South African Bank of Athens v 24 Hour Cash CC (A3027/2016) [2016] ZAGPJHC 217 (11 August 2016)

11REPUBLIC OF SOUTH AFRICA

IN THE HIGH COURT OF SOUTH AFRICA

GAUTENG LOCAL DIVISION, JOHANNESBURG

CASE NO: A3027/2016

DATE: 11 AUGUST 2016

In the matter between:

The South African Bank of Athens.........................................................................................Appellant

And

24 Hour Cash CC...................................................................................................................Respondent

Judgment

Van der Linde, J:

Introduction

[1] This is a case about a longstanding customer of good credit who walks into its bank to get a guarantee that it could present to its landlord as rental security. The bank duly issued the guarantee a few days later. In due time the guarantee was no longer needed. The customer told the bank this.

[2] However, and this is where the dispute comes in, the customer also said that it wanted back the money that it had paid in cash to the bank as security for the guarantee to be issued. The bank said the customer never paid any cash, and that the guarantee was issued on the basis of the customer's good credit with the bank. The central factual issue before the court a quo, and on appeal before this court, is where in this tale the probabilities lie.

[3] The customer's version is that its member, Mr Gouskos, walked into the bank, spoke to the manager a Mr Sardinah, and got told by him to deposit the required amount into a certain account. This Mr Goukos did, gave the deposit slip to Mr Sardinah, and was told to collect he guarantee a few days later. When he collected the guarantee it was not signed by Mr Sardinah but by someone else purporting to be the bank manager.

[4] Some three months later, when the guarantee had to be increased by a marginal amount, the process repeated itself, the increased guarantee being R67 670. The same individuals were involved then. This time the guarantee was in fact signed by Mr Sardinah in his capacity as bank manager.

[5] Also, it should be mentioned that five years earlier the guarantee was first established, but here the facts are different. The guarantee was for some R16 135, no cash was deposited, and Mr Sardinah was not involved. The cash was simply transferred from the customer's current account to a secure call account, on the back of which the initial guarantee was issued. As pointed out above, on no version is this what happened in December 2008 and March 2009.

The factual dispute

[6] In civil cases the measure of proof is a preponderance of probabilities. When the two competing versions intersect the question of credibility comes into play as well. Take the present matter. The narrow ambit of the facts is whether or not the customer was asked to deposit R51 235 on the December 2008 occasion. It is as narrow as that, because it stands fast that in December 2008 the customer asked for and was granted the guarantee, increased by R51 235. Two derivative factual issues are contested.

[7] First, that Mr Sardinah was involved. The bank contested this, saying that as a hard fact Mr Sardinah only came to the bank the next year. And it would seem that, as insistent as Mr Gouskos was on the point that he dealt with Mr Sardinah, the latter was only appointed bank manager at this branch in 2009. Does it follow, as was argued by the bank, that Mr Gouskos’ version must be rejected?

[8] The second subsidiary issue was whether the guarantee was for cash or for credit. The customer says it was for cash, and that he was given a specific account number, which was not his current account number, into which was to deposit the money, which he duly did. The bank says that they have no record at all of such a deposit having been made, and so they say the guarantee must have been issued on the basis of the customer's good credit alone, unusual as this was. Again, does it follow, as was argued by the bank, that Mr Gouskos’ version must be rejected?

The approach to assessing evidence

[9] We return to these questions after having first discussed the correct approach to the assessment of the evidence. A helpful discussion of this topic appears in the judgment of Van der Spuy, AJ in Selamolele v Makhado,[1]:

“The onus of proof and the legal requirements as to the discharge thereof

It is common cause that plaintiff bears the overall onus of proof, ie he must prove his version that he was pushed from behind and did not fall fortuitously backwards after a scuffle with defendant. It may be that defendant has some duty of adducing evidence in support of the latter version but the onus of proof in the overall case never shifts and remains on plaintiff. See Pillay v Krishna 1946 AD 946 at 952-3. A disagreement arose between counsel for the two parties, ie Mr Botha for plaintiff and Mr Pieterse for defendant, concerning that approach which I should adopt when determining whether plaintiff has discharged the onus of proving his version on a balance - preponderance - of probabilities. The disagreement arises from the well-known statement of the law in National Employers' Mutual General Insurance Association v Gany 1931 AD 187 at 199:

'Where there are two stories mutually destructive, before the onus is discharged, the Court must be satisfied that the story of the litigant upon whom the onus rests is true and the other false.'

