This appeal concerns a dispute between a spread betting operator and its customer concerning the margin arrangements applicable under its terms of business. It is primarily a matter of contractual construction.

Spread betting is not so much or not merely a bet, although it can be described as such, as a form of contract for differences. It enables a customer to take a position on a market (or an event) for a very small stake. Thus if the Dow Jones index is, say, at 10,000, one can “buy” or “sell” the market at a spread around the index of, for the sake of example, 10 points either way, 9990 to 10010. If one buys, one is betting that the market will rise above 10010. If one sells, one is betting that the market will fall below 9990. If one buys and the market rises, one stands to gain £1 for every point that the index exceeds 10010. If one sells and the market falls, one stands to gain £1 for every point that the index drops below 9990. If, however, one calls the market wrong, then one will stand to lose £1 for every point that the index exceeds the spread point in the wrong direction. Thus if one sells at 10,000 with a sell spread point at 9990, one will make £1 for every point the market falls below 9990 and lose £1 for every point the market rises above 9990. Until the bet or “trade” is closed, the gains and losses are merely “running” gains or losses. They are real enough, but constantly changing with every change in the index, and have not yet been fixed. Closing the bet will fix the position, win or lose. Unlike a classic bet, the customer can of course lose more than his stake. Indeed, on the example given, of a sale spread point of 9990 when the market is at 10,000, if the market does not move an inch, the customer will lose £10 for every £1 staked. Nor, again unlike a classic bet, are his winnings fixed at the outset by an agreement on odds. In theory winnings based on rising markets are infinite (in practice of course they are not) and losses based on falling markets are limited only in so far as they cannot exceed the consequences of a fall in the index to zero.

Normally, of course, to gain by £1 for every rise (or fall) of a single point in a stock market index such as the Dow Jones would take an investment of significantly more than £1. In effect, one’s £1 bet commands a position in the market significantly greater than the stake. In other words, there is a large element of gearing in the trade, and the situation is correspondingly volatile. Where the market in question is itself in a volatile phase, the risks become even greater. Thus, if the Dow Jones is capable of moving within a range of 100 or 200 points in a single day, the customer can be £100 to £200 richer or poorer per £1 stake within a matter of hours of his trade. On a trade of £100, those figures become £10,000 to £20,000.

The spread betting operator who accepts these trades does not bet against the customer, but lays off the trade elsewhere. Ultimately, I suspect, the trade is accumulated in some form of derivative transaction on a futures exchange, but I do not know. The operator, however, by laying off the bet elsewhere seeks to profit by means of the spread. The means by which it does that, and the terms on which it does that, however, are not a matter for the operator’s customer: nor, in the present case, have the applicable terms been disclosed.

If the customer’s trade is efficiently laid off, the spread betting operator does not retain a market risk, but, since its customer is open to volatile swings and losses which are potentially out of all proportion to his initial stake, it does retain a credit risk, which it has to be able to monitor closely. Typically, it seeks to limit that risk by controlling the level of its customers’ trading and by taking security for its customers’ exposure.

Such security, or margin arrangements, may take two forms, responding to two kinds of risk. Even at the outset of a trade, indeed at the outset of a relationship, the operator may require funds to be deposited with it as security for the customer’s potential losses. The size of such a deposit may reflect, of course, the level of the customer’s trading and also the volatility of a market in which that trading takes place. The more volatile the market, the greater can be the potential losses. Secondly, security for running losses already incurred in open trades may be required.

It will immediately be obvious that these two forms of security could either overlap or be accumulated. If the operator wishes to have the maximum security possible to be available at all times, then it will ensure that it always keeps in hand security for future potential losses: and, entirely separately, will ensure that each day, or perhaps even several times a day, it will demand further security for any running losses incurred in the course of the day on open trades. In that way, it will, as far as it can, end each day fully secured for such running losses and in addition will be able to begin each new day secured in advance for further potential losses yet to be incurred. On that basis, these two forms of security are cumulative, or, to put it another way, the security to be provided in advance merely as a condition of trading is always to be kept entirely insulated from being used as security in respect of running losses.

On the other hand, such demands may kill the goose that lays the golden eggs. If the customers are required to put up too much security, they may decline to trade, or be unable to trade. Moreover, on one view, such cumulative security, although perfectly understandable from the point of view of the operator, can also be seen as a form of double counting: since the running losses, if they occur, are precisely the contingency against which the advance security is taken, to take further security when such losses occur, save to the extent that the advance security fails to match such losses, may smack of greediness, or at any rate of seeking a second bite at setting the terms of trade in the midst of a deal. In the meantime the operator can always demand that it be fully secured against running losses as they occur.

The issue in this case is whether, on the applicable terms of trade, these two forms of security were entirely separate and cumulative, as the operator submits and the judge found, or overlapping, as the customer submits. The issue has, of course, to be determined as a matter of the construction of the parties’ contract.

On either view of this issue, one question which arises is as to the extent of the advance security required. This is settled by a concept known as “notional trading risk” or NTR. It may vary from index to index and from time to time, as well as from operator to operator. It is expressed as a factor of the amount staked. Thus a NTR of 100 means that an operator will in principle require a customer who stakes £1 on the relevant index to provide security of £100 as a condition of placing the bet.

An operator, much as he desires security, may also be willing to extend credit to a customer. On this basis, the amount of credit allowed will stand in place of security. The existence of such credit illustrates the balance which an operator has to be prepared to set between maximising security and encouraging business. Thus a customer with a credit limit of £1000 will be allowed, subject to further agreement, to stake a maximum £10 on a bet on an index with a NTR of 100. If the customer wishes to stake £20, he will either have to negotiate an increase in his credit limit to £2000 or will have to provide £1000 of security up front.

The credit allowed to and/or the security provided by a customer may be said to constitute his trading limit. He can enter into trades which, when multiplied by the applicable NTR, do not exceed his trading limit. Or, putting the matter another way, if an operator allows its customer to place a bet which exceeds his then trading limit, the operator may be able to stipulate for the right to require the customer to secure the level of his trading, by reference to the NTR, by calling for subsequent security.

