By James Kwak
One part of the Obama Administration’s financial reform plan is tighter regulation of credit default swaps – those previously unregulated derivatives that brought down AIG and nearly the entire financial sector with it. One of the problems with AIG was that its regulators were apparently unaware that it had amassed a huge, one-sided portfolio of credit default swaps that amounted to a massive bet the economy would do just fine; another problem was that, because credit default swaps were “over the counter,” custom transactions between individual private parties, they created a large amount of counterparty risk – the risk that the party you were trading with might not be there to honor the trade.
In response, the administration proposes to “require clearing of all standardized OTC derivatives through regulated central counterparties (CCPs).” In addition, “regulated financial institutions should be encouraged to make greater use of regulated exchange-traded derivatives.” Major players in the market will also be subject to conservative capital requirements (making sure they have enough money in case their trades go badly) and reporting requirements. These provisions aim to increase regulatory oversight and minimize the chances that a derivatives dealer will fail and take its counterparties down with it, and as far as they go they are a good thing.
However, there is one potential loophole that, according to UCLA law professor Lynn Stout (on Friday’s Morning Edition), is “potentially big enough to put the state of Texas into.” The loophole is that “customized bilateral OTC derivatives transactions” would remain out of the reach of both exchanges and CCPs.
Custom derivatives would still have to be reported to regulators, so what’s the problem? The small problem is that unnecessary customization of financial products is a great way for derivatives dealers to jack up unnecessary transaction fees. The big problem is that custom derivatives are by their nature harder to oversee. Regulators want to be able to estimate a firm’s potential exposure across all its tranactions under various scenarios; the more complex those transactions, the more difficult this becomes. Regulators already had the power to demand access to banks’ books before the financial crisis; the problem was that they lacked the staff and skills to understand the complex structured products those banks were manufacturing and trading. As a result, custom products become a way for market participants to hide risks from oversight, and a potential means for systemic risks to build up out of sight.
The conventional wisdom is that some firms have unique needs and therefore there have to be customized derivatives contracts. But the real question to ask is why we need customized derivatives in the first place. For example, you can only buy U.S. Treasury bills and bonds that mature on specific dates, and in specific denominations (although perhaps there are banks that will custom-manufacture a Treasury-like security for you, and charge you a transaction fee – while throwing in counterparty risk for good measure). What’s wrong with a world where you can buy a credit default swap on any fixed-income instrument, but only for certain maturities (including the maturity of the underlying instrument) and on standard terms (such as the definition of a credit event and how the swap will be settled)?
I know the high-level answer (in fact, I already said it): firms have unique hedging needs. But I want to hear some good examples. On that same Morning Edition segment, Cory Strupp, a lobbyist at the Securities Industry and Financial Markets Association, gave this example: “If you have a company that wants to hedge a credit exposure in an odd amount of money – $156,217.25 – they can enter into a credit default swap that covers exactly that amount of credit risk, to the penny.” This is a terrible example, because if I’m a company of any size, and I have a credit exposure of $156,217.25 but I can only buy credit default swaps in multiples of $10,000, that’s perfectly fine with me. Strupp must know it’s a terrible example, but must have decided the better examples were too complicated for NPR.
And once we know what the good examples are – and I imagine there are some – the question is whether the benefits they provide to the parties involved outweigh their costs. And by costs, I don’t mean just the transaction costs; I mean the fact that customized derivatives make regulation harder and increase the risks of a costly failure. This is an externality, pure and simple, and it should be deterred or taxed.
(I’m guessing someone will point out that the Constitution limits the ability of the government to interfere in the freedom of contract. But remember, this is a regulated industry. Custom derivatives increase the cost of regulation; it would be perfectly constitutional to say that firms that trade in custom derivatives must pay a hefty tax on those products to offset the increased regulatory cost. If the tax is big enough, that would deter banks from using custom derivatives when standardized ones would do.)
There’s actually an interesting point behind this question. Someone – I think it was Felix Salmon, but now I’m not so sure – said that the real problem wasn’t that firms weren’t perfectly hedged; it’s that they believed they could be perfectly hedged. Risk never goes away; if you think you’ve eliminated it, it’s just gone someplace else to hide. Instead of a system where companies think they can hedge their risks perfectly because financial products are infinitely flexible – and therefore don’t manage their risks effectively – it would be better to have a system where companies can’t hedge their risks perfectly, and know that, and behave accordingly.