity Fund 2003 (1) SA 1 (SCA); E Bertelsman et al Mars The Law of Insolvency in South Africa (9 ed) para12.16; Meskin Insolvency Law, para 5.3.2.2). But, as was rightly pointed out by KZN, it has thus far not been pertinently held that the section has no application outside the ambit of insurance contracts. The reason for the restriction in the practical application of the section may be, as suggested by KZN, that coincidentally the decided cases only dealt with factual situations which involved insurance policies. The conclusion appears to be justified, however, that over a period of 75 years, commercial practice in this country survived without the extension of s 156 beyond insurance policies.
[47] From a purposive perspective, the question is what goal was s 156 intending to achieve? With reference to insurance policies, the effect of s 156 had been explained by this court against the background of the position in common law (see eg Le Roux v Standard General Versekeringsmaatskappy Bpk supra 1046J-1047G). At common law, a contract of indemnity brings about a contractual link between the indemnifier and the indemnified. There is no privity of contract between the indemnifier and the third party. That also holds true for an indemnity brought about by an insurance policy. The third party therefore has no direct claim against the insurer, even if the insured should be sequestrated before it could settle the third party’s claim. In the absence of s 156, the third party would in that event have to prove a claim against the insolvent estate of the insured and be content with whatever dividend is paid to concurrent creditors. The insured’s rights under the policy, on the other hand, would vest in the trustee who would claim from the insurer for the benefit of the general body of creditors. Section 156 allows the third party, as it were, to leapfrog the concursus of creditors and to claim the full amount of the insurance policy directly from the insurer.
[48] The underlying purpose of the mechanism created by s 156 is best understood when it is borne in mind, as pointed out by J P van Niekerk (‘The Scope of Application of Section 156 of the Insolvency Act: Within or Beyond the Realm of Indemnity (Liability) Insurance Contracts?’ (2010) 22 SA Merc LJ 453 at 461) that it was ostensibly introduced at the time when compensation for motor vehicle accidents was dependent on the negligent driver having liability insurance cover. In this situation there was a clear need to protect the victims of motor vehicle accidents in the event that those against whom they had their claims became insolvent.
[49] One thing that is clear about the meaning of s 156, however, is that it is not limited to motor accident insurance, but that it at least extends to all liability insurance policies. In this light, KZN argued, the policy consideration as to why the third party should be allowed to leapfrog the insolvent estate of an insured, apply with equal force to an indemnity provided otherwise than by way of insurance. I do not believe, however, that this is necessarily so. In the present context, the essential elements of an indemnity insurance policy, as I see it, are these:
(a) The insurer undertakes, in return for the payment of an agreed premium, to pay a certain amount to the insured in the uncertain event of the latter incurring liability of a circumscribed kind to a third party.
(b) The insurance is, at least partly, for the benefit of the third party and not for the benefit of the insured’s creditors who fall outside the circumscribed category.
(c) The policy is a discreet contract of which the undertaking by the insurer constitutes the main purpose.
[50] Analysed in this way it is apparent that the application of the common law principles would in the event of the insured’s sequestration destroy the whole purpose of the indemnity insurance. Though the insurer still has to pay, the third party, who was intended to benefit, is left with the cold comfort of a concurrent claim. Conversely, other creditors who were not intended to benefit from the insurance will receive a windfall by sharing in the proceeds of the policy. The same considerations of policy do not necessarily apply outside the field of insurance. As I see it, this is illustrated by the facts of this case. Venfin received no separate benefit in exchange for the indemnity it gave. The undertaking to indemnify Fibalogic in the event of a claim by KZN, was not the main purpose of a separate agreement. It formed part of a much larger transaction. The undertaking was not intended for the benefit of KZN – with whom Venfin and Fibalogic were already at loggerheads at the time – but solely for the benefit of Fibalogic. In terms of clause 12.8 (referred to in para 38 above) Venfin’s liability would only arise once Fibalogic was actually obliged to pay. Because KZN did not proceed with its claim against Fibalogic, the liquidator of the latter will not have any claim against Venfin. In sum, the application of common law principles to these facts therefore leads to an end result which accords with the purpose of the undertaking. Venfin, who received no separate benefit for giving the undertaking, does not have to pay. The other creditors of Fibalogic receive no windfall. KZN, who was not intended to benefit from the undertaking, is no worse off. Its position is the same as that of any other unsecured creditor against the company in liquidation. Conversely, if Venfin were to be held liable to KZN under s 156, it would mean that the section bestowed a windfall of security on KZN for which it had never bargained nor paid.
[51] Lastly, from a policy perspective, I find it significant that in the other jurisdictions referred to by KZN in argument, legislative measures akin to our s 156 are limited to indemnities brought about by insurance policies. In those countries the legislatures therefore concluded that policy considerations do not require these measures to extend beyond the sphere of insurance. Apart from considerations of policy, I am swayed towards the interpretation of s 156 contended for by Venfin by a departure from the well recognised premise that an ambiguous statutory provision, such as this, must be construed in a way that causes the least interference with common law principles.
[52] This leads me to the conclusion that the court a quo was right in dismissing KZN’s claim based on s 156 and by ordering it to pay the costs resulting from this issue.
[53] A further consequence of the finding that KZN’s reliance on s 156 was unfounded is that its counterclaim should have been dismissed with costs. As I see it, the result of all this is that, while the appeal should be partly successful – to the extent that it results in the dismissal of KZN’s counterclaim – the cross-appeal should fail. In both cases I can see no reason why costs should not follow the event and why it should not be inclusive of the costs of two counsel. Save for the following reservation: I believe that, but for the unsuccessful part of Venfin’s appeal relating to the dismissal of its claim in convention, no more than 10 per cent of the record would have been required for the proper adjudication of the matter by this court. In this light, Venfin’s costs pertaining to the record recoverable from KZN should, in my view, be restricted to 10 per cent.
[54] The following order is made:
1. The appeal is upheld. The order of the court a quo is set aside and replaced with the following:
“(a) The plaintiff’s claim is dismissed with costs, including the costs of two counsel.
(b) The defendant’s counterclaim is dismissed with costs, including the costs of two counsel.”
2. The respondent is ordered to pay the appellant’s costs of appeal including the costs of two counsel, but the costs pertaining to the record is restricted to 10 per cent thereof.
3. The cross-appeal is dismissed with costs.
……………………….
F D J BRAND
JUDGE OF APPEAL
Counsel for Appellants: J G Dickerson SC
A Smalberger
Instructed by: Edward Nathan Sonnenbergs
DURBAN
Correspondents: Matsepes Inc
BLOEMFONTEIN
Counsel for Respondent: G Harpur SC
U Lennard
Instructed by: Lockhat & Associates
DURBAN
Correspondents: Lovius Block Attorneys
BLOEMFONTEIN