The customer is Dr Vijay Battu. At the relevant time, in September 2002, he was experienced in spread betting, and had used a number of operators to trade with. He was concentrating on the Dow Jones futures index (the “DJ index” or “index”), which he was betting to fall. He is the defendant in these proceedings and here the appellant.

The operator is Spreadex Limited, who provide spread betting facilities on sports events and financial indices. Their customers’ trading debts are enforceable by virtue of section 412 of the Financial Services and Markets Act 2000. They were regulated by the SFA, now the FSA. In these proceedings they are claiming just under £25,000 from Dr Battu. They are respondents to this appeal.

On 19, 23 and 24 September 2002 Dr Battu opened five trades with Spreadex, for a total of £500, each of them selling the DJ index. The first three trades were for relatively small amounts, namely £50, £30 and £20. The fourth trade (on 23 September) was for £200, and the fifth trade (on 24 September) was for £200 as well. The bets were fixed when the DJ index was respectively 7908, 7900, 7881, 7800 and 7672. Thus, as the index declined, Dr Battu was doing well.

At this time Dr Battu had a credit limit of £20,000 with Spreadex and Spreadex were setting a NTR for the DJ index of 120. It is said by Spreadex that this NTR was a computer error. In November 2000 the NTR had been at 80 (or at any rate Dr Battu had been personally informed that he could use his credit limit at the rate of £12.50 per Dow Jones point for every £1000 of credit, which is equivalent to a NTR of 80). In January 2002, however, a general Spreadex flyer had said that the DJ index NTR was 160, but Dr Battu disputed that this applied to him or qualified the offer made to him in the letter of 9 November 2000.

Spreadex say that at the time of the opening of the fourth trade, for £200, on 23 September, Dr Battu had a credit limit of £20,000, a positive cash balance on previously closed transactions of £3,840, and a running profit on his three previous open trades of £9,320, making a total of £33,160. However, at a NTR of 120, the £300 trading position required security of £36,000. Dr Battu was therefore asked to provide further security of £3,000, which he supplied then and there, over the telephone, by means of his credit card. It will be observed, however, that Spreadex’s real security was only £6,840 (the £3,840 cash in Dr Battu’s account plus the further £3,000 taken off his card). For the rest, Spreadex were relying on their agreed credit of £20,000 and the transient running profits on the open trades.

By the end of that day, however, or as at whatever point Dr Battu’s statement from Spreadex dated 23 September 2002 was compiled, the position had deteriorated. The index had now risen to 7865. The four open trades were now showing an overall running loss of £9,840. Otherwise the statement showed the cash of £6,840, the credit limit of £20,000 and the NTR (of £36,000), producing what was said to be “Available trading credit at 23/09/02” of “£-18,639.42”. Thus the headline figures at the top of the statement were set out as follows:

However, on 24 September, when Dr Battu wanted to open his fifth trade, for a further £200, the index had moved back again in his favour, for his selling spread point was at 7672, some 200 points lower than the closing figure of the previous day at 7865. At that lower level, his four open trades would have been showing running profits of some £46,000. Spreadex were willing to open the fifth trade without requiring any more security. Although the NTR on what was now a trading position totalling £500 would, at 120, have been £60,000, that was, apparently, regarded as well covered by a combination of Dr Battu’s running profits, plus credit of £20,000 plus cash of £6,840, a total of nearly £73,000. It will again be observed that at this time Spreadex’s real security remained £6,840, although they were no doubt taking comfort from his current running profits.

There is no statement dated 24 September 2002. However, the market closed at 7682, more or less in line with the selling spread point at which Dr Battu had fixed his fifth trade. At that level he continued to maintain his substantial running profits.

On 25 September the market opened at 7605, and Dr Battu was doing even better. From that point, however, the index started to climb. Spreadex have calculated that when it reached 7696, the NTR of 120 on the £500 trades, amounting to £60,000, was in balance with Dr Battu’s diminishing profits plus his credit and his cash. Above 7696, Dr Battu’s position was continuing to deteriorate by £500 per point. The market closed on 25 September at 7848. There is no statement dated 25 September 2002.

26 September was a day of crisis. The index opened at 7832. During the day it climbed as high as 7994. It did not fall below 7820. It closed at 7952.

At 09.40 on 26 September Lisa Mead, a credit and customer services supervisor at Spreadex, spoke on the telephone to Dr Battu. She told him that Spreadex were “looking for about fifty”, ie £50,000. She said:

During that morning Dr Battu paid Spreadex £29,000.

At 16.59 that afternoon, Miss Mead left a telephone message on Dr Battu’s answering machine to tell him that he had until 18.30 to “make a payment into your account of fifty thousand…to cover your open positions. Failure to do this will result in our closing all your open trades.” That would be a further £50,000 in addition to the £29,000 already received that morning.

At 17.04 Dr Battu called Spreadex and spoke to Miss Mead. He asked her to inform him of his running or current loss, which she told him was at £82,500. He told her he would close some of his positions and review the situation. She said she would transfer him to the dealing room, and after the trade he would be transferred back to her so she could tell him “how much is remaining to pay”.

At 17.11 Dr Battu closed his first three trades, totalling £100, at 7949. The losses on those closings amounted to £4,880, as a confirmation of that date addressed to Dr Battu reveals. When Dr Battu was reconnected with Miss Mead she revised her demand to £40,000, again by 18.30, otherwise his remaining positions would be closed. Dr Battu said he would do that. It was arranged that he would speak to the dealing room before 18.30.

At that time, at an index mark of 7949, Dr Battu’s position was as follows. His NTR on his outstanding trades totalling £400 was £48,000. His running losses on those trades were approximately £85,000. His credit limit remained at £20,000. His cash, less the realised losses on his first three trades, was now at £1,960, plus the £29,000 transferred that morning. That left a current deficit of approximately £34,000, not counting any security for further potential losses.