Mr Pieterse submits that this case presents 'a classic example of two mutually destructive versions. The one excludes the other.' With that submission I agree. Mr Pieterse then quotes cases in which the dicta in Gany's case have been interpreted, ie Koster Koöperatiewe Landboumaatskappy Bpk v Suid-Afrikaanse Spoorweë en Hawens 1974 (4) SA 420 (T) at 426 and African Eagle Life Assurance Co Ltd v Cainer 1980 (2) SA 234 (W) at 237 - 8, the latter case being cited with apparent approval but without comment in AA Mutual Insurance Association Ltd v Manjani 1982 (1) SA 790 (A) at 793G - H. He then submits that where there are probabilities, inherent or otherwise, the Court decides on the balance - preponderance - of probabilities. At the end of the day, so he contends:

"The question to be decided will always be: which of the versions of the particular witnesses is more probable considering all the evidence that was led by plaintiff and defendant and all their respective witnesses as well as all the surrounding circumstances of the case.'

It is clear to me that defendant's submissions tend to overemphasise a scrutiny of the probabilities of the matter as against findings on credibility. Mr Pieterse no doubt had in mind what Coetzee J stated in the case of Cainer (supra at 237F): 'Where there are probabilities, inherent or otherwise, there is no room for this approach.' (The Gany approach.) But one must be careful not to interpret those remarks as signifying that the Court's function of discerning the truth or falsity of witnesses' evidence becomes unnecessary where probabilities exist or less important when looking at the probabilities. One must not lose sight of the earlier

conclusion of the same learned judge at 237H when dealing with mutually destructive versions:

The position is simply that there is no proof, by any criterion, unless one is satisfied that one witness (sic witness's) evidence is true and that of the other is false.'

Ultimately the question is whether the onus on the party, who asserts a state of facts, has been discharged on a balance of probabilities and this depends not on a mechanical quantitative balancing out of the pans of the scale of probabilities but, firstly, on a qualitative assessment of the truth and/or inherent probabilities of the evidence of the witnesses and, secondly, an ascertainment of which of two versions is the more probable. See Maitland and Kensington Bus Co (Pty) Ltd v Jennings 1940 CPD 489 at 492 where Davis J stated:

For judgment to be given for the plaintiff the Court must be satisfied that sufficient reliance can be placed on his story for there to exist a strong probability that his version is the true one.'

(Italicised by me.) As pointed out by Clayden J. in International Tobacco Co (SA) Ltd v United Tobacco Co (South) Ltd (1) 1955 (2) SA 1 (W) at 13 - 14:

Though a "strong probability" may be less than "absolute reliance" it seems with respect that an unnecessary adjective has been introduced.'

It would therefore be correct for me to say that in order to give judgment for plaintiff I must be satisfied on adequate grounds that sufficient reliance can be placed on the story of the plaintiff and his witnesses, showing that their version is more probable than that of the defendant. But one still has to go through the process of considering the credibility of the witnesses and of assessing their weight or cogency and after these processes have been completed

what is being weighed in the "balance" is not quantities of evidence, but are probabilities arising from that evidence and all the circumstances of the case'.

