June 11, 2010 7 Comments
Jerome Kerviel, accused of being a rogue trader, is now on trial. SocGen lost $7bn in the incident which heads the list of major trading losses.
How did he do it?
This is actually a very similar situation to Nick Leeson at Barings – number 11 in the top list. They were both involved in forms of arbitrage, which exploits tiny differences in price which ‘shouldn’t’ really be there. In fact, pricing theory fairly obviously requires that there can’t be a price difference between two identical items. If that were false – say if one loaf of bread had a different price to an identical one – then I could make a risk free profit by buying at the low price and selling at the high price. And there can’t be a risk free profit because everyone would pile in. You can see that what would happen would be that the prices would equalise.
Now this is what the arbitrageurs exploit. It all hinges on what ‘identical’ means. Not quite identical introduces some risk. Leeson was buying one product in Osaka and selling the same product in Singapore. Clearly if the product is the same, exactly, there is no risk. You might ask what might cause a price difference – there might be transient local factors such as someone big in Osaka decides to buy something. And then there could be a delay before Singapore catches up. And that catch-up process is exactly what the arbs do.
Kerviel was involved in arbing equity index futures and underlying equities. Equities are stocks, indices are groups of stocks like the FTSE-100 and equity index futures is just a bet on where the FTSE-100 will be in six months from now. Clearly you can do that on a risk free basis if you, say, sell the index and buy all the stocks in it. [Incidentally, if you want to be an insider trader but don’t want to go to prison, maybe you could buy an index in which the stock you can’t trade figures and then sell everything in the index except the one you aren’t allowed to trade…but I don’t recommend it…]
Why is it dangerous?
There are two common factors between this case and Leeson. In both, the alleged misdeeds were possible because the trader and the back office person were effectively the same person. Leeson actually did his own monitoring, an extraordinary failure which rightly cost the jobs of many at Barings. I could go further and say it was so remarkable that everyone involved in the company deserved to lose all their cash, but I know there were lots of Barings debentures held by grannies and I suspect we can’t expect them to have known what they were doing. While Kerviel came from back office himself and knew the control systems and would have known how to defeat them. I also will claim that back office types are rather easy for front office traders to browbeat and this history will have played a part in Kerviel’s psychology and the desire to get somewhere fast.
Secondly, because you are exploiting tiny price differences, you need to trade in vast amounts. And all the time. The control problem comes when you do not have offsetting equal and opposite trades but wind up taking huge uncovered positions. Leeson sorted this out with a fax purportedly evidencing a large receivable from a hedge fund. Towards the end, he was drawing in funding from all over Asia, which should have alerted someone.
What is odd about this case?
You can’t make large amounts of money from arbitrage. You just can’t, because risk and reward are closely linked. You can see from the loaf of bread example that that has to be true. So if you are a manager in an I-bank, you need to get very concerned if your arbitrage desk is making large profits.
Now this leads to the strange consequence that Kerviel must have been concealing large profits. And this is what you see.
“During the largely procedural first day of the trial, Kerviel’s lawyer said Societe Generale would have been clearly able to see data showing Kerviel’s extraordinary profits of 1.4 billion euros at the end of 2007”
Note that this is profit not revenue, and that SocGen as a whole might typically make a net profit around EUR600m in a quarter. Do you think you could spot Kerviel in there?
“Seated on a plastic chair in front of rows of lawyers in black garb, the ex-trader said his annual salary at Societe Generale was 48,000 euros in 2006 with an annual bonus of 60,000 euros”
Now that is not a lot of money for traders. They might typically expect to make 5% to 10% of what they produce, or more in some cases where they are reliably producing large returns. Apparently Kerviel was expecting to make EUR300,000 for 08, on a declared profit of EUR60m. That’s a 0.5% return. You can see that this is not enough. Someone with that type of track record could just set up on their own, use the track record to raise funds, and trade themselves for maybe 50%. There is another type of arb there.
The GBPEUR exchange rate in 07 was 0.67, so we are talking about someone earning a salary of £32k. This is not far north of what we used to pay graduate trainees in London. So what we have here is someone being paid back office amounts, a French I-bank culture in which you shouldn’t really pay very much or have high quality people, and back office resentment of the flash and the furious.
“Lawyers also read a transcript of a conversation between Kerviel and SocGen’s ex-investment bank chief Jean-Pierre Mustier when the scandal broke, in which Mustier reportedly said: “If you won 1.4 billion euros, that means you’re very good. What you did was a pain, but it’s not a big deal.”
If Kerviel can make that out, then Mustier has failed in a stunning way to understand what arbitrage is. It is a French word, after all. It may be difficult to see how Kerviel can avoid jail, but he cannot have been on his own in this one.