At 18.25 Dr Battu received a call from Gavin Pugh, a trader in Spreadex’s dealing room. In the meantime the DJ index had fallen back somewhat, alleviating Dr Battu’s position. Thus his open trades were now showing a running loss of £53,500, as Mr Pugh informed him (some £30,000 better than before). This was not a great deal more than the sum of approximately £51,000 which, inclusive of his credit, Dr Battu had with Spreadex. Dr Battu therefore queried why Spreadex still required a further £40,000 from him. Mr Pugh replied: “because the open positions are losing you 53,000 and the notional risk is 64,000 on those trades…The bottom line is we need 40,000 by half past six otherwise we need to close the positions out”. Dr Battu protested, asked to speak to a manager, but offered to pay £7,000. Mr Pugh said the credit managers had gone home, and that even if £7,000 was paid, he would still close out £330 of the £400 trades. Dr Battu again asked for the running loss and was quoted £54,400. In the circumstances, there was deadlock and Mr Pugh proceeded to close out the £400, against the background of Dr Battu’s protests: “I object to this, I strongly object to this”.

Mr Pugh’s reference to a NTR of £64,000 was now applying a ratio of 160 to £1. Mr Pugh was intimating to Dr Battu that the NTR came on top of running losses.

At 18.39 Dr Battu’s fourth and fifth trades were closed out at 7875.11. This crystallised a loss of £55,644 on those trades and led to Spreadex’s claim in these proceedings in the sum of £24,923.42.

The following figures were agreed on the pleadings and were common ground at trial: that the shortfall required at the time of closure to make up Dr Battu’s then running loss was £4,683 (£55,644 less cash held of £30,960 and credit of £20,000). The primary issue remained as to whether Spreadex were entitled, in addition to security to cover the running losses, further security in the amount of the NTR, whether £64,000 or £48,000.

When Dr Battu applied for an account, he was told by the application form which he signed on 17 May 2000 that his “trading limit” could be supported by credit or a permanent deposit or a combination of both. He declared that he had received, read and understood “the Rule Book, the current Sports and Financial Brochures and the Two Way Customer Agreement” and had received and read the Risk Warning Notice. He acknowledged that the transactions he would enter into “will result in you having to provide margin payments as set out in Clause 5 of our Two Way Customer Agreement”.

Since that time Spreadex had published a new, January 2001 edition of their terms. I do not know what the previous edition said. The following provisions are agreed to be applicable. They are all contained in a single booklet which contains “Rules and Agreements applicable for Sports & Financial Spread Betting” or “Account Operating Rules” (the “Rules”), the “Two Way Customer Agreement” (“TWCA”) and a “Risk Warning Notice”.

The Rules provided inter alia as follows:

In addition TWCA provided as follows:

at any time and without notice, close all or any of your bets in accordance with Spreadex Limited Rules…

Finally, the Risk Warning Notice contained the following:

Dr Battu relied on the letter addressed personally to him from Spreadex dated 9 November 2000 and Spreadex relied on the flyer of 15 January 2002. Each was relevant to the concept of NTR on the DJ index.

The letter dated 9 November 2000 was written by the managing director of Spreadex. It began by pointing out that in the last few months financial markets had been very volatile. Its second sentence reflected the balance which an operator like Spreadex had to find between encouraging and controlling its customers, for it stated:

You can increase your credit limit…

You could transfer money onto a permanent deposit account to cover extra margin when required…

You could pay margin on that part of your trade that exceeds your credit limit.”

The flyer of 15 January 2002 was not addressed personally to Dr Battu, but he accepted that he probably received it together with a statement sent to him in that month. It was headed “Happy New Year from Spreadex!!” and contained a variety of news items about Spreadex’s business, including the following:

This is, I think, the first express reference to NTR in the papers, subject to this: a statement of the previous month, dated 3 December 2001, has its line for “Notional risk” in the form cited at para 18 above. I do not know when statements in this form began to be used.

The parties have taken up opposing positions concerning the primary issue identified above: is the initial security intended to provide Spreadex with comfort with respect to future potential losses separate from and entirely additional to further security to be provided in respect of running losses, or are the two forms of security overlapping?

On behalf of Dr Battu, Mr Christopher Heather submits that the two forms of security are overlapping and are capped by the amount of the running loss. He relies in particular on the expression to be found in Rule 6 “losses in excess of the trading limit will be dealt with by way of margin payments”, where he says that the “trading limit” reflects the security against potential losses (provided either by credit or deposit) and the additional margin payments to provide for running losses are capped by the concept that such additional payments are only to be in respect of “losses in excess” of that security. That he says is inconsistent with the opposing submission that the security in respect of potential future losses and additional security in respect of the full extent of running losses are cumulative. He says that his construction is also consistent with Rule 8’s reference to the requirement of margin payments where daily recalculations show that a client “has exceeded their trading limit”.

On behalf of Spreadex, on the other hand, Mr Alexander Pelling submits that the two forms of security are wholly separate and entirely insulated from one another, and are shown to be so by being defined by their own in different language. The first kind of security, in relation to potential losses, is strictly called a “deposit” and not “margin”. The second kind of security, in relation to running losses, is called “margin” rather than a “deposit”. The terms contemplate that “deposit” can be taken at any time, and in more than one tranche: see TWCA 6’s “either when the bet is opened or at any time thereafter” and “additional deposit payments on new or existing bets”. The clearest passage relied on for the insulated and cumulative status of the two forms of security is again, Mr Pelling submits, to be found in TWCA 6, which begins by referring to deposits and then goes on, after a semi-colon which he highlights, to deal with “margin payments” which have to be “sufficient to meet the amount which, when a movement adverse to your bet has taken place, you would lose on the bet”. This he says means that the margin payment has to equal the whole of the running loss, thus proving that deposit and margin must be separate and cumulative. He also submits that this is the logic of the security by way of deposit, which he says is intended to reflect NTR, the constant potential for future loss, at any state of the market due to its volatility.