See Hoffmann and Zeffertt SA Law of Evidence 3rd ed at 411. When, on the other hand, Mr Botha submits that

'in view of the overwhelming evidence tendered by plaintiff and in my view the unreliability of the evidence preferred by the defendant, the Court is not called upon to consider inherent probabilities in the matter other than those borne out by the evidence'

his submission is also not entirely acceptable. Because, in the process of determining credibility, the Court is charged with the investigation not only of seeking demonstrations of falsehood in the evidence of an individual witness but of weighing the inherent probabilities in his/her evidence showing that it may be false, though these improbabilities would, of course, be revealed by the circumstances of the case as a whole. But, as I have said, Mr Pieterse in emphasising the 'probabilities favouring defendant's version' might also have underestimated the Court's duty of examining the credibility of the witnesses on both sides. I must say something about the balance of probabilities or the preponderance of probabilities argued by both counsel. It is of course clear that the Court is not engaged at the end of the day in a mere mechanical process of balancing out the number of acceptable witnesses on the one side and the other because

'the object of the law is, or ought to be, to secure the sequence of I certain results upon certain objective facts'.

See Wigmore Evidence (1981 ed) para 2498. As to the degree of probability that is sufficient for plaintiff to discharge the onus, see the remarks of Denning J in Miller v Minister of Pensions [1947] 2 All ER 372 (KB) at 373 cited in Ocean Accident and Guarantee Corporation Ltd J v Koch 1963 (4) SA 147 (A) at 157D. If the acceptable evidence is such that I can safely say 'I think that it is more probable than not' the burden is discharged, but if the probabilities are equal, it is not.”

[10]The court a quo referred to the judgment in the Supreme Court of Appeal in Stellenbosch Farmers’ Winery Group Ltd and Another v Martell Et Cie and Others.[2] In that case Nienaber, JA summarised the legal position as follows:[3]

“[5] On the central issue, as to what the parties actually decided, there are two irreconcilable versions. So, too, on a number of peripheral areas of dispute which may have a bearing on the probabilities. The technique generally employed by courts in resolving factual disputes of this nature may conveniently be summarised as follows. To come to a conclusion on the disputed issues a court must make findings on (a) the credibility of the various factual witnesses; (b) their reliability; and (c) the probabilities. As to (a), the court's finding on the credibility of a particular witness will depend on its impression about the veracity of the witness. That in turn will depend on a variety of subsidiary factors, not necessarily in order of importance, such as (i) the witness' candour and demeanour in the witness-box, (ii) his bias, latent and blatant, (iii) internal contradictions in his evidence, (iv) external contradictions with what was pleaded or put on his behalf, or with established fact or with his own extracurial statements or actions, (v) the probability or improbability of particular aspects of his version, (vi) the calibre and cogency of his performance compared to that of other witnesses testifying about the same incident or events. As to (b), a witness' reliability will depend, apart from the factors mentioned under (a)(ii), (iv) and (v) above, on (i) the opportunities he had to experience or observe the event in question and (ii) the quality, integrity and independence of his recall thereof. As to (c), this necessitates an analysis and evaluation of the probability or improbability of each party's version on each of the disputed issues. In the light of its assessment of (a), (b) and (c) the court will then, as a final step, determine whether the party burdened with the onus of proof has succeeded in discharging it. The hard case, which will doubtless be the rare one, occurs when a court's credibility findings compel it in one direction and its evaluation of the general probabilities in another. The more convincing the former, the less

convincing will be the latter. But when all factors are equipoised probabilities prevail.”

Assessing the evidence

[11]One takes from these dicta then the cue that where versions collide, the three aspects of credibility, reliability and probability are intermixed, and all three must be examined. This endeavour is not to be equated with box-ticking; the constituent parts of the exercise are indicated merely to underscore the breadth of the field to be covered. The focal point of the exercise remains to find the truth of what had happened; these considerations are markers along the way.

Credibility

[12]Starting then with credibility, on appeal we suffer from the disadvantage of not having been steeped in the trial, and not having had the advantages of the trial court in seeing the witnesses testify. In this context the Constitutional Court has most recently in Makate v Vodacom (Pty) Ltd[4] restated the position.

[13]Trial courts have advantages that appeal courts do not have. The former are steeped in the matter; they observe witnesses, and are able and required to assess probabilities as they manifest within the circumstances prevailing, and as they apply to the particular witnesses testifying. In the result, unless the factual findings of the trial court are clearly wrong, or unless the trial court will have misdirected itself, those findings are not to be upset on appeal.

