大象传媒

SocGen unhedged

• Robert Peston
• 28 Jan 08, 09:56 AM

I鈥檓 not sure whether yesterday鈥檚 鈥渆xplanatory note鈥� by Societe Generale about what it describes as its 鈥渆xceptional fraud鈥� was supposed to be reassuring.

But it didn鈥檛 do the trick with me. It left me feeling more anxious than ever, not only about how SocGen got into its mess but about whether controls are appropriate at other large banks that are stewards of our cash.

Here鈥檚 why.

As I explained last week on 大象传媒 television, Jerome Kerviel鈥檚 alleged method was to take very big bets on the direction of markets, and then only pretend to hedge them by making fictitious bets in the other direction.

This fictitious hedging made it look to his employers, for a year or so, that his net trading position was close to neutral 鈥� or that if that one bet swung massively out of the money, it would be offset by a swing into the money by the matching bet (and vice versa).

Let鈥檚 push to one side, for now, the mystery of his motivation. And that is perhaps the most gripping mystery of all, because even if 鈥� as his lawyers claimed yesterday 鈥� his real trades were in profit to the tune of 鈧�1.5bn (拢1.1bn) at the end of last year, there was no way he would ever have trousered any of this: that gain would have been sitting in his employer鈥檚 account, not his.

Also, let鈥檚 not fixate at this moment on what may enrage SocGen鈥檚 shareholders, which is the bank鈥檚 disclosure that as of mid-day January 18, M. Kerviel鈥檚 position was in balance 鈥� but that the massive loss of 拢3.7bn was only made real by the bank鈥檚 actions of selling M. Kerviel鈥檚 positions into a declining market.

Now I have been told by a banker advising SocGen that the Banque de France, France鈥檚 central bank, instructed SocGen to close out the position as quickly as possible.

But the official explanation from SocGen is that its chairman, Daniel Bouton, opted to sell tout de suite.

If so, SocGen鈥檚 shareholders may well be feeling less than enamoured of M. Bouton.

And, I have to say, if SocGen were a British bank, my instinct is that M. Bouton would already have been ejected from the boardroom.

But all of that, to an extent, is a sideshow.

The only thing that really matters is how M. Kerviel was able to bet his entire bank on the direction of stock markets and whether the flaws in SocGen鈥檚 controls also exist in other big banks.

What doesn鈥檛 provide comfort is that a senior banker with one of the world鈥檚 largest and most respected financial institutions 鈥� who has been drafted in by SocGen to help sort out its mess 鈥� told me he didn鈥檛 think SocGen鈥檚 risk-control techniques were any different or worse from most other banks.

The point is that SocGen鈥檚 method for controlling the risk to the bank from the activities of traders like M. Kerviel relied on controlling their respective net exposures to markets, rather than taking into account their respective gross exposures.

In fact, SocGen鈥檚 latest statement says that M. Kerviel was encouraged to engage in a 鈥渧ery large amount of operations involving very high total nominal amounts鈥�.

SocGen probably thinks it is just stating the bloomin鈥� obvious, given that M. Kerviel鈥檚 official job was in equity-markets arbitrage.

And the various traders who read this column will probably accuse me of being needlessly alarmist.

They will point out that M. Kerviel was supposed to generate profits for the bank by exploiting minute differences in the buying and selling prices of almost-identical financial instruments, in his case European stock index futures or instruments with near-identical characteristics to those futures.

And in order to do that, he needed to trade on an enormous scale.

If done properly, they鈥檒l say, the risks are minimal.

Perhaps it鈥檚 understandable that SocGen hadn鈥檛 properly allowed for the mad-genius risk, that a trader should have the desire and ability to gull the bank for no apparent personal gain.

But my fear is that SocGen, and possibly other banks, may be unduly optimistic about their ability to hedge a large position so that it becomes, in net terms, a small position.

Here we have to get into the detail of what M. Kerviel did.

According to SocGen, his fictitious hedging portfolio consisted of 鈥渧ery specific operations with no cash movements or margin call and which did not require immediate confirmation鈥�.

The relevance of this is that it shows how M. Kerviel sustained his alleged game for so long.

If M Kerviel had hedged his bets in a normal way, there would have been margin calls 鈥� or demands for cash from those he was dealing with, known as counterparties 鈥� when those hedging bets moved against him.

So alarm bells should have rung at SocGen when there were no margin calls.

However, M Kerviel silenced the alarm bells by putting in place special hedges which involved no margin calls.

Now, his ability to do this is troubling enough.

But I find it difficult to believe that even if those hedges had been real, as opposed to figments constructed in SocGen鈥檚 computers, they would have been safe.

A hedge requiring 鈥渘o cash movements, or margin call鈥� doesn鈥檛 sound like a very liquid hedge, or one that could have been traded out when needed.

Forgive me for this long pre-amble, but we are now at the nub of what worries me 鈥� that SocGen was allowing its traders far too much exposure to liquidity risk.

Let鈥檚 just say that everything M. Kerviel had done had been real, can we really be confident that his portfolio would have withstood a drying-up of liquidity of the sort we have experienced in the credit markets since last August?

Was there a degree of naivet茅 built into SocGen鈥檚 assessment of the risks being run by all its traders?

We need to know a great deal more about the hedges it permits in its arbitrage department.

In particular, we need to know that when a trader legitimately makes a 鈧�50bn bet, his or her hedge really will behave in the way that it says on the label 鈥� because if it turns out to be an illiquid hedge, it might well be no hedge at all.

These may sound like technical issues. But they matter to all of us, because they directly concern the robustness of the controls imposed by all the banks on which we rely.

PS. M. Kerviel鈥檚 day job in equity index arbitrage sounds complicated, but it鈥檚 not really.

Think of it as buying a load of apples at 50p a pound from market-trader Fred on one side of the street, and then selling them to trader Boris on the other side of the street for 50陆 p per pound.

If you sell enough apples to Boris before he reduces his price to 50p, you will have pocketed a few pennies.

In fact, what M. Kerviel did was a little bit more complicated than that. He was dealing in futures.

In effect, he made a deal with Fred that he would buy a load of Granny Smiths from him at 50p in the pound in two weeks time, and he simultaneously agreed to sell Boris an equivalent amount of Grannies at 50 陆 p, also for future delivery.

Now you can鈥檛 get rich by selling ten apples at a 陆 pence profit.

But if a bank buys and sells 50 billion apples, that鈥檚 a lot of 陆 pences.

Which is why M. Kerviel had a licence to take enormous 鈥渘ominal鈥� positions.

M. Kerviel was allowed to buy 50 billion apples from one group of traders, so long as he simultaneously sold 50 billion apples 鈥� or some similar fruit, exhibiting similar price behaviour to apples 鈥� to another set of traders.

Unfortunately what M. Kerviel actually did was buy all those apples, while only pretending that he matched all those purchases with sales.

And that鈥檚 how he upset the apple cart.

• Permalink
• Comments