Mr Heather submits that on his construction of Spreadex’s terms, Spreadex were only entitled to be secured for (a) Dr Battu’s “trading limit”, whatever that was, plus (b) any running loss in excess of that. Since Dr Battu’s total trading losses at 18.30 on 26 September were only £55,564, his liability to secure Spreadex was capped at that figure, unless perchance his trading limit itself exceeded that figure. Even if “trading limit” reflects the same concept as NTR (an expression not to be found in Spreadex’s terms), the NTR on the open trades totalling £400, at a ratio of 120, was only £48,000, and even at a ratio of 160 amounted to no more than £64,000. At most, therefore, Spreadex were entitled to security for either £55,564 or £64,000. However, it already had cash of £30,960 and had allowed Dr Battu credit of a further £20,000, to give a total of just under £51,000. Therefore, they were only entitled to a further sum of either under £5,000 or at most some £13,000. Dr Battu had offered £7,000, more than the former sum, and in any event Spreadex had demanded £40,000 and refused to take any less. Therefore, they were not entitled under the contract to close Dr Battu’s account for any failure to provide the £40,000 demanded.

Mr Heather also had alternative submissions, based on Spreadex’s letter to Dr Battu dated 9 November 2000, but it is unnecessary at this point to go into them, save to note that the letter is relied on as curtailing Spreadex’s NTR rate in any event to a ratio of even less than 120, namely to one of 80.

If, however, Mr Pelling is correct, then Spreadex were entitled to demand a “deposit” of at least £48,000 (or even £64,000) plus “margin” of some £55,564, which cumulatively far exceeded the sum of what Spreadex already had (just under £51,000) plus the further £40,000 demanded, a total of just under £91,000.

This appeal is taken from the judgment handed down on 6 August 2004 by HHJ Chambers QC, sitting as a judge of the Queen’s Bench Division in the Cardiff District Registry, Mercantile Court. Dr. Battu appeals with the leave of the court from the judgment for Spreadex in the sum of £28,636.22.

The judgment essentially proceeds by first accepting the evidence given on behalf of Spreadex that the two forms of security are intended to be separate and cumulative and then concluding, with regard to TWCA 6, that that is indeed what Spreadex’s terms have achieved by reference to two entirely different concepts, “deposit” and “margin”. However, there is, I fear, not much analysis as to whether the two different forms of security are necessarily to be viewed as cumulative as distinct from overlapping; and also, little by way of construction of the language of the Rules and TWCA. On the contrary, the judge sees but accepts that the language used does not fit his understanding of the scheme, so that one finds (at para 19): “It is an irritating feature of this case that the expression “margin” is used promiscuously as between its correct meaning and meaning deposit”. A little later he states (at para 20): “Payment of margin would only require to be made to the extent that there was insufficient security in the hands of Spreadex after taking any credit limit into account”: an innominate reference to Rule 6, but without any discussion of a provision apparently inconsistent with his conclusion. When later he refers to NTR, he again comments (at para 29): “It is another example of the confusing habit of Spreadex in using different terms to mean the same thing and the same thing to mean different things”. He appears to have been influenced by his understanding of Dr Battu’s evidence as being that he “accepted the need both to supply a deposit and margin” (at para 30), but again without pondering whether the need for both was cumulative or overlapping.

Ultimately, the essential issue debated by counsel on this appeal is a matter of construction. Either solution would make perfectly good commercial sense. The existence of two different concepts or functions of security is not in doubt, but the question is what Spreadex’s terms have provided for.

The first thing is to consider whether under the Rules and TWCA “deposit” and “margin” are indeed two entirely separate concepts, each referred to by a special term, or both part of a single scheme for securing losses, potential or current. To such a question, it seems to me that only one answer is possible, namely the latter. A temporary deposit, which, like a permanent deposit, can support a trading limit, is itself referred to in Rule 6 as “initial margin”, and Rule 8 refers to a margin payment to cover running losses in excess of trading limit as “additional margin”. TWCA 6 is headed “Margining arrangements” and refers to both deposits and margin payments. Its second paragraph expressly states that “margin payments” outside a customer’s credit limit must be provided “by way of deposit”. Other examples could be given from Spreadex’s contractual terms, eg from TWCA 9, headed “Credit”, where the general concept juxtaposed to “credit” or “credit limit” is “margin” or “margined monies” and there is no talk of deposits.

Moreover, under Dr Battu’s signed application form, he was told that transactions would result in him having “to provide margin payments” with nothing said as to deposits. Furthermore, the Risk Warning Notice speaks of a “margining system” including “comparatively modest deposit or margin” to start with leading to the possibility, in the event of losses, of being called upon to pay “substantial additional margin”. And for good measure, the letter of 9 November 2000, which provides the immediate background to the January 2001 edition of Spreadex’s booklet of terms, refers to putting up a deposit, permanent or temporary, as a way, in addition to or as an alternative to a credit limit, of providing margin or extra margin; and the flyer of 15 January 2002 speaks of “initial margin” in the context of NTR.

It seems to me quite obvious that credit, deposits and margin payments are all forms of margin: and are all designed to give security (or, where credit is concerned, trust in lieu of security) inter alia for running losses. That is, in my view, a bad beginning for seeking to find two wholly separate concepts, insulated from one another, such that only margin or margin payments stand against running losses, while deposits (and credits) are maintained solely against the ever-constantly renewable potential for future, as distinct from current or running, losses.

Secondly, and critically, in my judgment Mr Heather is right to focus on the words in Rule 6 “losses in excess of the trading limit will be dealt with by margin payments”. If the Spreadex submission had been right, one would expect this sentence to say quite generally that “all losses will be dealt with by margin payments”. Instead it is only some losses that will be dealt with in such a way, namely “losses in excess of the trading limit”. What is the trading limit? There is no express definition, but Rule 6 states in its second sentence that it “will be based upon credit awarded or a deposit of cleared funds or a combination of both”. Thus, the customer is to understand that a trading limit is something expressed in financial terms, quantified by credit, deposit, or both. If, for example, a customer has a credit limit of £10,000 and a deposit of £10,000, then, as it seems to me, he has a trading limit of £20,000. But what does that mean? In the absence of some such notion as NTR, which weights the potential risks of various markets, the only meaning which could be given to a trading limit is a limit on losses: thus a customer cannot trade further, either in the sense of opening new trades, or even in the sense of keeping his current trades open, without providing fresh security or margin. Neither the Rules nor TWCA describe a system of NTR. But even if the NTR concept is understood to be built into, or capable of being inserted into, the concept of a trading limit, and there is some support for that view in the explanation given in Rule 6 about temporary deposits (“to reflect the likely volatility of the event”), nevertheless it is very difficult to see why running losses should not enter into the calculation of the use of a trading limit as well, especially given the language “losses in excess of the trading limit”.

