Former Société Générale rogue trader Jérôme Kerviel owes the bank $6.3 billion. Here's what his case tells us about financial reform.
He can earn million-dollar gains without anybody knowing. He can execute make-believe trades by sending fake emails from hacked computers. He doesn't always lose money. But when he does, he loses more than $6 billion. He is ... the most indebted man in the world.
Jérôme Kerviel is learning one of life's harsher lessons: It stinks to be $6.3 billion in debt. That's how much his fraudulent trades in 2007 and 2008 cost French bank Société Générale, and how much he has been ordered to pay in restitution -- after he gets out of jail in three years. A judge recently upheld these terms on appeal.
Stay risky, my friends?
But how exactly do you carry out €50 billion ($73 billion at the time) of unauthorized trades? With lots of computer hacking and not a lot of vacation. Kerviel was an arbitrage trader at Société Générale who wanted to take out the "arbitrage" part. In plain English, arbitrage just means taking advantage of discrepancies when things should have the same price, but don't. The idea is to buy the cheaper one, sell the more expensive one, and then wait for them to converge. The beauty is it doesn't matter whether markets go up or down -- you're both long and short -- just that the prices actually converge.
This is what Kerviel was supposed to do with European stock futures. He only did half of it. In other words, he took, say, a long position and then pretended to offset it with a fictitious short position. That's where the computer hacking came in. But how did he end up betting such mind-boggling sums? Well, arbitrageurs are usually exploiting such small price discrepancies that the only way to make decent money at it is to bet lots and lots of money. And Kerviel certainly did that. Société Générale didn't notice because it only monitored the net, and not the gross, value of these paired trades -- which again, were not really paired. So the bank took Kerviel's actual bets and subtracted his fake bets. Voilà, €50 billion worth of trades disappeared. Of course, it's impossible to maintain a continuous fraud like this if you're not at work. That's where vacation, or the lack thereof, comes in. That's another banking no-no, precisely to prevent this kind of chicanery. By the time Société Générale uncovered the fraud, Kerviel had built up huge one-way bets and huge one-way losses.
Kerviel certainly isn't the first rogue trader, nor will he be the last. But is there anything we can do about these massive failures of bank risk management? And just what happens when you owe billions and billions of dollars? I talked to Frank Partnoy, a former investment banker and current law professor at the University of San Diego whose book Infectious Greed offers a detailed financial history of the past few decades, about this and more.
First off, how does a $6 billion fine even work? Will they garnish his wages for life?
Well, he's obviously not going to be able to pay the fine. What happened to Kerviel is the financial equivalent of sentencing someone to life plus 100 years. Garnishing is usually what they do in cases like this. There's the fine, and then there will be a structured settlement. They'll likely reach some kind of agreement where a significant percentage of any money he makes for the rest of his life will be paid into a fund to cover the fine. He'll be like Sisyphus pushing the boulder up the hill every day for the rest of his life. Now, one of the things this shows is the sword of Damocles will be hanging over you forever if you're convicted of this sort of offense.
How does Kerviel's fine compare to what we've seen for some of the big banks in the wake of the financial crisis?
Once you get above ten million dollars or so for individuals, the numbers no longer matter all that much. But in terms of comparison, it's worth noting that this single fine for a single trader is significantly higher than all of the other settlements that have come out of the subprime crisis, at least so far. For instance, it's 12 times larger than what Goldman Sachs settled for.
Why do big banks seem to have so much trouble managing hot-shot traders like Kerviel?
The big picture point here is there's a continuous thread running through the past few decades with these kinds of rogue traders. They keep coming up with new mousetraps and new ways to evade detection, so we shouldn't be surprised that it keeps happening. There are two possible answers for why this is. The first is just the sheer size of the big banks. These are large organizations and it's impossible to police everyone's conduct. Inevitably, your'e going to have a handful of people who will evade your attempts to detect them. In a way, technology makes it worse, because there are more complicated financial products and computer systems. If people really want to violate your internal trading rules and guidelines, you'll catch many of them, but some of them will succeed. As a footnote, we hear about Jérôme Kerviel and we hear about the London Whale, because they involve multiple billions of dollars, but there are many Jerome Kerviels happening all the time. They just aren't large enough to be deemed material, so they aren't disclosed.This is part of the fallout of having even larger too-big-to-fail banks -- as they get bigger and bigger, these kind of rogue traders become arguably immaterial, so we don't hear about them. The banks are well aware that rogue trading is kind of a cost of doing business in the complex financial world that we live in.
