Kennedy v Dresdner Kleinwort Wasserstein

The Claimant (Mr Kennedy) claims against the Defendants (“DKW”), his former employers, for sums to which he says he is entitled as bonuses for the years ended 31 December 1999 and 2000. In 1999 Mr Kennedy was entitled to a “formula” bonus based on results. In 2000 he was at most “entitled” to a discretionary bonus.

Mr Kennedy was first employed by DKW from 1 April 1997. In 1999 and 2000 he was a Senior Trader of the Global Foreign Exchange Options business (GFXO) of DKW. He worked in London in 1999. In early 2000 he was seconded to DKW’s Singapore office and worked there for the rest of the year. He returned to London in January 2001 and moved to a seat on the proprietary trading desk, Global Markets.

Mr Kennedy was employed pursuant to a written offer letter dated 17 March 1997 signed by him in acceptance on 24 March 1997. The letter provided that the terms set out in it were to be for an initial period until 31 December 1999 when the terms would be subject to renegotiation.

The terms set out in the letter included a provision that Mr Kennedy would receive an annual base salary and would be eligible to participate in a performance based bonus scheme. The performance based bonus was to be calculated by reference to the Net Operating Profit (NOP) of the GFXO and was to be based on 2% of NOP subject to a cap of US$ 575,000 total compensation.

For the year 1999, Mr Kennedy was paid his base salary and a bonus of £125,500. He claims that NOP was €15m (alternatively €12.7m) in 1999 and so his bonus at 2% should have been €300,000 or £183,390 (alternatively £155,270) at the exchange rate of £0.6113 to €1 at 27 February 2000 the date on which his bonus was advised to him. The shortfall alleged is £57,890 (or £29,770). The reason for the alternative figures is that a booking error discovered at the year-end meant revenues were €2.3m more than had been recorded. However the evidence is that the error was duly corrected in the figures on which DKW relies.

DKW claims that NOP for 1999 was €9,028,971 and that the 2% bonus should have been €180,579 or £110,388, less than the bonus in fact paid to Mr Kennedy.

In December 1999 Mr Kennedy had a conversation with Jorge Villon, then the Global Head of GFXO. Mr Villon told him that DKW would no longer grant formulaic bonuses and bonuses would be discretionary. Although a new written contract for 2000 was not entered into there is no dispute that Mr Kennedy was entitled to be considered for a discretionary bonus for the year 2000. It is also agreed, in accord with Clark v Nomura International plc [2000] 1RLR 766, that it was an implied term of Mr Kennedy’s contract of employment that DKW would not exercise that discretion “irrationally or perversely”. No new written employment contract was agreed until June 2001. That contract related to 2001 not 2000.

During March and April 2000 DKW was involved in merger discussions with Deutsche Bank. That caused uncertainty. In the year 2000 Mr Kennedy was paid a base salary of £100,000 and an expatriate package of £25,000. He was also awarded further payments of £62,750 (paid in July 2000) and £125,500 for 2000. The former was, it is agreed, a “retention” payment made to secure the loyalty of staff in the context of the merger discussions. There is an issue whether or not the £125,500 was also and only a retention payment, as Mr Kennedy contends, or a guaranteed minimum bonus payment as DKW contends. It was paid as to £115,500 on 27 February 2001 with the balance of £10,000 deferred.

Mr Kennedy’s case is that whether or not the £125,500 was a bonus, and even if DKW did (as it contends) consider payment of a further sum in excess of that amount, in the light of his performance in 2000 a decision not to pay him any further sum was indeed irrational and perverse.

Mr Kennedy contends that in 2000 he identified and implemented “a strategy” involving a series of USD/JPY barrier trades which generated a profit for GFXO in the region of US$ 17m. On that basis it is said that a bonus of about 15% (or $2.5m) should have been awarded to him. It was to the amount of a bonus on the assumption that profits were earned of that order that the expert evidence was directed. $2.5m at the exchange rate at 28 February 2001 would have amounted to about £1.73m.

It is DKW’s case that Mr Kennedy did not generate profits of $17m or anything like that amount. In particular it is said that it is not possible to calculate or attribute profit to particular trades for the very reason that they form part of a book of business which is risk managed as such by others than the trader who makes the initial trade. DKW’s case is that Mr Kennedy’s performance did not justify the payment of any further sum in excess of £125,500. It is not disputed that if he did indeed generate a profit of $17m or a sum of that order a decision to pay a bonus of only £125,500 (or nothing) would indeed have been perverse.

Mr Kennedy left DKW in May 2002. The present claim was issued and served on 12 November 2002.

The Global Head of Foreign Exchange (and Local Markets and Commodities) at DKW was Achilles Macris. Mr Villon was responsible for risk management of GFXO and proprietary activities from June to December 1999 when he became Global Head of GFXO. He also had other responsibilities. Mr Villon gave evidence at the request of DKW. In 1999 the Co-Heads of GFXO itself were Mr O’Keefe and Mr deCarlucci. From January 2000 they were replaced as Co-Heads by Tim Cartledge and David Cooney who had both been Senior Traders in GFXO (like Mr Kennedy) from April 1997 until that time. Mr Cartledge and Mr Cooney then became the immediate superiors of Mr Kennedy. Mr Cartledge gave evidence at the request of DKW. When he did so he was on “gardening leave” pending his departure from DKW to work for a competitor in the City. Stephane Coquillaud is and has been since February 2000 Head of FX Risk Management. He gave evidence at the request of DKW. Guillermo Giminez was a junior GFXO trader from December 1999 to January 2001 when he was appointed Head of Exotic Options Trading. He left DKW in November 2002 and moved to HSBC. At the time he gave evidence he was about to return to DKW as Head of the Options Desk. He also gave evidence at the request of DKW as did Jane Davidson an Associate in the Management Information and Systems Team within the Human Resources Department of DKW.