In this connection, the language of Rule 8 is worth considering. The opening sentence of Rule 8 speaks of the familiar process of recalculating all open positions every day on a “mark to market” basis. The second sentence then says that such a calculation may indicate that a customer has “exceeded their trading limit”. On Spreadex’s doctrine, however, a trading limit is never affected by a movement in the market, although it could be by a change in the NTR – if such a change were permitted in the mid-stream of a bet. On Spreadex’s submission, Rule 8 is solely concerned with “margin payments” properly so called, which are designed to deal with running losses. On Dr Battu’s submission, however, a movement in the market may well cause a customer to exceed a trading limit, as running losses exceed the amount of deposit or credit available on the account. Thus Rule 8, as well as Rule 6, supports Dr Battu’s case and is inconsistent with Spreadex’s case.

Mr Pelling nevertheless submits that the “trading limit” concept is nothing but the NTR system and has nothing at all to do with actual running losses, only with the potential for such losses. All that a trading limit means is that you can enter into bets with potential losses, as measured on the NTR system, up to a certain limit, admittedly fixed by your credit, deposit or a combination of the two. For these purposes, it does not matter whether you have even made use of your trading limit or not. Running losses are necessarily something entirely different. That, however, is Spreadex’s same argument, restated. It rests on a pure, theoretical distinction between potential future losses and current running losses. It seems to me, however, that such a distinction is not a necessary one, and does not fit the language of Rules 6 and 8. There is no reason why a provision for the possibility of potential future losses cannot overlap with the incurring of running losses: the latter is merely the occurrence of the risk foreseen in the former.

Thirdly, there is the important provision in TWCA 6, much relied on by Mr Pelling, to the effect that Spreadex may demand “margin payments sufficient to meet the amount which…you would lose on the bet”. Mr Pelling submits that this means “margin payments equal to the amount of running losses” and thus is consistent only with a system in which deposits against potential losses and margin payments against current losses are entirely insulated from one another. However, this is not what TWCA 6 says, although it is what it could very easily have said. Instead, it uses the more complex formula “sufficient to meet”. Assume a trading limit of £10,000, a deposit of £10,000, a £50 bet on a NTR of 100 to 1 and running losses of £10,000, how much is “sufficient to meet” the running losses of £10,000? In a very real sense nothing, for there is already margin in place to the extent of £10,000. It is only if one starts from the assumption that the “deposit” is entirely insulated from being used to secure running losses that one can begin to think in terms of a full £10,000 being needed as “sufficient to meet” a running loss of £10,000. And yet TWCA contains nothing whatsoever to suggest that deposits are not available (to both parties) to secure running losses. Normally, one would think that deposits, under the general heading of “Margining arrangements”, would be available to secure running losses. It is only by importing the concept of a trading limit from the Rules and a NTR concept from outside both Rules and TWCA that one might think it possible that not all Spreadex’s security would be available to meet running losses. But that possibility runs straight back into the difficulty of Rule 6’s concept of “losses in excess of the trading limit”.

If, therefore, TWCA 6 had used the expression “equal to” rather than “sufficient to meet”, then it would be possible that Rule 6 (“losses in excess of the trading limit will be dealt with by way of margin payments”) and TWCA 6 (“margin payments equal to the amount which…you would lose on the bet”) would be in conflict with one another. In such circumstances, the question would arise how such a conflict could be reconciled. I am sceptical that such a conflict should be reconciled in favour of Spreadex and against their customer, especially as the Rules contain the more sophisticated analysis of the counterplay between a “trading limit” and “margin payments” in Rules 6 and 8, whereas TWCA does not speak of trading limits, but only of “margining arrangements” and “credit” (TWCA 6 and 9). As it is, however, TWCA 6 does not use the expression “equal to”, and therefore there is no prima facie inconsistency between TWCA and the Rules.

Fourthly, there is the purposive approach adopted by Mr Pelling when he stresses the function of NTR as always looking forward to future potential losses. He therefore submits that it is inherent in that function that deposits – or indeed whatever they are called, say initial margin – are kept insulated from running losses. For, at any stage of the market, there is always the potential for further loss. That is correct, but in my judgment that insight is not decisive. The NTR concept has not been expressly made a part of the Rules and TWCA. It follows that even the basic concept from which the inherent logic could be derived has not been incorporated into a contractual text. A fortiori, the inherent logic relied on has not been made express. It could have been. Moreover, that logic does not necessarily survive in a business world where there are competing tensions. As I have said above, the operator wants security, but first and foremost he wants active customers. Even in terms of the point’s own logic, the inference is not, however, necessarily compelling. The point about volatility in markets is that they churn: they can of course keep going straight in one direction, but more often they turn and turn about, as occurred in the present case. Even without complete insulation, therefore, the NTR concept achieves two purposes. (1) It provides a control mechanism on the customer’s trading: by quantifying the risk value of a bet, it provides a means by which the operator can limit the number and value of a customer’s bets and give effect to the concept of a trading limit. This is a valuable discipline where the stake of any individual bet is in itself small. (2) It quantifies the amount of security the operator desires to hold in respect of a bet in case of losses. But, if the losses go beyond the anticipated amount, it does not necessarily follow that the operator has contracted for additional security quantified in the sum of the whole of the loss as distinct from that part of the loss in excess of the security already provided.

Let me try to illustrate the problems inherent in the debate between the parties by considering, by reference to specific figures, three different possible approaches to the question of security.

