By Wednesday 23 January the unwind was complete and the total loss – €4.9bn ($7.1bn) – gave Société Générale’s Jérôme Kerviel the dubious honour of precipitating the largest ever single trading loss, eclipsing even the $6.5bn collapse of Amaranth Partners, the hedge fund, in 2006. SocGen told the market on Thursday 24.
A post mortem, using information from the bank’s own reports into the incident, and those with first hand experience of previous rogue trading scandals, show that the Kerviel saga shares remarkable similarities in their circumstances and in their execution, and that some fundamental lessons remain to be learned by the industry. Furthermore, issues that in this incident, and in the infamous cases of Allfirst and Barings, caused massive financial losses and damage to reputation, are reportedly endemic to the industry. Given recent volatility across a number of markets, it was only a matter of time before it happened to someone.
Philippe Carrel, EVP of Reuters Trade and Risk Management, is among those of the belief that had it not been SocGen, it could have been one of its rivals. “I’m not surprised at all,” he says. “What we’re looking at here is basically an operational failure. From what I understand, a person from the front office has been able to square positions against fake counterparties.” While he claims no specific knowledge of SocGen’s systems, he says that there is a general lack of readiness across the industry when it comes to limit systems for futures trading.
“From my experience, when people were setting up the limit systems in the first place, it was always with credit risk in mind. With futures, they were thinking that there is no credit risk as your counterparty is a clearing house.” Typically, limit systems in futures trading monitor P&L, with daily checks on measures of exposure that are based on a trader’s net position. “Provided you square your position before end of day, then it will not show that you are exposed,” says Mr Carrel.
“Am I surprised? No. Because of the way we look at futures. Was it bound to happen? Yes, because in times of volatility these things appear – you can’t hide them.”
Unusually volatile market conditions tend to shake out risky trading strategies, Mr Carrel says, pointing to Nick Leeson, whose 1995 fraud brought down Barings and was revealed in the aftermath of the Kobe earthquake; and Long Term Capital Management, which collapsed in stormy markets, precipitated by the 1998 Russian debt crisis. Whenever there are dramatic price movements, rogue traders, bad traders and errors tend to be exposed.
“What this commands is that for very fast moving markets, like futures and equities – especially now when it tends to be automated and given to algorithmic trading – then you need to have an internal real time view of the overall exposure, not end of day,” he says. While many financial institutions may claim sophisticated processes and platforms, “How many limit systems monitor and mark in real time? How many compute the volatility surfaces to calculate your delta exposure when you have a book of OTC [over-the-counter] options?” he asks. “I think this is going to change as people get more agile and they will accept that processes have to be reviewed based on market conditions.”
Review underway
The good news is that lessons appear to have been learned, Mr Carrel adds. “The massive review that’s going to take place has already started… Some of the big banks did not even wait for the weekend [of January 26th], they started on Thursday or Friday to launch a review of their own limit systems.”
Mr Kerviel joined SocGen in 2000 in its middle office, where, the bank alleges, he amassed the knowledge of internal processes that would allow him to skirt controls some years later. He moved to the front office in 2005 as a junior trader in the bank’s delta one arbitrage desk, trading in European equity index futures. In theory, this strategy does not allow the trader to take on significant directional risk, as any directional positions are hedged.
According to SocGen’s “Mission Green” report into the incident, Mr Kerviel began to take bets on the market in 2005. To do this, he took genuine directional positions and created fictitious hedges, buying securities and warrants with deferred start dates and futures with a counterparty that did not require instant confirmation. Using other employees’ access details, he was able to later delete trades from SocGen’s system, leaving him with massive exposures, but fooling the monitoring tools into thinking that his P&L was relatively flat.
Although he began to systematically circumvent procedures in 2005, his activities were not discovered until January 18, when the internal investigation began in earnest. “The failure to identify the fraud until that date can be attributed firstly to the efficiency and variety of the concealment techniques employed by the fraudster, secondly to the fact that operating staff did not carry out more detailed checks, and finally to the absence of certain controls that were not provided for and which might have identified the fraud.”
Eurex, the derivatives exchange, flagged up in November 2007, that Mr Kerviel’s positions showed some irregularities. The Mission Green report says that compliance and the trader’s managers “were satisfied, without verification, with the trader’s explanations, in contradiction to Eurex’s assertions.”
The Mission Green report has a chronology of other incidents that were detected but either not escalated or acted upon with sufficient force to warrant an investigation, including questions raised at Fimat, the brokerage that was at that time a subsidiary of SocGen.
For some, this is simply an internal control issue, although the apparently complex nature of the fraud seems to suggest that it was unavoidable – undetectable without massive manual intervention. Not so, says Bruno Piers de Raveschoot, head of Actimize Europe, which provides compliance and fraud monitoring software for financial services clients, including Credit Agricole Asset Management. There were, he says, clear warning signs, not least the trader’s unusual behaviour and concentration of instruments with low margin calls. Any abnormal practices should have been flagged up, as should the dominant position he had taken in the market, according to Mr Piers de Raveschoot.
However, rather than individual failings in the bank’s fraud measures, Mr Piers de Raveschoot believes that SocGen’s woes were caused by a failure of those systems to communicate with each other to create an aggregated case file: “I think in this specific case, the problem is more that there was no specific system in place that was linking all the problems to one instance. So, for example, they received letters from Eurex warning that their strategy was very dangerous, they knew that some anti-theft system did spot that this trader was using the password of some other employee, but those isolated cases were not strong enough to stop everything. If these had been aggregated into one single case, they would have stopped it a long time ago.”