Both Mr Kennedy and DKW adduced expert evidence on the criteria for and level of discretionary bonus payments paid by investment banks in the City in 2000. Mr George Hacker, who retired in June 1996 as Senior Managing Director and Senior Executive Officer of Bear, Stearns International Limited London, gave expert evidence at the request of Mr Kennedy. Mr Craig Kersey, the founder and principal of Guildhall Limited, a business advisory firm, gave evidence on behalf of DKW. Both experts were well qualified to address the subjects of their evidence and in the event there was not much of relevance at issue between them. It was Mr Hacker’s evidence that the expected level of discretionary bonus to be paid to a trader who achieved revenue of $17m would be in the range £1.2m to £2m.

In 2000 the particular type of trade on which Mr Kennedy relies in support of his claim was an “exotic option” described as a “hedged barrier trade”. The parties helpfully provided descriptions of these trades which I have gratefully adopted and to a minor extent adapted for what follows.

Exotic Options are contrasted with more simple or “vanilla” options only by reason of a more complex structure. A vanilla option is a contract between a seller and buyer whereby the buyer, in consideration for the payment of a fixed sum (the “premium”) acquires the right but not the obligation for a period of time (until the “Option Expiry Date”) to enter into a second contract on predetermined terms at a future date. In FX Options Trading, the second contract would typically be to buy or sell a foreign currency at a future date in exchange for another foreign currency at a predetermined exchange rate. The commercial purpose (if there is one apart from the hope of gain on the transaction itself) is to limit the exposure to exchange rate movements over the time of the option.

Mr Kennedy relies upon 53 options the great majority of which (all but 3) were US$ put options. The buyer of the options (DKW) purchases for a premium an option to sell (put) predetermined (say 100m) amounts of US$ in exchange for Japanese Yen at a pre-set exchange rate (the “strike rate”) on a pre-set date in the future.

The “Barrier” is a particular refinement and may be one of three types: “Up and Out”, “Up and In” and “One Touch”. The options could only be exercised on a particular date in the future (“the Expiry Date”). The Barrier is a condition which determines whether or not the options can be exercised then. The condition is a USD/JPY exchange rate, also fixed at the time of the purchase of the option, which could be higher or lower than the strike rate. If the barrier is “up and out” the option can only be exercised if the exchange rate has not moved up to the barrier rate before the Expiry Date. If it does reach that rate the option is “knocked out” and will never be exercisable. 34 of the trades were of this type. If the barrier is “up and in” the option cannot be exercised at all unless and until the exchange rate goes up to the barrier rate. If it does go up to that rate it is “knocked in” and can be exercised on the Expiry Date. 16 of the 53 trades were of this type,

In either case the exercise of the options will of course depend on the spot exchange rate on the Expiry Date. If it is favourable, so that more Dollars would then have to be paid to buy the Yen than the strike price, the option will be “in the money” and exercised; if it is unfavourable so that the Yen can be bought for fewer dollars, the option will be “out of the money” and will be allowed to expire unexercised.

The other 3 of the 53 trades were “One Touch Options” which mandated a fixed cash payment at the Expiry Date if the chosen barrier was touched at any time during the life of the option. All 53 trades were interbank not customer trades.

A barrier trade will always include “a long spot position” in addition to the barrier option. The one provides a partial or initial hedge against the other. The object of the long spot position is to cover the purchase of dollars with Yen to provide for the cost of exercise of the option on the Expiry Date. If the option is exercised DKW will need to have the amount of dollars it has committed to sell on that date. If the option is knocked out or out of the money or never knocked in DKW “loses” the premium it has paid for the option. As the exchange rate moves DKW will need to hedge the currency risks of the cost of the premium and the risk of having too much of the delivery currency to meet the long spot position or because the option is knocked out or never knocked in. Thus both the premium and the long spot position create an exchange rate risk during the existence of the trade.

At the inception of the trade, in addition to the barrier option and the long spot position, a “delta exchange agreement” or “delta gap hedge” was also frequently entered into. The agreement would be to buy JPY and sell dollars in order to seek to close at the barrier rate the transaction that created the long spot position in the event that the option is knocked out. The risk of having to deliver the dollars will not then arise but DKW will have bought the dollars in order to hedge against that risk and will need to sell them.

The long spot position may also enable the accrual of interest (“carry”) at the difference between the (in fact) higher US$ interest rate and the lower JPY interest rate. The dollars bought earn interest at (say) 6% but no interest is paid on the Yen sold as the Yen interest rate is (say) zero.

In practice the long spot position is financed not by lending dollars and borrowing Yen at overnight interest rates every day but by daily financing trades known as “Tom Next Rolls”. These comprise a pair of spot trades for Yen and dollars permitting the daily accrual of the cash equivalent of the interest differential which would have been earned on the long spot position.

It is an important part of DKW’s case and not in dispute that the further hedging required throughout the life of the option was carried out on an aggregate or book basis and not in relation to specific trades. So, too, the Tom Next Rolls are carried out on the net position created by DKW’s overall requirements.

Murex is an options trading risk management and accounting system. It values positions on the basis of mark rate parameters entered into the database on a daily basis. When a barrier trade is entered, Murex shows the impact of the new position by way of the aggregate value of the option and long spot position. It also aggregates all other USD/JPY spot and option positions held by GFXO.