(1) The first is that proposed on behalf of Spreadex. NTR, reflected in a deposit, is to be kept entirely separate and insulated from margin for running losses. Thus suppose again a deposit of £10,000, a trading limit of £10,000, and a £50 bet on a market with a NTR of 100. Running losses of £10,000 are incurred on the bet. Mr Pelling would say that those losses have to be separately met by margin in the full amount of those running losses. In addition, there remains the NTR of £5,000 represented by the £50 bet on a NTR of 100 to 1. That £5,000 NTR is within the deposit of £10,000, and so presents no problem. The deposit even allows for a further bet of £50 at a NTR of 100 to 1. However, in the absence of such a further bet, the customer gets no benefit from the (as it were) reserve of £5,000 held within his deposit. That deposit is for meeting the future potential risks of his trades, if any, not as security for running losses. It may be that the operator will be willing at various times temporarily to waive its full requirements, eg it might, as it did in this case, allow running profits to stand in the place of an initial deposit for NTR purposes: however, it is entitled to enforce its full rights as the occasion demands. This thesis emphasises the language in TWCA 6, glossed to mean that margin payments equal to the full amount of running losses are required, in addition to deposits, which are in turn understood to be devoted to NTR.

(2) The second is that proposed on behalf of Dr Battu. All payments are part of margin, the purpose of which is ultimately to secure the operator against running losses. Some such payments may be demanded in advance of the bet, when there are no running losses, because of the future possibility of such losses. That is reflected in the concept of NTR and in trading limits. But if such running losses are incurred, additional security is only required to the extent that such losses exceed the trading limit security. Thus, given (as before) a deposit of £10,000, a £50 bet at a NTR of 100 to 1, and running losses of £10,000, there is no requirement of another £10,000 or of anything by way of further margin. There are no running losses at all in excess of the trading limit of £10,000 represented by the deposit in that sum. On this doctrine, the trading limit deposit and the running losses have to overlap completely before any further margin is required. This thesis thus emphasises the language of Rule 6 and its talk of “losses in excess of the trading limit”.

(3) The third possibility is one for which neither side contended, and represents something of a compromise between their two positions. This is that NTR and running losses count against all forms of security, whether “deposit” or “margin”. Thus given, as before, a deposit of £10,000, a £50 bet at a NTR of 100 to 1, and running losses of £10,000, all that is required by way of extra margin is another £5,000: on the basis that the balance of the deposit not used for the purpose of NTR can stand as part of the additional security needed to meet the running losses. On this thesis, a deposit could perform a duel or chameleon role, both as initial margin and as additional margin.

This third possibility would be a perfectly sensible system to adopt, and would go a long way towards meeting the purposive ground-rule adopted by Spreadex here and below, namely that there should always be cumulative security for both running losses and future potential losses. However, it is not the same as the thesis for which Spreadex contended, and it leads to different results, as shown above. Moreover, although it meets Mr Pelling’s purposive approach, it is difficult to fit into the contractual language highlighted in the opposing submissions. Thus, in Rule 6, “losses in excess of the trading limit”, would have to be glossed as referring to “losses in excess of that part of the trading limit which has not been committed by [the concept of] NTR to open trades” (on my example, the balance of £5,000), which is not what the words say. And in TWCA 6, “margin payments sufficient to meet the amount which…you would lose on the bet” could only be made consistent with this third thesis if “sufficient to meet” is not given the meaning of “equal to” espoused by Mr Pelling. Nor is this third thesis consistent with Spreadex’s doctrine that a “deposit” properly so called is only available for the purposes of trading limits and NTR, as distinct from “margin payments” properly so called which are required to meet running losses. Thus, contrary to that doctrine, where part of a deposit is currently surplus to NTR and is used as part of a “margin payment” to meet running losses, the status of that security is changed, and the customer in effect no longer has a trading limit of £10,000 based on a deposit of £10,000, but only a trading limit of £5,000.

Even so, Mr Pelling’s submissions at one stage appeared to elide possibilities (1) and (3). Thus he relied on Spreadex’s statements, an example of which is set out at para 18 above, as demonstrating Spreadex’s operation, to the knowledge of Dr Battu, of possibility (1). In fact, however, such statements seem to me to demonstrate the operation of something closer to possibility (3). Thus, to work on the example at para 18: if possibility (1) had been operated, then, with a “Credit limit” of £20,000 and a “Deposit account balance” of £0, Dr Battu should not have been allowed to trade with a NTR (“Notional risk”) of greater than £20,000 (or possibly £26,840 if the “Trading account balance” of £6,840 were to be counted as a deposit). The NTR on the open trades was £36,000. “Open positions” showed running losses of £9,480. The demand should therefore have been for an additional deposit of either £16,000 (or £9,160, if the £6,840 counts as a form of deposit) plus additional margin of £9,480 (or of £2,640, if the trading account balance counts against additional margin rather than as a deposit). In fact, all forms of security/credit and all forms of liability were rolled up together, to give a single balancing figure called “Available trading credit”, here given a negative figure of “-£18,639”.

An even clearer example of a statement demonstrating the operation of something closer to possibility (3) is the statement of 27 August 2002, which reads as follows:

At times, Mr Pelling’s submissions appeared to me to rest not so much on questions of construction as on findings, or would-be findings, as to how Spreadex, with the knowledge of Dr Battu, operated their business. It seems to me, however, that, subject to a degree of background context which appears to have been uncontentious, the actual operation of Dr Battu’s account does not assist in the question of construction that has to be faced on this appeal. Even if it were permissible to seek to construe Spreadex’s terms by reference to what they did, as to which I am very doubtful, the case against Dr Battu was not based on a contract defined by conduct, nor are the findings necessary to such a case available to Spreadex.