Now, there's a second way of looking at this problem. According to this story, senior bank managers are well aware of the risk of rogue trading, and have, in fact, set up the risk management systems in a way that makes at least some rogue trading expected and almost inevitable. The idea is they give their traders leeway. As a manager, you don't want your traders to be risk-averse. You want your traders to be risk-seeking. If you really constrain your traders, you keep an eye on them, and you check all of their trading marks in detail so you understand every single trading position -- if you do that, in the aggregate, your traders aren't going to make as much money. It's an argument about the entrepreneurial spirit. You want to free up your employees to take risk. Inevitably, when you free them up, traders will start to aggressively mark their positions and take risks they maybe shouldn't -- like Kerviel. He had this arbitrage strategy, which is supposed to be neutral, but how do you really monitor neutrality? Well, somebody takes a little risk here, have a slightly mismatched option position that makes some money, they lose some money so they mismark to hide that loss, but then it works out in the end, because they make money on some other trade. If you have 100 traders you might make more money even if one of those 100 rips you off -- the 99 of them will make so much more money than they otherwise would. If you look at the history of the past two decades, there's some truth to that. Bank trading operations are massively profitable. The way traders make money are often dubious, sometimes illegal, and we see these repeated instances of rogue trading.
Does it really matter whether banks have trouble controlling traders because they're too big or because they don't want to?
I don't think so. You can call the first story "we can't" and the second one "we won't". In either case, any regulatory attempt that relies on top-down line-drawing is going to have a very difficult time. Take the Volcker rule. If you tell banks that they can't engage in propriety trading, how will they react? And what is a hedge versus a proprietary trade? If you believe the first story, it might make the risk even more dangerous, because you think you've eliminated proprietary trading, but you've still got this possibility of risk that you can't monitor. But the second story might be true as well -- banks might say that they can't engage in proprietary trading, so they'll "hedge for profit". Banks will empower traders to engage in arbitrage, then wink and nod when things get out of control. If somebody makes a million dollars on a trade that wasn't really arbitrage, maybe the bank will slap them on the wrist and tell them not to do it -- but good job, and we'll pay you a $500,000 bonus!
Should we be thinking about bringing back Glass-Steagall or will banks always find a way to trade? Or should we not worry about it?
I think it's a concern because it goes to the deeper question about trust in banks and the riskiness of banks. Where there's smoke there's fire, and I think it's a mistake to take these incidents as one-offs. They reflect deep risk within the banks. But I do think trading will keep showing up. It has shown up in other places. Even if you brought back Glass-Steagall, there would be pressure to create exceptions, like with Bankers Trust back in the 80s and 90s. It's kind of like a piece of Swiss cheese, where there's pressure for the holes to get bigger and bigger. There's just too much profit in trading for banks to agree to forgo all those profits. If you implemented Glass-Steagall, the commercial banks would want to keep as much of this as they could, and they would do it by labeling it "hedging" or something else.
What can we do to get banks to rein in some of this riskier, or rogue, trading?
Rather than just looking at Value at Risk (VaR) or required risk reporting, we need bank boards to start to ask questions about worst-case scenarios and hold managers' feet to the fire. The board might ask what will happen to all of our positions if home prices decline by 30 percent or equity prices decline by 20 percent or equity volatility goes up, and we're going to hold you responsible if what you tell us turns out not to be the case. What's missing is monitoring within the institution, and creating an incentive to do so.
What do you think it will take to create that incentive for the board? The last crisis is still fresh in our minds, so boards might be more vigilant now, but what about when things return to normal-ish, and there's a risk of institutional forgetting?
The only thing that really works are real ex post consequences. In the aftermath of the Great Crash of 1929, we created this system which was embroidered by judges over time, and ultimately provided a real threat for board members and for people within companies to worry that they might actually be punished -- to think about the future more. Think about it from the trader's perspective. If you're a trader and go rogue, it's like your'e spinning a roulette wheel and hoping it doesn't land on double zero. The probability of being caught and then being prosecuted, and then the prosecutors winning is very, very small. The Kerviel case actually reinforces that -- the difficulty of the case, and all the evidentiary questions, and the length of the case make it so no government will have the capacity to bring very many of these cases.
Is it a matter of expanding the scope for civil lawsuits?
I think it's both civil lawsuits and criminal liability, so that there's more frequency. It's more like other kinds of fraud or criminal activity. If you were to go out and start a meth lab in your house, what's the probability that you would ultimately end up doing jail time? It's going to be a lot higher than if you massively defraud a financial institution. It's an endemic problem within the justice system now. Civil lawsuits are a lot harder than they were, and the resources for prosecutors are a lot less. Prosecutors are bringing these cases and often losing, which doesn't help. It's a pretty simple solution idea-wise, but it's very difficult to implement. And there isn't a lot of political will for it. Notice this hasn't been a topic within this presidential election at all.
We want to hear what you think about this article. Submit a letter to the editor or write to firstname.lastname@example.org.
Matthew O'Brien is a former senior associate editor at The Atlantic.