It was this aggregate currency position in the book which required to be hedged. Some trades might, in effect, hedge each other. The risk in the aggregate position would be hedged daily usually by spot trades. It was the task and skill of the risk managers (in 2000 Messrs Cooney and Cartledge) not Mr Kennedy to determine what daily hedging was required. On the evidence of Mr Cartledge, which I accept, risk management itself provides an opportunity for profit and a risk of loss. It was for Messrs Cooney and Cartledge to make the judgments as to the degree of risk to be run in the GFXO book and so to determine the position they wished to achieve and to manage the book accordingly. Even this explanation risks an over simplification of the matter. It is not right to look only at a pair of currencies in isolation in the context of hedging and the need to hedge. Positions held in USD/Euro and JPY/Euro, for example, may impact on USD/JPY positions. To use the expression used in evidence there may be “leakage across currency pairs”.

It is perhaps no wonder that those who trade in exotic options and risk manage FX positions are by training and qualification skilled as much in computer sciences, statistics and the like as other disciplines. The opportunities for generating profits and losses from such trades are analytically four-fold. First, the amount of the premium required by counter-parties may undervalue the Option. The evidence is that such was indeed the case with USD/JPY barrier options in at least the first half of the year 2000. The “undervalue”, on the evidence, can be deduced from a more accurate assessment and calculation of the risks and rewards made by DKW than others in the market. After a time the market will catch on but in the meantime there is an “edge” to be earned on such transactions.

The second potential source of profit or loss is the “carry” arising from the interest rate differential on dollars and yen. The extent to which it is earned will depend on the actual and aggregate extent to which dollars are held by DKW. The third source is the profit or loss on the long spot position after exercise of the Option. The fourth source is the cost of hedging of the aggregate positions of the book. It is, however, DKW’s case that the last three sources are not capable of analysis, and were not in fact analysed, by transaction nor attributable to specific transactions because they were managed as part of the book of GFXO as a whole. It is also DKW’s case that the task of a trader, such as Mr Kennedy, was to find transactions with “edge” but that it was not part of his responsibility or task to address the other potential sources of profit and loss. The barrier trades undoubtedly had edge; indeed it was readily accepted by Mr Cartledge that the level of edge was exceptionally high. But, as Mr Cartledge put it “the role of our traders is to capture edge and not to reason about what is going to happen in the world”.

Mr Giminez said that even if (contrary to the fact) the trades were viewed in isolation they would make money from these other sources only if the trades remained alive (were not knocked out) for a substantial time, there was low volatility and a higher spot rate at the Expiry Date. These are matters for subjective judgment and cannot be objectively assessed at the time of the trade as is the case with edge. The ability to earn carry comes at the price of the cost of the premium and a foreign exchange risk.

Mr Kennedy was entitled under his contract of employment to a bonus of 2% of NOP of the GFXO business. The offer letter dated 17 March 1997 also provided a mechanism for the determination of disputes in relation to the calculation of bonus. They were to be determined by “the appointed Financial Controller [of DKW] in accordance with its prevailing accounting principles and practices and the decision will be final”.

Granted that Mr Kennedy’s contract expressly referred to NOP it could have been expected that DKW would be able to establish the relevant figure with some certainty. In fact the evidence is that this was far from being the case. Indeed the bonus in fact paid was not calculated by reference to NOP at all but was awarded on a discretionary basis. That was explained by Mr Villon to Mr Kennedy at a meeting on 1 February 2000 which Mr Kennedy surreptitiously recorded.

The amount of the bonus was subsequently approved in the sum of £125,500 of which £115,500 was paid on 30 March 2000 with the balance deferred in accordance with DKW’s policy also stated in the offer letter. Mr Kennedy made no complaint about the amount of the bonus until his solicitors raised the matter in August 2002.

It is Mr Villon’s evidence that the NOP of the GFXO department for 1999 was €9,028,971.02. That figure is taken from a document dated 28 April 2000 entitled “Profit and loss account as of December 31, 1999.” The document itself is not very informative in the sense of providing any detailed breakdown of the sums shown but the net figure is calculated after “administrative expenses location” of some €11.755m and “administrative expenses DB Frankfurt” of some €3.289m. Those expenses are made up wholly of “indirect costs and mark up” in the case of Frankfurt and include such costs and mark up in the case of “location” of some €4m. Mr Villon said the document had been “signed off by central management in Germany” and was “the official Profit and Loss account for the Global Book of the GFXO department” for the year 1999. One of the signatories held the position of Financial Controller referred to in Mr Kennedy’s contract. Whilst DKW’s accounts for 1999 were audited by Price Waterhouse Coopers, Mr Villon was unable to say what, if any, part the document or the information in it played in the audit nor was he involved personally in its preparation.

Mr Kennedy’s “best estimate” of NOP for 1999 was derived from an estimate of trading revenue of €23m reduced by €8m of “allowable direct costs”. These figures were said to have been derived from observation of Murex figures at the end of 1999 and conversations with Mr O’Keefe and Mr deCarlucci. Mr Kennedy also said that Mr Cooney told him in December 1999 that they were to be paid on the basis of a NOP of US$10m which would produce a bonus of about £125,500.

Whilst Mr Villon’s evidence about NOP is of a general nature I see no reason to doubt that it is accurate. Nor do I think in principle, let alone in a competitive market, it would have been in DKW’s interests to underpay bonuses. The matters on which Mr Kennedy relies for his figures were known to him by or about the 1999 year end and I can see no good reason why he did not raise them at the time he was informed of his bonus or invoke the provision for reference to the Financial Controller but only made a complaint over two years later.

There was some debate about the inclusion of indirect expenses and mark up but in principle I would expect NOP to provide for such expenses and that DKW’s accounting principles and practices would reflect some allowance for such matters. Just what might be included and at what level could indeed be a matter of considerable debate and the experts disagreed upon it but the actual figures put forward by Mr Villon do not strike me as in any way surprising or exorbitant. Nor has Mr Kennedy sought to make a case to the contrary.