The above discussion leads to a fifth reason which militates against Spreadex’s submission as to the effect of their terms. The examples given above demonstrate that one of the consequences of possibility (1), namely of holding deposits (for the purpose of NTR) and margin payments (to meet running losses) entirely separate, is that it may be critical to know whether a payment held or demanded by Spreadex bears the imprint of one kind of security or the other. This is especially the case since Spreadex appear to reserve the right to forgo calling up the whole of a “deposit” at the initial stage and to require the deficit to be provided at a later stage as “additional deposit”. Suppose therefore a £100 bet at a NTR of 100 to 1: that would call for a deposit (assuming no credit has been agreed) of £10,000. However, Spreadex do not require that deposit to be placed up front, perhaps because (as in the present case) Spreadex have been relying on running profits on another open trade (which, of course, are not true security). The £100 bet incurs running losses of £10,000. Spreadex are now entitled, on their construction of their contract terms, to (a) a “deposit” of £10,000 and (b) “margin payments” of £10,000, a total of £20,000. They call for a payment of £10,000, without making it clear whether they are calling for a deposit or a margin payment, and it is provided. The next day the bet moves back into profit. If the £10,000 was provided as margin, the customer is prima facie entitled to its release. If, however, it was provided as deposit, he is not.

This does not mean that if, properly construed, Spreadex’s terms required a system of insulated deposits, such a system could not have been operated. And the fact that it appears that it was not operated I assume is besides the point. It would, however, have been a complex system in which there would have been much room for misunderstanding. The status of each demand and each payment would have to be defined. The fact that Spreadex’s terms of contract, and their other documents by which they sought to implement, define or vary those terms, speak so haphazardly on the subject of deposit and margin, is completely at odds with the system which they submit is there to be found.

To return to the express language of the text, therefore, Rule 6 speaks of “losses in excess of the trading limit” and Rule 8 speaks of daily calculations which may indicate that a customer has “exceeded their trading limit”. It is in my judgment simply not possible to derive from those words the consequences for which Spreadex contend. On the contrary, they seem to me fully to support the alternative, and entirely commercial scenario, for which Dr Battu contends. If there was doubt, which at the end of the day I do not think there is, it would have to be resolved against Spreadex, who seek on the basis of their own terms either to demand payments from their customers or to close their customers’ bets without their customers’ authority.

It is next necessary to consider the consequences, for Dr Battu’s trades and for Spreadex’s claim, of the construction of Spreadex’s contractual terms which I prefer.

In my judgment Spreadex were entitled to security or margin (less the amount of his credit limit) to cover Dr Battu’s running losses. They were entitled to security or margin in the form of a deposit (less the amount of the credit limit) to cover Dr Battu’s trading. They were not entitled to both cumulatively, but to the higher of the two figures.

The figure pleaded by Spreadex for Dr Battu’s running losses at approximately 18.30 on 26 September when Miss Mead was demanding £40,000 was £54,400 (amended reply at para 10.8). This reflects the last figure for running losses quoted to Dr Battu just before his trades were closed out (see para 29 above). I am prepared to work on Spreadex’s pleaded figure. Spreadex therefore go on to allege that at that time they were entitled to demand £67,440 (made up of running losses of £54,400 plus a “NTR deposit” of £64,000 less £50,960 in credit and cash), even if they limited the demand to £40,000.

Spreadex’s claim therefore raises two further questions. One is as to the relationship of NTR and trading limits. The second is as to the appropriate NTR ratio. Subject to those questions, Spreadex were, in the light of my view of the effect of their terms, entitled at most to the higher of the deposit of £64,000 or the running losses of £54,400: a further sum (over and above the £50,960) of at most only £13,040. If the NTR ratio is 120, requiring a deposit of £48,000, then the running losses of £54,400 become the higher sum, and the further sum required becomes £3,440. If the NTR ratio is only 80, the further sum required remains at £3,340. If the figures referred to in paragraph 32 above are used, the further sums required are not materially different, viz either £4,683 or at most £13,040.

Since Spreadex demanded £40,000, their demand was therefore prima facie unlawful. This is subject to an argument raised by Spreadex in their respondents’ notice that Dr Battu had in any event bound himself by contract to pay £40,000. The judge did not deal with this aspect of Spreadex’s claim, other than to record that during his earlier telephone conversation with Miss Mead Dr Battu had promised to pay £40,000 by 18.30. I will deal with that cross-appeal below. Subject to that, it was not disputed that the demand for £40,000 was, on Dr Battu’s construction of Spreadex’s terms, which I have now adopted, unlawful. Nevertheless, Mr Pelling submitted that even so Spreadex were entitled to close Dr Battu’s trades because he had failed to pay the sum actually due, although less than £40,000, in circumstances where Spreadex were willing to accept less, indeed any sum at all. He relied in this respect on the principle that –

In these circumstances, it does not seem to me to matter what precisely was the relationship between NTR and trading limits or whether the applicable NTR was 160 or 120 (or 80). In general, however, it seems to me that Mr Heather is probably right to accept that a trading limit can be established by the level of credit and/or deposit (see Rule 6), so that the £50,960 in Spreadex’s hands at the crisis point on 26 September would support trading to the value of £50,960, measured at whatever may be the appropriate NTR. As for the appropriate rate, at the outset of the trades it had been established, as Spreadex themselves assert, at a ratio of 120 to 1, although that is said to have been an error. The question arises whether that NTR could be altered in the middle of an open trade. The judge thought that Spreadex could alter the terms of trading unilaterally at any time and without notice, relying on Rule 11. I am doubtful about that. Rule 11 itself seems to have some doubts about itself (“to the extent that these rules can be altered without the written agreement of both parties under the SFA Rules”). Normally, and as a matter of fundamental principle, a contract can only be varied by agreement: but I suppose that it is possible for parties to agree to give one party a unilateral power to alter some part of their arrangement, provided the effect of that is not simply to leave what purports to be a contract as something which is merely left in the air at the whim of one of the parties. An option is perhaps a clear example of an agreement which gives one party a right to alter a contractual arrangement, but there the variation is spelled out in the option.

Spreadex’s terms have their oddities in purporting to give Spreadex something like carte blanche (see Rules 11 and 21), but in general I would be reluctant to conclude that such rules gave Spreadex the right to alter the terms of trading in the midst of trading, as distinct from for the future. At any rate, given the possibility that an operation like Spreadex might wish to alter the terms of trading either with immediate effect on existing trades or only for future trades, I would be reluctant to construe a right to alter as applying to existing trades unless the contractual terms clearly expressed the right in such a way. Thus I would hesitate to say that Spreadex were entitled to alter the NTR in the middle of a trade, unless the contract clearly gave Spreadex that right. Otherwise Spreadex could destroy a customer’s trading in mid stream. As it is, the concept of NTR is inadequately, if at all, dealt with in the Rules/TWCA, and its relationship with the concept of trading limits, which are there dealt with, suggests that it is primarily concerned with considerations at the outset of a trade. However, it seems to me that it is unnecessary to decide these further issues.