According to Actimize’s figures, such systems are yet to achieve significant penetration across the industry. “Globally, less than 50 percent of large financial groups have fraud monitoring systems in place, and less than 8 percent have sophisticated systems,” says Mr Piers de Raveschoot. After the latest storm, it is likely that this will change. “We already have information that people are shifting their budgets. There is a very strong and positive reaction by the industry,” he says.
 | The need for aggregation of risk and control information is a view that is shared by Chris Leong, operations director, treasury and capital markets at Misys. Having studied the SocGen report, Mr Leong believes that the lack of a single, independent monitoring system that proactively flagged anomalies across business lines. “[At SocGen] whenever a discrepancy or an anomaly was reported to the business lines, it didn’t go any further, it was contained within lines of business,” he says. |
“Any time an anomaly or discrepancy occurs… you want to make the people who are ultimately accountable aware of this risk. If you study the transcripts of the SocGen report, often the trader would make some explanation to that and it would be either dismissed, or the case was closed,” says Mr Leong. “At no stage I saw evidence of senior management being involved in the decision making process. I think the notion of independent monitoring – getting information further up the tree – is going to be very important going forward for most of these firms.”
SocGen has begun to reinforce its control procedures in the wake of the incident. This includes using biometric identification systems to prevent the misuse of passwords, reinforcing alert procedures so that information is passed between units and reaches managers, and “strengthening the organisational structure and governance of the operational risk prevention system to develop its cross-functional nature and better take account of the fraud risk, including from a human resources perspective.”
The last point could be the most profound. FT Mandate spoke to two individuals involved with risk management and IT at Barings in the run-up to and the aftermath of the 1995 Nick Leeson trading scandal that ended with the bank sold for £1 to ING. Both revealed that there were startling similarities between Barings and SocGen, and both were adamant that improving systems alone will not shield the financial services industry against future losses of this kind.
Front vs middle office
The case-by-case similarities between SocGen and Barings, and indeed, John Ruscnak’s fraudulent trading at Allfirst, could mask the main issue, which is the dynamic that exists between the front and middle offices: the money makers and the cost centres.
“I’ve seen it happen time and time again… where you have somebody who’s at the front end of the business wants to do a trade that doesn’t quite conform with the rules that have been laid down by management… and they know the guys in middle office, and they ask ‘can you just let it through on this occasion,’” says one of the risk professionals. “The poor guys in middle office are in a difficult position. If they don’t feel they have the support of the management, then they are under a huge amount of pressure to allow trades to go through.”
The majority of trading firms are set up to support the risk takers that make the firm’s money. If a trader is pulling in profits, the back and middle office find themselves routinely bullied or sidelined, according to both of those familiar with the Barings debacle. Both remain in the industry, and continue to see the same cultural issues, where the back and middle office, who struggle to quantify their return on investment at the best of times, forced to, by direct action from traders or by culture, kowtow for fear of being labelled obstructive. In many incidences, they say, while a trader is making money, he is almost untouchable.
The emergence of the C-level risk officer at some firms is a major step forward, but, as one professional says, in the end, the ideal solution would be to entice more traders to take on risk roles, bringing with them gravitas and knowledge of the tricks of the trade. However, he says, “it comes down to how you pay people.” Traders get paid more than support staff, so unless the lifestyle benefits of a role in the back office can be played up, there is little chance of this happening systematically. Until that point, he says, “the biggest risk to the bank is its management.”
“They’re looking at what specific control was missed, what specific area was not working properly, and what’s missed in a root cause analysis is that cultural aspect,” says John MacKessy, MD in the forensic and litigation practice at FTI Consulting. “Ultimately when we talk about the risk management system, we’re talking about the entire control environment, and I think sometimes when we talk about what technology they had in place what gets lost is the structure, and that really starts at the top.” There is still an over-reliance on technology amongst management, he says, and a tacit acceptance that a rogue trader is only “rogue” when he loses money. Cultures are hard to change, and even if escalation procedures for middle office staff are set down and chains of responsibility reinforced, examples may need to be made. “Once you enforce a sanction, one thing that I haven’t seen at the trading firms, is the publication of those sanctions,” Mr MacKessy says. “I think that could have a reverberation across the industry.”
When FT Mandate asked if the apparently growing momentum behind a movement to “explain your profits as well as your losses” could ever throw up the intriguing case of a rogue trader fired for making his firm a hefty profit, most risk professionals gave similar responses: “You’ll be waiting a long time.”
SIMILARITIES WITH THE BARINGS 'ROGUE TRADER'
The Nick Leeson incident at Barings has become the archetypal ‘rogue trader’ story, inspiring a film of the same name. Mr Leeson operated out of the Singapore office of Barings bank, trading futures and options on the Osaka Securities Exchange and Singapore International Monetary Exchange. Over the course of two years, he secretly piled up losses in an account – number 88888 – while taking progressively larger bets to cover himself.
The similarities between Mr Kerviel and Mr Leeson are immediately obvious. Firstly, both had a familiarity with the back office, which enabled them to successfully disguise their activities. In fact, Mr Leeson was in charge of settlement for his own trading.
Secondly, Mr Leeson’s management believed, according to a senior risk operative, then at Barings, that the Singapore office was pure arbitrage, and hence low risk. Little in-depth analysis of his positions was performed, even when, as in 1994, external auditors raised concerns. Mr Leeson provided forged faxes to cover his trail. Thirteen years on, Mr Kerviel is said to have falsified emails to protect himself.
Thirdly, both traders were successful up to a point, but both were uncovered as short-term volatility forced them into a catastrophic endgame that revealed the extent of their deception. Of the $1.3bn loss that Mr Leeson accumulated, two-thirds was due to an ill-fated $3.6bn bet on the Nikkei in early 1995. At his best, Mr Kerviel was in the black to the tune of €1.5bn, before the market turned.