Mr Kennedy did say that Mr O’Keefe had told him that only direct costs were to be deducted. In fact the contracts of Mr O’Keefe and Mr deCarlucci, whilst also providing for performance based bonuses calculated by reference to NOP, contained express provisions as to how NOP was to be calculated (“gross trading revenues less operating costs”) and provided that:

“Operating costs will include those outlined in Schedule A attached and any other operating costs and provisions reasonably allocated to the [GFXO] business as agreed between senior [DKW] management and yourselves. In year one, operating costs will be capped at US$8 million. If NOP is negative in any year, it will be carried forward to the following year.”

Schedule A, albeit using the language of “direct cost” included some matters (such as “Treasury” and “IT Support” and “Accommodation”) which could, I think, equally readily have been described as “indirect” cost. In fact had the “carry forward” provision been applied there would have been a nil bonus in 1999 under these contracts because of the losses GFXO made in 1998. The contracts also expressly acknowledge the allocation of other reasonable operating costs and provisions of which again (as I construe the wording) the ultimate arbiter was to be the Financial Controller in accordance with DKW’s prevailing accounting principles and practices.

Mr Villon was also, and understandably, criticised for apparent inconsistencies in DKW’s response to Mr Kennedy’s claim. It was stated on instructions in September 2002 by DKW’s solicitors that GFXO had made a negative NOP in 1999. Mr Villon said that claim had been based on a document which purported to set out the bonus calculation at October 1999 and a projection for the year which did indeed show a cumulative negative figure. Mr Villon said the estimates for costs were only assumptions and he knew they were unreliable. When the statement that NOP was negative was corrected by DKW’s solicitors in a letter dated 17 January 2003 it was also asserted that the figure of €9,028,974 was “audited by KPMG”. In fact it was neither audited nor were the auditors KPMG. But Mr Villon was entirely straightforward in his evidence about the figure and, as stated, I see no reason to doubt what he said about it.

In my judgment Mr Kennedy has failed to make out a case that the application of 2% of NOP would have resulted in a bonus payment any larger than that he in fact received. Indeed I am satisfied as a matter of probability that it would only have resulted in a payment of the order of £110,000. The claim for 1999 therefore fails.

There is no doubt that Mr Kennedy was disappointed when he was not promoted and Mr Cartledge and Mr Cooney were. He also seems to have suspected that he was at risk of being treated unfairly at least in the sense that they might take the credit for any good ideas he might have. Mr Villon’s reasons, which were not challenged, for promoting Mr Cartledge and Mr Cooney were that they had risk management experience which Mr Kennedy did not. Mr Villon said, and I accept, that he had not considered Mr Kennedy for the post nor told Mr Kennedy that he would. Mr Kennedy nonetheless felt he had been told that he would be informed when the process of looking for a new Head of Desk began and felt let down when the appointments were announced without warning.

Mr Kennedy was asked to go to Singapore, on a substantial salary, principally to support the sales team in that region to develop customer business. DKW’s declared trading policy was to concentrate more on customer trading than interbank trading. There had been concerns about Mr Kennedy’s relationship with the sales teams in Frankfurt and London but not such as to merit any formal meeting or written record beyond some general comments on his 1998 Appraisal. As Mr Villon put it, and I accept, the secondment to Singapore was seen as giving Mr Kennedy an opportunity for a new start. Mr Kennedy’s role was discussed with Mr Villon on 18 January 2000 and set out in an e-mail copied to him and sent to Ken Lee the Head of Sales in Tokyo on the same day.

Before he went to Singapore, Mr Kennedy asked Mr Villon to give him some independent trading limits to take advantage of opportunities he saw. Mr Villon asked him to set out his ideas.

There is an important distinction between a trader who operates as part of “the desk” and one who has his own book. At DKW the latter was known as a “proprietary trader”. The key features of a proprietary trader at DKW are that the trader is given a limit on the extent to which he can risk the bank’s capital and is personally responsible for the risk management of the trading undertaken within that limit. In terms of performance the proprietary trader would have records of the results achieved. It is an important and undisputed feature of this case that despite his request Mr Kennedy was at no time in 2000 granted the status of a proprietary trader with a trading limit and indeed that he was expressly refused such a role.

Mr Kennedy responded to Mr Villon’s request to set out his ideas in an e-mail sent on 19 January 2000. Mr Kennedy set out an example of a barrier option trade and suggested Mr Villon consult Mr Giminez if he wanted another opinion on the trades. It is Mr Kennedy’s own case however that whilst the “concept” of his “strategy” did not differ in substance from what was stated in this e-mail “the strategy could be and was implemented by means other than that shown in the example in the e-mail”. Mr Villon responded by e-mail about an hour later. He said:

“This sounds like a trade which obviously falls outside the market-making but for which we should have risk appetite. I certainly do not have a problem with you running positions like this given that I understand the analysis and risk/reward. Have you presented these ideas to Dave/Tim? I would be very happy to support setting up a “back-book” for lack of a better word for smart ideas. Can we discuss with Tim/Dave? ”

Mr Kennedy responded “sure”. “Tim/Dave” were Mr Cartledge and Mr Cooney. Mr Villon said, and I unhesitatingly accept, that he meant by the word “given” “provided that” and not that he had understood the analysis and risk/reward. He said (rightly) that Mr Kennedy had given no figures in his e-mail and risk was about figures. He also said, and I also accept, that in no sense was he giving approval to Mr Kennedy: hence the reference to discussing it with Messrs Cartledge and Cooney. Mr Kennedy himself, and despite the contrary inference in his witness statement, acknowledged that Mr Villon’s e-mail was not “a definitive approval”. He never did discuss his ideas with Messrs Cartledge and Cooney. Indeed, as is apparent from the exchanges in August 2000, he did not do so deliberately for fear they would take the credit for them. To quote his witness statement (paragraph 8.3) Mr Kennedy said “having got Mr Villon’s approval, I decided to work the strategy as part of my normal trading in the main book until such time as the Heads of Desk chose to designate a separate book for trades in the strategy”. But, I am satisfied, he had no “approval” and the Heads of Desk knew nothing of any “strategy”.