[14]It is instructive to quote the relevant portion of the judgment of Jafta, J in Makate:

“[37] In these circumstances, interference with the factual findings made by the trial Court is neither necessary nor justified. Ordinarily appeal courts in our law are reluctant to interfere with factual findings made by trial courts, more particularly if the factual findings depended upon the credibility of the witnesses who testified at the trial. In Bitcon, Wessels CJ said:

‘[T]he trial judge is not concerned with what is or is not probable when dealing with abstract business men or normal men, but is concerned with what is probable and what is not probable as regards the particular individuals situated in the particular circumstances in which they were.’

[38] In our system, as in many similar systems of appeal, the cold record placed before the appeal court does not capture all that occurred at the trial. The disadvantage is that the appeal court is denied the opportunity of observing witnesses testify and drawing its own inferences from their demeanour and body language. On the contrary, this is the advantage enjoyed by every trial court. Hence an appeal court must defer to the trial court when it comes to factual findings. In Powell & Wife, Lord Wright formulated the principle thus:

‘Not to have seen the witnesses puts appellate judges in a permanent position of disadvantage as against the trial judges, and, unless it can be shown that he has failed to use or has palpably misused his advantage, the higher court ought not to take the responsibility of reversing conclusions so arrived at, merely on the result of their own comparisons and criticisms of the witnesses and of their own view of the probabilities of the case.’ ”

[15]The court a quo was favourably impressed with both the respondent’s witnesses. The court considered that Mr Gouskos gave his evidence in a clear and concise manner, and made appropriate concessions. It held that the cross-examination did not discredit him; and that his evidence contained no inconsistencies or contradictions. The court also held that Mr Fritz corroborated Mr Gouskos on material aspects, and that he too was a credible witness.

[16]Testing the credibility on appeal, one asks why the respondent’s witnesses would have wanted to fabricate the version to which they testified? One knows that their deliberately falsely implicating Mr Sardinah would hold no advantage for them, because they must have considered that he would be available to testify for the bank. Far better for them, if their version was a deliberate fabrication, to say that they cannot remember the name of the person with whom they spoke.

[17]Yet they steadfastly insisted that the person they dealt with was Mr Sardinah. This conduct is more consistent with a person who is convinced about the accuracy of his (albeit incorrect) recollection, as one has so often experienced, than with a person deliberately fabricating a version.

[18]Likewise, the evidence about the arrangement being for cash rather than on credit: if the respondent’s witnesses were deliberately fabricating the cash version, they would have had to have catered for the possibility that the bank will produce documentary evidence corroborating the fact that internal contemporaneous notes would prove a credit arrangement. They did not, and no such rebutting evidence was forthcoming.

[19]The court a quo was less favourably impressed with Mr Sardinah. Although the court held that he had testified in a clear and concise manner, it held that he was not independent and objective. He was held to appear biased and gave evidence on issues on which he did not have knowledge.

Reliability

[20]Moving on to reliability, the respondent’s witnesses were obviously suffering from the expected minor memory lapses, having regard to the passage of time. As a general proposition, however, it bears remarking that the two of them were testifying to events in which they were directly involved, whereas the appellant’s witness was, on his evidence, not there. The respondent’s witnesses may therefore be expected to provide more reliable evidence than the absent witness of the bank.

[21]In fact, not only did Mr Sardinah testify to events at which he was not, but he also distanced himself professionally, saying that he was a salesperson, and that others in the bank would have undertaken the preparatory work necessary for the issuing of the guarantees. He could not explain the absence in court of the documentation, although he vouched for their existence.

Probabilities

[22]Turning finally then to probabilities, these must be examined in view of the two subsidiary questions identified above, being whether Mr Sardinah was involved in December 2008, and whether that guarantee was for cash or on credit. Concerning Mr Sardinah’s involvement, its relevance to the issue is really only as to the credibility of the respondent’s two witnesses, because there must have been a person who on behalf of the bank interviewed and interacted with the two gentlemen when they came asking for the increased guarantee.

