Disclosure (or apparent disclousues, who knows if we will ever learn the true story) of how equity derivatives trader Jerome Kerviel caused the biggest trading loss in banking history continues to dribble out.
Today, Bloomberg in “Societe Generale Says Trader Built Up Positions of EU50 Billion,” gives more detail on how the trader caused so much damage:
Societe Generale SA said trader Jerome Kerviel built up positions in European stock index futures of 50 billion euros ($73 billion) before the French bank discovered the trades and unwound them last week.
Kerviel, 31, who is being questioned by France’s financial police for a third day, took advantage of the bank’s practice of checking only net trading positions rather than gross bets to conceal his subterfuge with phony hedges, Paris-based Societe Generale said yesterday. The trading loss of 4.9 billion euros was the biggest in banking history.
The failure to look at gross as well as net exposures is a basic shortcoming and the press and competitors will make a great deal of noise about that.
While further comments will no doubt leak out about the specific failings in SocGen’s risk control system, one glaring lapse leaped out:
He took only four days off last August and postponed a vacation at the end of the year, Societe Generale said.
Huh? It is Trading 101 to make staff take a minimum of a full week off at least once a year, better twice, precisely because it is well nigh impossible for sole actors to keep frauds going from afar. For example, the $1.8 billion in losses at Sumitomo Corporation were generated by a copper trader who hadn’t taken a vacation in over two years.
This factoid alone says that SocGen’s systems (and by that I mean not just computerized risk controls, but broader management practices) were woefully deficient. Managing traders isn’t just about designing computer systems to find anomalous patterns; it also involves a certain amount of judgment about overall risk, staff capabilities, procedures, and organization design. The discussion of how Kerviel did what he did appears to be focusing unduly on computer and system failures, and missing the point that they might have reflected or even reinforced larger organizational design failures.
Note I don’t dismiss the possibility that the charges against Kerviel are either trumped up or considerably exaggerated. If they fail to hire an outside firm like Eugene Ludwig’s Promontory Capital, which is full of former regulators and has conducted a number of high profile rogue trader investigations, to prepare a report and recommend improvements, that suggests there may be more to this story than SocGen is revealing.
17 comments
Comments are closed.
Yves
The FT (or bloomberg) also pointed out promises by banking authorities to remedy the glaring error of just checking trades for net figures, which is probably a more pertinent issue in the SG case.Just blogged on that, not repeating that. What’s worrying is which other bank has such loopholes?
foesskewered,
The gross/net difference is a big lapse too, but the media is all over that one. The failure to insist that staff take vacation is so basic and low tech that it suggests sloppiness on other managerial fronts.
SocGen is also saying they caught him in suspicious trades more than once, but he cancelled them which allayed their suspicions. This too suggests a lack of adequate management procedures, although you’d have to dig deeper to be sure.
From the same article:
Kerviel’s positions, mostly on Germany’s DAX Index and the pan-European Euro Stoxx 50, had losses of 1.4 billion euros when Societe Generale discovered the fraud on Jan. 18. The bank said it lost 3.5 billion euros more liquidating the positions.
Societe Generale also used a single, very senior trader to unwind Kerviel’s positions.
I guess the question is, could they have handled this more adroitly and minimized their losses, unwinding over a longer time period? Was the magnitude of their loss partly self-inflicted?
The short answer is yes. LTCM, which everyone knew had a major major underwater positions, was quite deliberately wound down over a long period of time.
As long as no competitors knew how exposed SocGen was, they have taken their time unwinding the trades. The risk is that someone finds out you are positioned in a way you are trying to escape. They know you have an overhang; they’ll sell just to get out at a better price than you (if they fear that their positions will be hurt by your liquidations) or they might sell opportunistically just to cause pain.
I can tell you from painful personal experience, the vast majority of the time, selling into a panicked market is a very bad idea.
My Question is how many traders are fired or arested for profitabile trades? I think some discusion about that “problem” also needs to be had.
50 bn€ is a very big position.
Maybe SocGen has closed the position partly by buying out of the money puts and now some other party has a problem?
I am not an expert on liquidity in equity futures – but to cover a 50bn position into a market breakdown looks quite difficult to me….
Maybe they sold correlated futures (US? Nikkei? )too, bought bond-futures, sold equity calls, whatever……
Any suggestions?
Dear Sir,
Was it possible to hedge such a position ?
Apparently, Mr. Bouton informed the Autorités des Marchés Financiers (AMF = french SEC) and Banque de France on sunday.
Do you think they could have force him to sell ?
Happy New Year anyway !
“If they fail to hire an outside firm like Eugene Ludwig’s Promontory Capital, which is full of former regulators and has conducted a number of high profile rogue trader investigations, to prepare a report and recommend improvements, that suggests there may be more to this story than SocGen is revealing.”
Oh bring on the consultants, by all means. They need to be fed.