Before he left for Singapore, on 1 February, Mr Kennedy had the informal meeting with Mr Villon which he tape recorded without informing Mr Villon that he was doing so. That is some measure of the concerns and fears which Mr Kennedy was harbouring. The recording is not very good to judge by the transcript of it prepared by Mr Kennedy. But significantly it does contain a request by Mr Kennedy to have his own “book P/L broken out then I got something I can be judged on” and Mr Villon’s response that it would be “a different story” if he wanted to be a proprietary trader. Mr Kennedy was and was to remain a member of GFXO when working in Singapore.

In early March 2000 the prospect of a merger of DKW with Deutsche Bank became public. In early April it was announced that merger negotiations had failed. In the interim Messrs Cartledge and Cooney managed the GFXO book so as to reduce the risk levels to a minimum and payments were made to staff in the hope of retaining their loyalty. The terms of the letter to Mr Kennedy informing him of this payment are material to the question whether or not the subsequent award of £125,500 was a further “retention payment” or a bonus payment. So far as material the letter, dated 17 March 2000, stated:

“I am pleased to confirm that, as part of the Company’s commitment to staff during this process of merger, you will receive a retention payment of £62,750. This will be paid in full … on Friday 28 July 2000. This payment is distinct from and in addition to the Company’s normal discretionary performance based remuneration arrangements.”

The letter set out certain conditions for the payment the major one being that Mr Kennedy had not given notice to resign before 28 July.

Mr Kennedy made one of the 53 barrier trades in April 2000 which came to Mr Cooney’s attention because it showed a sizeable loss. Mr Kennedy sought to explain the trade in an e-mail to Mr Cooney sent on 12 April saying he thought it was “trading as usual again after the merger was off”. Although Mr Knowles sought to rely on this e-mail as some evidence of Mr Cooney’s knowledge of Mr Kennedy’s so called “strategy” I do not think the reliance is justified. Mr Kennedy said this was the first occasion he had explained “the mechanics” of the trade but acknowledged that neither Mr Cooney nor Mr Cartledge knew why he was doing the trades and said with reference to this e-mail (wrongly referred to as dated 14 April) that he did not think Mr Cooney “took it all in”. The true picture also emerges clearly from the exchanges later in the year in August.

The material documentation relating to the award of £125,500 is as follows.

An internal form completed on 6 June 2000 in the Human Resources Department records for Mr Kennedy an “Amended Stage II retention bonus – Feb 2001. £125,500”. On the same day a letter was sent to Mr Kennedy by the Human Resources Department. So far as material it stated:

“I am pleased to confirm that, in respect of the year 2000, you will receive an award valued at no less that £125,500 as follows:

a)

Property with a value of no less than £115,500 to which you shall become entitled in February 2001.

b)

A deferred award of property with a value of no less than £10,000 which will vest in three instalments in March 2002, 2003 and 2004 ….

The above award is in addition to the retention payment advised in our letter of 17 March 2000.”

The letter also set out conditions for the payment of “the award” the major one being that Mr Kennedy had not given notice to resign “before the due date”.

Mr Kennedy did not see the internal form. The form refers to a “retention bonus”. So also did an e-mail sent by Human Resources to Mr Kennedy to inform him that “a stage two retention bonus letter” was on its way to him. The letter itself is in my judgment sufficiently clear that the payment to which it refers is indeed a “bonus” albeit one which is guaranteed. The contrast between the terms of the 6 June letter and the “retention payment” letter dated 17 March is apparent. The former refers to an “award” “in respect of the year 2000” payable substantially in February 2001 (the usual bonus payment date) and in a split form (also usual for bonuses). The latter refers to a “retention payment” payable during the year itself and expressly contrasts the payment with the normal bonus.

£115,500 was duly paid to Mr Kennedy in February 2001. The letter informing him of the payment referred in terms to “your contractual bonus”. He complained about “my bonus this year” to Mr Macris in an e-mail sent on 21 May. His complaint was that he had only received “the minimum guaranteed loyalty bonus”.

I have no doubt that DKW considered the £125,500 to be a bonus and that Mr Kennedy knew and accepted as much. The fact, which is also clear, that the bonus was notified in June and guaranteed in order further to encourage staff to remain with DKW does not, in my judgment, affect the status of the payment as a bonus.

On 9 August Mr Kennedy sent an e-mail to Mr Villon enquiring what Mr Villon’s thoughts were on “his future role in FX options”. He added “surely you are aware that the desk has been making some money in the jpy books. You may not know that a lot of it is due to a big barrier position that I (with Guillermo’s help) have built since February. If you recall, I sent you a mail in January telling you what type of trade I wanted to put on this year, and that’s exactly what Guillermo and I have done”.

Mr Villon’s response on 11 August was:

“I am aware of the $/jpy accrual positions in the book as Dave spoke to me about them at the time they were initiated and I do recall this was a suggestion you had in January. I am delighted it has worked out for the book.”

Also on 9 August Mr Kennedy had sent an e-mail to “Global FX Options Trading” asking if “the barriers” could be moved to “the global hedge book”. The next day he sent a further e-mail stating:

“this barrier position has a chunky fwd risk. If BOJ raises rates we lose on funds. We need to fwd hedge this stuff …. Does anyone object to putting the trades in the GH book?”

This e-mail plainly came to the attention of Mr Cooney who replied to Mr Kennedy on 10 August:

“I don’t want to move these trades to global hedge. We are not actively trading that book. I would like to leave them as they are to make it easier to view the consolidated risk….”

In effect, Mr Kennedy was seeking to create a separate “book” for the barrier trades whilst recognising the need to hedge them, but Mr Cooney was not prepared to do so for the very reason that the trades were better managed as part of the overall GFXO book.