[23]There is also a complicating dimension on this issue, which is that the fact that Mr Sardinah was not then employed by the bank, was not taken up in cross-examination of the respondent’s witnesses. In argument before us, Mr Peter, SC who appeared for the bank, argued that in fact it was taken up in cross-examination.

[24]With respect, the passages relied on by counsel deal with the identity of the person who executed the guarantee on behalf of the bank, and not with the question whether it was at all conceivably possible for Mr Sardinah to have interacted with the two witnesses. That is important, because it is possible that an exculpatory explanation for the memory lapse, or incorrect recollection, might have been forthcoming.

[25]Indeed, the argument now advanced is not that someone other than Mr Sardinah had signed the guarantee; that was never an issue. The argument now advanced is that the respondent’s two witnesses are dishonest, because Mr Sardinah was in fact not there in December 2008. This is not permissible; as was held in Small v Smith:[5]

“It is, in my opinion, elementary and standard practice for a party to put to each opposing witness so much of his own case or defence as concerns that witness and if need be to inform him, if he has not been given notice thereof, that other witnesses will contradict him, so as to give him fair warning and an opportunity of explaining the contradiction and defending his own character. It is grossly unfair and improper to let a witness's evidence go unchallenged in cross-examination and afterwards argue that he must be disbelieved.”

[26]Since on anyone's version there must have been a body representing the bank when the customer walked in there that December day to get an increased guarantee, there was at least potentially a witness available to the bank to come rebut the version that the guarantee was cash based. Yet the bank was reticent to offer to the respondent’s witnesses in cross-examination and to the court in argument who that person might have been.

[27]That introduces the probabilities surrounding the second issue, being whether the guarantee was for cash or for credit. The respondent’s two witnesses gave direct evidence that it was for cash. The bank set out to rebut that version. They did so by testifying that the only accounts within their system into which the cash deposits could conceivably have gone, do not reflect the money.

[28]But Mr Sardinah explained that in the case where cash deposits are received as security for a guarantee, back-office, in this case head-office, would take steps to ensure that the funds will have been transferred into the appropriate security account before the guarantee would be issued.

[29]He explained too that issuing a guarantee without money received as security is not common practice; but when this does occur, the same process is followed by back-office, meaning that there would have been back-office credit assessments in place and necessary authority would have been given; “you cannot lend money to anybody without having that proof,” he said.

[30]The bank did not discover such documentation. The submission on appeal in response to the absence of this material was that a claim for repayment of a deposit was not the pleaded case; the pleaded case was encashment of a guarantee. I am not sure where this submission goes, in the end. If the case were simple encashment of a guarantee, and the defence was that the guarantee had not been given for cash but on credit, the evidence required to substantiate that defence would be the same. Then too the bank would have had to have discovered the documentation substantiating the credit afforded.

[31]In any event, the particulars of claim, even if they could have been drawn better, do convey that the respondent had deposited money for the guarantee, no longer requires the guarantee, and now wants the money back. On the bank’s plea, those allegations were squarely disputed, so that whether or not moneys had been deposited in December 2008 was an issue defined for trial determination.

Concluding discussion

[32]Where does a consideration of the three issues of credibility, reliability and probability then leave one? The respondent’s witnesses were credible, and so too Mr Sardinah. But the respondent’s witnesses testified to events which the bank’s witness was not in a position personally to rebut, apart of course from his own alleged involvement. But, as pointed out above, whether or not he was personally involved does not discredit the respondent’s version.

[33]From a reliability perspective, the tally also favours the respondent’s witnesses, who were on the scene, and not like Mr Sardinah, who was not at the coalface. The probabilities are, at an objective level, in favour of the respondent, if only because issuing a guarantee on the back of cash is the norm, not the exception, as is the case of issuing a guarantee on credit.