“Huh? It is Trading 101 to make staff take a minimum of a full week off at least once a year, better twice….”
He took 4 days off in August (1 day less than the “full week”) and since a “full year” goes back to January 2007, it’s not clear to me that he didn’t take off the time you say is required, since he probably took off time between Jan 2007 and April 2007 (April 2007 because that meets the 8 months, if I recall correctly, of “no vacation” which Mianne mentioned in his speech to staff).
On the other hand, this is France, and SocGen traders are entitled to 50 days per year. So it is true he was on the short end of the stick in terms of vacation leave.
As to gross versus net trading limits, c’mon. The first limits put in by SocGen for basket trading were nominal (gross) limits, like 15 years ago. There are limits on the number of front-month futures, etc. The trader was in a different sub-activity, I believe Delta One listed products, and he probably was supposed to have such a small position that it was inconceivable that he would have anything like the position he actually got.
But there was and is a problem with systems. Management never figured out how to close the circle between the real world (cash flows, margin calls, and fees) and positions shown in the front-office system. They just weren’t systems guys, and worse, they didn’t realize they weren’t.
a,
SocGen under attack from multiple fronts: did Kerviel really do this much damage, or is management trying to attribute a whole heap of bad positions to him? Were more senior people asleep at the switch? The one way to clear matters up with the world at large is to bring someone in.
And FWI, Luwdig makes his real dough from principal investing. He is in the consulting business more to burnish his reputation (but yes, it more than pays for itself).
Anon of 6:45 PM,
The article never said that the four days Kerviel took in August were consecutive days off. It could just as well have been four Fridays off.
Kerviel was supposed to have a small NET position because he was doing arbitrage. You can rack up very large gross positions and have a small net.
His fraud, at least as reported, did not trigger any warning signals via a failed reconciliation with the cash system because (at least as reported) his moves did not involve any trades that led to cash payments. There may have been other failures to cross check, but that wasn’t one of them.
“Did Kerviel really do this much damage?”
Yes
“Is management trying to attribute a whole heap of bad positions to him?” No and a laughable claim.
“Were more senior people asleep at the switch?” Sure, you don’t lose that much money without something not being checked which should have been checked.
“The article never said that the four days Kerviel took in August were consecutive days off. It could just as well have been four Fridays off.” You’re the one who got high and mighty about Trading 101. The burden of proof is on you.
“Kerviel was supposed to have a small NET position.” Duh, yes.
“His fraud, at least as reported, did not trigger any warning signals via a failed reconciliation with the cash system.” What are you talking about? The guy was long thousands of futures, matched by fictitious deals. He was not supposed to have this position size (of real futures). SocGen’s failure was not matching the fees for buying the futures and the margins that the positions created to the trader. Instead the Back looked at the fees and the margins and thought, “Oh that’s still normal for us,” without every breaking it down to the trader level, where management could have seen that the fees and margins were being created by a trader who shouldn’t have been doing this.
Anon of 9;34 AM,
I frankly don’t know what you are taking about when you mention “fees” for futures. SocGen almost certainly has seats on all the relevant exchanges. They have direct access to the order system, so there is no broker involved.
As for your other comments, I turn the mike over to FT Alphaville:
firstly, he ensured that the characteristics of the fictitious operations limited the chances of a control; for example he chose very specific operations with no cash movements or margin call and which did not require immediate confirmation;
he misappropriated the IT access codes belonging to operators in order to cancel certain operations;
he falsified documents allowing him to justify the entry of fictitious operations.
he ensured that the fictitious operations involved a different financial instrument to the one he had just cancelled, in order to increase his chances of not being controlled.
By all means turn over the mike, but it’s clear you have no idea what happened, so perhaps you should be quiet before you start to speculate about conspiracy theories.
Fictitious operations cannot create losses in the real world. What they can do is hide risks induced by real operations, which can create real losses. On the one hand we have a position of x Eurostoxx futures, counterbalanced by a position of fictitious operations. The position of x Eurostoxx futures – real not fictitious operations – created the 5 billion euro loss. Ask yourself why this real position was not discovered and how it could have been discovered.
It seems you are much more interested in proving yourself right than in dealing with facts. Had you bothered to click on the FT Alphaville link, you would have found out that the text came from SocGen itself. So if you think you know what happened better than they dio, I suggest you volunteer your services to them.
Yves,
You should have slapped around 3:18 harder. He can’t even read the stuff you quoted correctly, he is more interested in flaming you.
And Kerviel himself agrees about the vacation thing:
Le simple fait de ne pas prendre de jours de congés en 2007 (4 jours en 2007) aurait dû alerter ma direction. C’est une des règles primaires du contrôle interne. Un trader qui ne prend pas de vacances est un trader qui ne veut pas laisser son book à un autre.
(from Le Monde today)
Wow! Thanks.