Similarly, my conclusion as to the primary issue between the parties makes it unnecessary for me to consider Dr Battu’s further grounds of appeal based on the letter of 9 November 2000.

That leaves Spreadex’s cross-appeal for consideration.

Spreadex submit that Dr Battu’s trades were closed under an oral agreement made between him and Miss Mead earlier that afternoon that those trades would be closed unless he had paid £40,000 by 18.30: and that the significance of that agreement was emphasised by Miss Mead telling Dr Battu that at that hour there would be no one around to authorise anything different.

The judge did not deal with this submission, other than to record Dr Battu’s promise to pay the £40,000 by 18.30. No damages were claimed for breach of the agreement, which was also pleaded as amounting to an estoppel preventing Dr Battu from complaining of closure of his trades in the absence of the promised payment. The intended effect of the plea, therefore, whether seen as an agreement or an estoppel, is to render the demand of £40,000, and the closure of the trades in the absence of its payment, as in any event lawful.

In my judgment, however, the submission fails. Without detailed findings by the judge, we are left with the transcripts of the relevant telephone conversations. The premise of Miss Mead’s and Dr Battu’s conversation was that at the then current state of the market, Dr Battu had running losses of well over £80,000. By 18.30 the position might be even worse (“we need the forty to cover your position at the moment as well as this evening”). I observe that every 10 point rise made for a loss of £4,000. In their morning conversation, when the demand was for £50,000 (that was before the £29,000 was paid), and when the running losses were “about thirty-five”, Miss Mead acknowledged that the demand contemplated and anticipated a deteriorating situation but was subject to a change back in Dr Battu’s favour (“if the market moves one way we’re gonna be back on the phone to you so we need to have at least fifty on there which will prevent us having to call you every day…obviously if the market moves dramatically back in your favour and the, you know, trades go back into credit you can claim the money back at any time, that won’t be a problem”). There was no discussion between Miss Mead and Dr Battu about NTR on top of margin (a point made only at the last moment by Mr Pugh). The underlying assumption therefore was that the market would be worse not better at 18.30. In fact, the market was considerably better at 18.30, and Dr Battu claimed the benefit of that. He made it clear that he was willing to pay £7,000 which would more than meet his current open losses, and that he was objecting to his trades being closed out against him. In these circumstances there was no agreement, in any binding contractual sense, that required a payment of £40,000 in all circumstances, and any estoppel is undermined by a combination of the market change in Dr Battu’s favour plus his clear objection. Another change was that the £40,000 was now being supported by Mr Pugh by reference for the first time to the need for cover for the notional risk and the citing of that notional risk at a ratio of 160. Neither Spreadex’s pleading nor Mr Pelling’s submissions addressed the question of what prejudice made it inequitable for Dr Battu to substitute his offer of £7,000 for his previous promise to pay £40,000.

In the circumstances I am unable to find that Spreadex have made good a binding contract to pay £40,000 or an estoppel to the same effect.

It follows that the closure of Dr Battu’s trades that evening was uncontractual and unauthorised, and that Spreadex’s claim should fail. To that extent, in my judgment this appeal should succeed.

On the basis that Spreadex were in breach of contract in closing Dr Battu’s trades, he has a counterclaim, which the judge did not reach. The pleaded counterclaim states that Dr Battu would have kept his positions open until 4 October 2002, when the DJ index quoted by Spreadex was 7,470 – 7484, at which point he would have closed them.

There may be significant difficulties in that counterclaim. Thus the index closed on 26 September at 7952, opened on 27 September at 7943 and hit a high that day at 7960. On 1, 2 and 3 October it hit highs of over 7900 (7983, 7957 and 7903 respectively). Dr Battu may not have been able to sustain such losses. On the other hand, the index also plumbed lows as follows: on 27 September of 7656, on 30 September of 7429, on 1 October of 7530, and of 7710, 7667 and 7447 on the next three days. Such markets raise the question of whether Dr Battu could have mitigated Spreadex’s breach by restoring his positions at favourably high levels of the index so as to make his projected gains in any event, as the market periodically plunged. It is not clear to me whether his pleaded counterclaim is put forward on the basis that a repudiation by Spreadex had or had not been accepted, if such an analysis is even relevant. Nor is it clear what Spreadex might have known about his resources: see Chitty on Contracts, 29^th ed, 2004, Vol I, at para 26-096. The case of Ata v. American Express Bank Ltd (17 June 1998, CA, unreported: see McGregor on Damages, 17^th ed, 2003, at para 29-068) may be relevant.

In these circumstances, I do not think that the counterclaim can be dealt with other than by way of remission to the court of trial, although I would caution the parties over the prospect of incurring fresh expenditure of costs in arguing about what may well be speculative matters. I do not agree with Mr Pelling, however, that this court is in a position now summarily to say that Dr Battu’s counterclaim must fail.

In sum, if Spreadex had intended to require their customers to secure their trading by reference to a known NTR and in addition to provide further security for the whole of running losses, I do not think that they have achieved this on their current terms.

I can understand that in the fast moving world of spread betting on financial markets, the operator needs to retain powers to control and limit its customers’ risks, for the sake of both parties, and to stipulate for adequate security in its own interests. On the other hand, it is only fair, again in the interests of both parties and of the industry in general, that the terms of trade, and the security requirements, are properly set out and not left to contractual and conceptual black holes. It may be difficult to achieve the necessary combination of flexibility and clarity; but the effort should be made. We have been given to understand that the existence of TWCA in addition to the Rules reflects the requirements of the FSA, but it would, as it seems to me, be preferable if the applicable terms could be found in a single document, rather than, as here, spread between the two sets of provisions.

I agree.

I also agree.