Mr Kennedy’s unhappiness and resentment was manifested in a long e-mail he sent to Mr Macris on 21 August. It is apparent from the e-mail that Mr Kennedy thought Mr Macris did not know much about him and so he introduced himself at some length including the statement that he was “the best trader you have on the desk” but was also “slowly being forced to leave the bank”. Mr Kennedy said that Messrs Villon, Cartledge and Cooney had been trying to squeeze him out since the beginning of the year and “I believe they don’t want senior guys because they don’t want anyone who may show them up and they don’t want to share any significant bonus”. Mr Kennedy continued:

“You may have noticed that the USDJPY book is making a lot of very consistent money. You may also have noticed that the risk … is low. This money is coming from a volatility arbitrage position. The arbitrage is only possible to achieve by using barrier options … because less experienced traders are mispricing it ….

At present the desk is up approx 6 million euros directly attributable to the barrier position. At present we are making about 160k euros a day ….

The only one who understood this position until recently was Guillermo …. I told Jorge several times at the end of 1999 that I would like to move to the prop. desk. In the beginning of January of this year I sent Jorge an e-mail describing this barrier trade and why it was perfect for the USDJPY market conditions. He told me to go to Dave. After the way I felt at the end of the year, I had no interest in giving Dave my good ideas so he could put on the trade. I didn’t explain it to anyone else except Guillermo because I knew it was a big winner and I needed to build up my own credibility….

The reason I am stressing the point that neither Jorge nor Dave nor Tim knew what Guillermo and I were doing is because I want my contribution to be properly acknowledged. Last week Jorge stated in an e-mail to me that Dave told him about this position when it was initiated. That’s a blatant attempt to claim that they were somehow involved in putting this position on….

…. I certainly don’t want to hang around until March if we can’t agree numerically on how much I have already made for the desk and what bonus I can expect ….

I have been here as long as both Dave and Tim, and I feel I have contributed more. I don’t see a position for me working under Dave and Tim. I am too senior and too experienced to be happy doing that. I want my own P/L ….”

Mr Macris responded saying it was “heavy stuff” and asking for “a private call for me to get more context”.

Mr Kennedy’s evidence was that the private call took place that evening London time. Mr Macris said he would get involved in determining Mr Kennedy’s bonus and make sure Mr Kennedy was treated fairly and that “the hundred thousands number was right”. Mr Kennedy also said Mr Macris told him to “keep trading and try and make as much money as you can”.

Mr Knowles, QC for Mr Kennedy submitted that it was significant that Mr Macris had not said anything to Mr Kennedy to the effect that he should not be doing trades of this nature or concerning himself with anything other than the initial edge on the trades. But that is to ignore the fact which Mr Kennedy’s e-mail demonstrates and is unsurprising that Mr Macris, at the very senior level in DKW at which he operated, had no prior knowledge of the issues. It is also to ignore what happened a few days later.

As Mr Kennedy put it “right after the phone call with Macris” he did two up and out barrier trades in a total of US$ 130m. Mr Cartledge spoke to him and said he should not do any more of the trades because they no longer had edge. That led Mr Kennedy to e-mail Mr Villon on 24 August. He said he had had a discussion with Mr Macris “about my future at the Bank” and that Mr Macris had offered him the opportunity to move to a proprietary trading position. He also said “on Tuesday I did a 100 USD trade that Dave and Tim felt was marginal. They do not want me to build the position any more unless I can get the trades on at an impossible spread.” The e-mail contained none of the figures, supposedly attributable to the trades, which Mr Kennedy had provided to Mr Macris, let alone the request to agree what they had made for the desk.

Mr Villon replied the same day. He wrote:

“I spoke with Tim and Dave about the US$ 100m trade. They explained to me that the trade has far less edge at inception than the previous trades you had executed to build the position and that they did not feel, at these levels, it had a great positive expectancy. The position has been built to US$750m and they, as risk managers, felt this was sufficient given current levels. This conversation came about because I asked them for an explanation of the mark-to-market loss sustained to put the trade on. I don’t think this is unreasonable. You have done a good job at building a position with positive edge but it seems to me like the market liquidity has now been absorbed. I don’t know why you would be depressed after having built a good position and realising that maybe you shouldn’t go back to the well one last time.”

Mr Villon also expressed courteous disappointment that Mr Kennedy had spoken to Mr Macris regarding a role in proprietary trading rather than himself but said he would talk to Mr Macris about it.

Mr Kennedy’s response about the $100m trade was that he agreed “liquidity is pretty much gone” and that “our position is large enough now that further building it at fair-ish value is not wise”.

These exchanges are, I think, of some significance. The “position” for which Mr Kennedy was credited for a “good job” was “a position with positive edge”. When the edge went he was, in effect, told to stop doing the barrier trades and did so. That is entirely consistent with Mr Cartledge’s evidence as to the role of a trader, to capture edge. It is not, however, consistent with Mr Kennedy’s case that edge was only one of the profit constituents of his “strategy”. Mr Cartledge said, and I accept, that it was only in August that he had “even the remotest suspicion that Mr Kennedy was more concerned about the result of the trade than the edge” and he made it clear to Mr Kennedy that “was not how we were to conduct our business”.

When the bonus (as I find it was) was paid to Mr Kennedy and after he complained about it to Mr Macris (paragraph 56) a meeting was held on 19 June between Messrs Macris, Villon, and Pullman (of Human Resources) and Mr Kennedy. Mr Kennedy prepared a note of his recollection of what was said at this meeting. Essentially there were two subjects, the bonus and the fact that Mr Kennedy did not have a current legal contract. Mr Macris wanted to deal only with the contract first and left the meeting telling the others to sort it out. They did and a contract was produced shortly thereafter. The bonus issue was left over. The new contract was signed by Mr Kennedy on 21 June and related to his employment thereafter as a Senior Proprietary Trader.

