Société Générale

Jan 31st 2008
From The Economist print edition

The humiliated French bank has plenty more explaining to do before putting its rogue-trader scandal behind it

AN OLD line of Hank Paulson's has been dusted off since news broke of a €4.9 billion ($7.2 billion) trading loss at Société Générale, France's second-largest bank. “We will never eliminate people doing bad things,” the former head of Goldman Sachs, now America's treasury secretary, once said. “In a town of 20,000 people, there's a jail.” The question now being asked of SocGen is: shouldn't there also be a police force?

In fact, SocGen has plenty of internal cops at its high-security headquarters in the La Défense enclave of Paris. The bank's annual report for 2006 devotes 26 reassuring pages to its risk-management practices; more than 2,000 staff worked in the function that year, and lots more bodies were added in 2007. Yet none of them stopped Jérôme Kerviel, the trader accused of taking enormous unauthorised bets, from building an unhedged €50 billion exposure to European futures markets (Mr Kerviel reportedly alleges that his supervisors were aware of his activities).

On January 28th Mr Kerviel was placed under formal investigation for abuse of trust, breaching computer security and falsifying documents. Two days later Daniel Bouton, SocGen's chairman and co-chief executive, survived a board meeting to consider his handling of the affair. He was lucky. Holes have not only appeared in the bank's accounts; its initial version of events is also looking threadbare.

Mr Bouton's description of Mr Kerviel as having “an extraordinary talent for dissimulation” certainly looks less convincing as more details emerge. Although Mr Kerviel was properly found out on January 18th, he had tripped alarms inside the bank well before then. “When challenged, he was clever enough to say, for example, that he had made a mistake,” says Jean-Pierre Mustier, the head of SocGen's investment-banking arm. Clever, indeed. Rival banks, admittedly fortified by anonymity, say that their traders would not have been able to keep getting away with that kind of explanation. Some people wonder if SocGen would have blown the whistle at all had the bets been profitable. Outsiders raised other suspicions. Eurex, Europe's largest futures exchange, contacted SocGen about oddities in trading patterns in late 2007, which the Paris prosecutor says referred to Mr Kerviel's positions. The bank says that Eurex's questions were rather technical in nature and that it had responded to them.

The sheer size of Mr Kerviel's exposure, the losses on which tripled as SocGen frenetically unwound its positions between January 21st and 23rd, has caused most bafflement among veterans of the futures markets. Two big blind spots saved him from detection. The first was the bank's focus on traders' net exposure, the difference between the portfolios that are being arbitraged. Mr Kerviel did not have a defined gross-exposure limit. By creating a fictitious portfolio of trades that appeared to balance those he was really making, his net exposure stayed within set ranges and he remained below the radar.

Why didn't the margin calls on Mr Kerviel's real trades (likely to have been of the order of €2.5 billion on a €50 billion position) trigger alarms? This was the second blind spot. According to Mr Mustier, margin data from Eurex showed only consolidated positions. These positions were “not a different order of magnitude” from the volumes expected of a big investment bank. “One lesson of this is that it is important to see what is attributable to each trader,” he says. In truth, that should not have been too difficult: as well as consolidated figures, Eurex says it does already send data that tie margins to specific traders on a daily basis.

Mr Kerviel's credentials as a supervillain look less impressive in other ways too. His fictitious portfolio did not just comprise over-the-counter transactions with big banks, where agreed credit limits meant he could avoid margin calls. Embarrassingly, it also included trades with other parts of SocGen. Access to risk-control codes did not necessarily require the skills of a seasoned hacker; the bank says he may simply have offered to input details of trades on behalf of middle-office people when there was lots of activity on the trading floor. Some control procedures appear to have been predictable, being timed to take place shortly before the settlement date of futures contracts. Mr Kerviel's limited holidays and late nights should also have raised red flags.

Despite the resounding support of the board, Mr Bouton has been severely weakened. Nicolas Sarkozy, the French president, has made pointed calls for SocGen executives to face “consequences” (see article). The chairman may not survive the findings of an internal investigation into the loss. The future of the bank itself is also in doubt. Its shares have slumped since the start of the year (see chart) and its credibility has been shredded, not just by the trading loss but also by write-downs of subprime-related investments. Analysts are trying to work out who might be in a position to buy the bank; one option is a joint approach by two other French institutions, BNP Paribas and Crédit Agricole, with BNP taking SocGen's retail operations and Crédit Agricole the investment bank.

The question now is whether the flaws that seem to have torpedoed SocGen are endemic to other banks. There are some reasons to think not. SocGen's management was being criticised for poor standards of disclosure well before news of the Kerviel affair broke: “They had not imposed the important discipline of transparency on themselves,” says John Raymond of CreditSights, a research firm. In common with other French banks, SocGen was also thought by many to take an overly mathematical approach to risk. “‘It may work in practice but does it work in theory?' is the stereotype of a French bank,” says one industry consultant. More obvious visual cues—beads of sweat running down Mr Kerviel's face, say—may get overlooked in this type of environment. And the sheer size of the loss partly reflects the bank's pre-eminent position in the equity-derivatives field.

But there is no cause for complacency. Most observers, regulators included, concede that banks can do little to stop determined individuals from sidling around controls, at least for a while.

And in the midst of the subprime crisis, the SocGen saga resonates for other reasons too. One concerns the status of risk managers within banks. Mr Kerviel is alleged to have outfoxed rather than pulled rank on risk managers, but swaggering traders can find it all too easy to ignore the concerns of meek back-office types. Many Wall Street banks have responded to the meltdown in structured credit by strengthening their risk teams. The SocGen loss will accelerate that trend.

Events at SocGen will also fuel an old debate about bankers' pay—Mr Kerviel was reportedly motivated by his desire to win a higher bonus—and a newer one about the ability of banks to keep pace with the dizzying growth of derivatives markets. It will also reinforce concerns about how “fat-tail”, or extreme, risks correlate: might SocGen's risk managers have been too distracted by its subprime woes to keep watch on the futures desk?