[34]One must accept that had a deposit been made, a paper trail would have followed as night follows day; but then again, the approval of a guarantee on credit would also have left a paper spoor, as night follows day. Both are absent from the evidence; but that is not to say that either does not exist. Specifically, I am not convinced that the bank has produced evidence of the absence of a deposit, as was submitted by Mr Peter, for a reason mentioned below.

[35]Although an attempt was made to prove the absence of a deposit, no attempt was made to prove the presence of credit approval. No explanation was offered through the bank’s witness as to why it was not, particularly since the bank was able to produce what it said were the accounts into which money would have been deposited. The bank cannot then begrudge the fact that an inference is drawn that proof positive of credit approval does not exist.

[36]All that then really remains of the probabilities is that guarantees are rather issued on the strength of security than on credit. That obviously favours the respondent’s version.

[37]On the question of whether the bank produced evidence of the absence of a deposit: is it possible that the bank made a mistake? That an account number was given to the respondent’s witnesses that belonged to another customer?

[38]But whether or not the bank made a mistake, the concluding words of Nienaber, JA quoted above seem apposite: “The hard case, which will doubtless be the rare one, occurs when a court's credibility findings compel it in one direction and its evaluation of the general probabilities in another. The more convincing the former, the less convincing will be the latter. But when all factors are equipoised probabilities prevail.”

[39]In this matter credibility ultimately tips the scales. Mr Gouskos’s evidence cannot, in the light of the findings a quo, and having regard to the considerations raised above, be rejected without more. On the other side of the scales, there is the bank’s having put up the accounts into which they say the deposit would have been received had it been paid.

[40]But two potential flaws remain in the persuasiveness of that rebuttal evidence: first, where is the evidence about the guarantee being for credit? And what if the moneys were actually deposited into another customer’s account?

[41]In the result the appeal cannot succeed. Interest on the judgment debt must run from three days after the receipt of the demand of 7 October 2013. Leaving seven days[6] for receipt, the date of mora would be 21 October 2013. The balance currently standing to the respondent’s credit in the call account must be deducted, since this formed part of the R67 670.49. I have deducted R18 177.23, being the balance at 31 October 2013, from the capital amount on which mora interest is to run.

[42]In the result the following order is made.

(a) The appeal against the judgment in the capital amount of R67 670.49 is allowed.

(b) The remainder of the appeal is dismissed, with costs.

(c) The judgment of the court a quo is deleted, and the following is substituted for it:

“Judgment is entered against the defendant in favour of the plaintiff,

(i) in the amount of R67 670.49, less the amount currently standing to the credit of the plaintiff in defendant’s books under call account 20000153379; plus

(ii) interest on R49 493.26 at 15,5% p.a. from 21 October 2013 to 31 July 2014; at 9% p.a. from 1 August 2014 to 29 February 2016; and at 10.25% p.a. from 1 March 2016 to date of payment; plus

(iii) costs of suit.”

WHG van der Linde

Judge, High Court

Johannesburg

I agree.

KLaM Manamela

Acting Judge, High Court

For the appellant: Adv. J. Peter, SC

Instructed by: Hirschowitz Flionis Attorneys

First Floor, Upper Ground Level

8 Arnold Road

Rosebank

Tel: 011 880 3300

Ref: Mr A Flionis/sp/bg/T325.09

For the respondent: Adv. D.J. Winterton

Instructed by: KG Tserkeris

Unit 1, Ground Floor

280 Kent Avenue

Ferndale

Tel: 011 285 3500

Ref: D Tserkezis/sr/24 HOUR

Date argued: 2 August, 2016

Date of judgement: 11 August, 2016

[1] 1988(2) SA 372 (V), at 374.

[2] 2003(1) SA 11 SCA.

[3] At [5].

[4] [2016] ZACC 13 at [37] to [41], the judgment in the Please Call Me case, handed down on 26 April 2016.

[5] 1954 (3) SA 434 (SWA), at 438.

[6] Compare s.7 of the Interpretation Act 33 of 1957.