On 19 September Mr Kennedy reminded Mr Pullman that the issue of his bonus for the year 2000 had not been resolved. He was told in response that “after discussions with David Cooney and Jorge Villon we believe the bonus payment for 2000 was appropriate for the level of contribution to the business”.

It was Mr Villon’s evidence, and the evidence of Mr Cartledge, that in late 2000 and early 2001 they and Mr Cooney had discussed discretionary bonus levels for members of the GFXO desk. Market value ranges had been discussed and individual performance. They had agreed that £125,500 was a fair bonus for Mr Kennedy. It was Mr Cartledge’s evidence that he and Mr Cooney took account in particular of the good performance of Mr Kennedy in obtaining edge on the barrier trades when considering the level of bonus to be awarded to him. Mr Cartledge’s “rough sense” was that the edge was worth up to US$1.5 to 1.75m on 1.5 to 1.75 billion dollars traded. Mr Villon and Mr Cartledge were disappointed at Mr Kennedy’s work (or lack of it) with the sales force in Singapore and Japan to build a customer business in exotic options but it is not suggested that any disappointment could begin to justify a failure to pay a much larger bonus if Mr Kennedy had truly earned the sort of substantial sums for the desk for which he contends. It was also Mr Villon’s evidence that following the meeting on 19 June 2001 he had discussed the level of Mr Kennedy’s bonus with Mr Macris, Human Resources, and Mr Cooney, but it was again concluded that the payment made was appropriate for Mr Kennedy’s contribution to the business. The e-mail dated 25 September had then been sent.

Although there was some challenge in cross-examination to this evidence about the way in which the bonus was assessed and reviewed, and in particular the involvement of Mr Macris, I am quite satisfied it was truthful and accurate. Much of the process is not documented but I accept that was usual because the issues were both delicate and confidential.

Mr Kennedy’s claim that the best estimate of the profit on the trades within his strategy was US$17m was substantially founded on a document known as a “Risk Report” dated 22 November 2000 for the USD/JPY currency “family” which showed a “P&L since 30.12.99” of US$17,208,175, the then equivalent of about €20.3m.

The origin of these Risk Reports is a print out from Murex. Mr Kennedy produced print outs for dates between August and November 2000 and further ones at the end of January 2001. The copies were faxed to Singapore from the New York office of DKW. Mr Kennedy frankly said he kept them because he thought there would be “problems in being paid”. It is some indication of the significance of these documents that they were not retained and are not recreatable by DKW itself nor were they or the information in them used for any accounting purpose. They represent realised results on closed positions and mark to market figures on outstanding positions. They cover all USD/JPY transactions. There are no equivalent figures available at the year-end.

It is Mr Kennedy’s case that the figures are net of all hedging costs of the book to the date shown. In principle I think that should be right, as they show a “P&L” figure, albeit Mr Kennedy’s own evidence cast some doubt upon it. The figures include “sales credits” (edge and add-on income) from customer sales and edge from all the interbank trades made by other traders in GFXO. Sufficient for the claim in these proceedings, the evidence suggests that figures of €4m for customer trades and €2m for edge on interbank trades are appropriate, of which €1.5m to €1.75m was attributable to the 53 trades. Again for practical purpose the € and $ can be taken at year end at parity. Thus some 4 to 4.5m of the 17.208m (treating that as a full year figure) was not, even on Mr Kennedy’s case, attributable to him. But it is his case that the balance (say 13m) was earned from ‘his’ 53 trades.

The basis for that case is the assertion that it was only the 53 trades which can be regarded as making money and every other trade must be seen (in effect) as hedging those trades. That assumption is said by DKW’s witnesses not only to be unjustified of itself but also to ignore the fact that the degree of risk and so opportunity for profit in the USD/JPY book was the responsibility of Messrs Cartledge and Cooney not Mr Kennedy. As Mr Cartledge put it, any profit derived otherwise than from edge is made from their positioning of the USD/JPY book: “it is my evidence that the positions of our books were our desired positions and that there were no changes in our positions because of the addition of [Mr Kennedy’s] trades. We would have exactly the same position during the year whether those trades were in or out of the book”.

DKW produced a spreadsheet showing the count of trades done by currency pair and option type in the year 2000. The total number of trades in the USD/JPY book was 6888. It was the second highest number of all currency pairs (EUR/USD was the highest). Of these 243 were barrier trades (including the 53 trades on which Mr Kennedy relies); 3765 were spot trades probably representing daily hedging, and 2788 were vanilla trades. Mr Cartledge said, and I think these figures bear him out in principle, that there were a number of types of trade, particularly vanilla trades, which were available to the risk managers to achieve the overall position of the book on which they had determined. As he put it, if the 53 trades had not been done, others could have been to achieve that position.

There is a further spreadsheet produced by DKW which sets out (in Euro) the profit and loss of GFXO for the year 2000. Total operating income is shown as €36.262m. After direct costs and allocations, profit before tax is shown as €16.419 or £9.85m. The Risk Report, produced by Mr Kennedy, for the composite portfolio at 22 November 2000 showed a P&L figure for the year to date of €42.855m. It is Mr Kennedy’s case that the difference between €42m or €36m and $17m is the cost of hedging.

The DKW spreadsheet showed “total operating costs” of €19.843m (including allocations of €9.616m). Mr Knowles questioned the figure of €8.252m included in the total for personnel costs on the basis that the record of salaries and bonuses paid to GFXO staff in the year amounted to some £2m less than this sum. Whilst it is right to record that when this apparent discrepancy was raised at the beginning of the trial, Mr Underhill QC for DKW said he hoped to be able to explain it, he did not adduce any evidence to do so save that Mr Villon said (no doubt rightly) that some small part of it might be due to the cost of pension and National Insurance contributions and other benefits. Nonetheless, I see no reason to doubt the figure shown. There are a number of possible explanations. For example, no part of the salary or bonus paid to Mr Villon himself is included in the record of payment to staff of GFXO.

