Mark Gongloff has found a new IMF working paper (pdf), by Arnoud Boot and Lev Ratnovski, which basically comes to the conclusion that banks shouldn’t be trading in financial markets. This is a conclusion others have come to as well, of course — most prominently, the Volcker Rule in the US and the Vickers Report in the UK both attempt to legislate such things, on the grounds that it’s simply not just for banks to engage in risky trading activities, safe in the knowledge that if they blow up they’ll end up getting bailed out by the government.
The IMF paper, by contrast, takes a different approach. It just looks at the way that banks work, under a few simplified assumptions. Let’s say, for instance, that banking — lending money to steady customers — is a profitable business to be in. And let’s say that a large part of what banks do is to offer credit lines. At any given point in time, the bank will have a large number of undrawn credit lines outstanding. And so it’s easy to see how the bank would be tempted to take that money, before it’s drawn down by the customer, and use it to make some short-term profits in the markets. If your trading book is closed out every day, then as soon as your customer asks for the money, you can provide it. And in the meantime, you’ve made a bit more money.
Even if trading is less profitable than banking, then, it still makes sense to trade — as a use for surplus cash which might be waiting to get put to good use.
On top of that, trading is a way of giving excess risk-adjusted returns to shareholders. Let’s stay in the world where banking is steadily profitable — if you hold on to your loans to maturity, and develop healthy relationships with a diversified group of customers, then you’ll end up with a valuable long-term franchise, where shareholders take a certain amount of risk and get a commensurate return. Meanwhile, the bank itself borrows cheaply in the wholesale market, just because it’s so safe.
Once again, at the margin, it makes sense to put a little money into trading. The bank’s cost of funds is low, for one thing. And for another thing, the risks of trading are asymmetrical. If it works, the shareholders make money. But if it fails, the losses can be so huge that the whole institution blows up — which means that not only do shareholders lose money, but so do bondholders. This, weirdly, is a good thing from the shareholders’ point of view, because it means they don’t need to bear the full cost of trading blowups. And just about any bet, if you don’t have to bear the full potential downside, thereby becomes much more attractive.
So banks therefore have two big reasons to move into trading: the paper calls the first one “time inconsistency”, and the second one “risk shifting”. The problem is that both of them make only a limited amount of sense at the margin, in small doses. In large doses, the benefit goes away — as we can see by the fact that trader-heavy investment banks nearly always trade on the stock market at very low multiples of earnings or of book value. (Even mighty Goldman Sachs (GS), these days, is trading at less than its book value.) But the problem is the inexorable logic of marginal thinking: wherever a bank might be, a tiny bit more trading is perceived to be a good thing rather than a bad thing. And so the amount of money the bank trades just goes ever upwards, long past the point at which it’s actually a good idea.
The IMF paper does a good job of listing the consequences. UBS. Barings. Citi (C). Bear Stearns. Lehman Brothers. Merrill Lynch. Washington Mutual. Wachovia. Even JP Morgan (JPM), with its infamous Whale trade. Not all of them were commercial banks, but all of them got far too exposed to tradable market instruments, and suffered enormous damage as a result.
The problem here, as I see it, is that it’s pretty much impossible for banks not to be exposed to tradable market instruments. Call it the curse of the credit default swap: ever since JP Morgan invented the synthetic CDO, bank risk managers have been able to mark their credit portfolios to market, and to hedge those portfolios using derivatives.
I should take this opportunity to plug, not for the first time, Nick Dunbar’s wonderful book about the financial crisis, The Devil’s Derivatives. It is very rich on many levels, but one of the things that Dunbar does in his book is give the best explanation I’ve ever read for why it’s incredibly dangerous when banks start marking their portfolios to market.
Once again, the logic is invidious: if no bank marks its credit portfolio to market, and then one bank comes along and starts doing exactly that, the one bank which is marking to market is likely to have a significant advantage over everybody else. And so one bank moves to a mark-to-market system, and then the other banks have to follow suit to remain competitive, and the result is that everybody ends up in a state where they’d all be better off if none of them did it.
According to Gongloff, the IMF paper says that banks “should be allowed to hedge their bets” with “small trading positions”. But hedging is trading — as we saw, most clearly, at JP Morgan’s Chief Investment Office. And trading is just as dangerous when it’s done for hedging purposes as it is when it’s done for absolute-return purposes.
In an ideal world, then, banks simply wouldn’t be allowed to trade at all. What’s more, in that world the banks would quite possibly make more money than they’re making right now. But you’d need globally-coordinated regulation to get there, and it’s simply not going to happen. Which is why trading blow-ups are here to stay — and regulators are always going to be on the back foot when it comes to trying to prevent them.