There is a further DKW spreadsheet which records that the total (in all currencies) of “sales credits” earned in the year was €25.672m. Taking the operating income figure of €36.262m on this (admittedly simplistic) view the balance leaves no room for $17m to have been made by Mr Kennedy.

It was Mr Villon’s evidence that the combined notional value of the 53 trades, (that is assuming they were viewed as a separate entity) was US$ 1.75billion. Mr Villon said, and I unhesitatingly accept, that if Mr Kennedy “had been acting as a proprietary trader in 2000 he would never have been permitted to place trades worth even a fraction of that amount”. Mr Coquillaud’s evidence was to the same effect and equally compelling. Indeed he said that if the 53 trades had existed as a stand alone portfolio they would, without hedging, have breached the limits for the entire desk several hundreds of times.

There is attached to this judgment a schedule of payments of salary (“base”) and bonus in fact made to staff of GFXO for the year 2000. It will be seen that the Risk Managers were awarded bonuses 4 or 5 times the amount awarded to any other staff member. Those awards were not, on the evidence, in any way affected by what Mr Kennedy was paid. Mr Kennedy was the trader with both the highest salary and the largest bonus award. If he had been paid or was entitled to a bonus of between £1.2m and £2m, as Mr Hacker thought appropriate for a trader who achieved revenue of $17m, it can be seen that he would have received between 60 and 100% of the total of bonuses in fact paid. It would also represent between 12 and 20% of the profit before tax shown in the spreadsheet referred to in paragraph 81.

It is accepted that Mr Giminez in fact initiated a proportion of the 53 trades. Mr Kennedy thought perhaps half of them, Mr Giminez rather less. The two were at issue as to the extent to which Mr Kennedy could justifiably take credit for the trades done by Mr Giminez. Mr Kennedy was the senior man (but not the ‘boss’ of Mr Giminez) and claimed to have devised “the strategy”. Mr Giminez said, and Mr Kennedy agreed, that he did his own trades on his own assessment of edge. The only relevance of the issue is, I think, as a further illustration of the uncertainties created by seeking to apportion credit to individuals for the overall performance of a book of business to which there are many contributors. Whilst the court does not know which specific figure applies to him, Mr Giminez, as a trader, was paid substantially less than Mr Kennedy for 2000 as is apparent from the attachment to this judgment.

I am quite satisfied that it would be wrong to attribute any sum beyond the edge on the 53 transactions to Mr Kennedy. It is notable that in August when the edge ceased to exist, the trades were stopped. Once the trades were booked they ceased to be Mr Kennedy’s responsibility. Whatever his “strategy” not only were Messrs Cartledge and Cooney unaware of it, but it was no part of Mr Kennedy’s role to have a strategy as regards interest rate or exchange rate risks. That was the responsibility of Messrs Cartledge and Cooney. It is artificial to seek to isolate the 53 trades when they were addressed as part of the GFXO book and managed as such. The artificiality is, I think, well illustrated by the concomitant assumption (paragraph 79) that no other trades in the USD/JPY book made any profit at all. It is also illustrated by the gross breaches of risk limits (paragraph 84) which Mr Kennedy’s figures require. On the evidence, it is not correct (save in the case of edge) to attribute profit to particular trades. In this case edge would be a figure of only about $1.75m even without crediting some part of it to Mr Giminez. Mr Kennedy has in effect sought to present himself as a proprietary trader, and a proprietary trader without risk limits, when he was expressly refused such a role and would never have been allowed it on such terms. Mr Knowles submitted that Mr Kennedy was in effect subject to a risk limit: the limit imposed by Messrs Cartledge and Cooney from time to time as they managed the book. But that is, I think, a false analogy. The judgment as to the size and achievement of that limit was entirely the judgment of the risk managers and made on a continuing basis. A proprietary trader has a limit imposed which is to be managed by the trader. No doubt the limit could be altered in the course of a year if circumstances justified it but because the trader manages it the risk of failing to meet the limit is the risk of the trader just as any success engendered from operating within the limit can be separately assessed and redounds to the credit of the proprietary trader.

“Credit” for the gain arising from the risk management of the book, which, together with leakage across currency pairs, Mr Cartledge said substantially made up the balance of the $17m figure, belongs with the risk managers. The unreality of the claim made by Mr Kennedy is also demonstrated by the size of the bonus he claims relative both to others and the profit before tax made by GFXO (paragraph 85).

It is in any event for Mr Kennedy to establish that in 2000 DKW acted irrationally or perversely in setting his bonus at the level of £125,500. Even if it were properly arguable that the 53 trades created the opportunity for profit of the order of $13m the question would remain whether a decision to award a bonus by reference to the edge on the trades and no more could be said to be irrational or perverse. I do not think it could. DKW was in my judgment not acting in any way irrationally or even unreasonably in crediting the results of risk management to the risk managers and not to Mr Kennedy. Whatever his wishes Mr Kennedy remained a trader as part of GFXO and not a proprietary trader.

Mr Kennedy’s claim for payment of a further bonus for the year 2000 fails. In no relevant sense did the 53 trades “earn” $17m or any sum approaching that figure. At most he was entitled to credit for earning about $1.75m of edge. There was nothing irrational or perverse in the decision to award him a bonus no greater than the guaranteed figure of £125,500.

Mr Kennedy’s claims for both 1999 and 2000 fail and must be dismissed. I will hear the parties on any matters consequential upon this judgment when it is formally handed down.