The following op-ed by Professor Mark Roe, “Derivatives Clearinghouses are No Magic Bullet,” appeared in the May 6, 2010, edition of the Wall Street Journal. Roe looks at the Senate financial overhaul bill, part of which is built around an emerging Washington consensus that a clearinghouse for derivatives could stem a financial crisis such as that which we just experienced.  Roe argues that while a clearinghouse can be a useful step in the right direction, it’s incomplete, with other legal improvements needed to make it work well.

Derivatives Clearinghouses are No Magic Bullet

As the Senate finalizes its financial reform legislation, a consensus is developing that if we could just get derivatives traded through a centralized clearinghouse we could avoid a financial crisis like the one we just went through.

This is false. Clearinghouses provide efficiencies in transparency and trading, but they are no cure-all. They can even exacerbate problems in a financial crisis.

If I agree to sell you a product next month through a clearinghouse, I’ll deliver the product to the clearinghouse and you’ll deliver the cash to the clearinghouse on the due date. Let’s say we both have many trades going through the clearinghouse and we’ve posted collateral to cover any single trade that fails. This is more efficient than each of us posting collateral privately for each trade. Moreover, we’re not worried that I won’t deliver or you won’t pay because we both count on the clearinghouse to deliver and pay up if one of us doesn’t.

This clearing system makes trading more efficient. If you default, the cost is spread through the clearinghouse so I don’t get hurt severely. And if the clearinghouse has enough collateral from you, there’s no loss to spread. But there’s also a potential downside: The clearinghouse reduces our incentives to worry about counterparty risk. Your business might collapse before you need to pay up, but that’s not my problem because the clearinghouse pays me anyway. The clearinghouse weakens private market discipline.

Still, if the clearinghouse is as good or better at checking up on your creditworthiness as I am, all will be well. But one has to wonder how good a clearinghouse will be, or can be.

Consider two of our biggest derivatives-related failures-Long-Term Capital Management in 1998 and the subprime market in 2008. When Russia’s ruble dropped unexpectedly, LTCM was exposed on its more than $1 trillion in interest-rate and foreign-exchange derivatives. It could not pay up and collapsed. Ten years later the market rapidly revalued subprime mortgage securities, rendering several institutions insolvent. AIG, for example, was over-exposed in credit default swaps tied to the value of subprime mortgages.

Could a clearinghouse really have been ahead of the curve in getting sufficient capital posted before these problems became serious and well-known? I’m not so sure. Worse yet, major types of derivatives have built-in discontinuities-“jump-to-default” in derivatives-speak.

For a credit default swap, one counterparty guarantees the debt of another company to you, in return for you paying a fee for that guarantee. If no one goes bankrupt, the counterparty just collects the fees from you. But if the guarantee is called because the company you were worried about goes bankrupt, the counterparty must all of a sudden pay out a huge amount immediately. Yet the guarantor is often called upon to pay in a weak economy, just when it can itself be too weak to pay. You get credit default protection on your real-estate investments from me, just in case the economy turns sour. But just when you need me the most, in a sour economy, I turn out to be so over-extended I can’t pay up. Collateralizing and monitoring such discontinuous obligations will not be so easy for the clearinghouse.

Moreover, if trillions of dollars of derivatives trading goes through a clearinghouse, we will have created another institution that’s too big to fail. Regulators worried that an interconnected Bear or AIG could drag down the economy. Imagine what an interconnected clearinghouse’s failure could do.

AIG needed $85 billion in government cash to avoid defaulting on its debts, including its derivatives obligations. Could one clearinghouse meet even a fraction of that call without backup from the U.S.? True, we could have many clearinghouses, each not too big to fail-but then maybe each would be too small to do enough good.

The Senate bill would allow a clearinghouse to grab new collateral out from failing derivatives-trading banks to cover old, but suddenly toxic, debts the banks owe to the clearinghouse. This could harm other creditors and cause the firm to suffer a run. Nevertheless, to protect itself in a declining market, a clearinghouse would have to make those big collateral calls. That’s good if it protects the clearinghouse. But it’s bad if it starts a run on a weakened but important bank.

One key but missing element in the search for reform hasn’t yet obtained traction in Washington. Derivatives players obtained exceptions from typical bankruptcy and bank resolution rules in the past few decades for their contracts with a bankrupt counterparty. This allowed them to grab and keep collateral other creditors cannot. That gives derivatives traders reason to pay less attention to their counterparties’ riskiness and weakens market discipline. These rules should be changed before the Senate is done.

To say that a clearinghouse solution is very incomplete is not to say there is an easy solution out there. We may be unable to do more than to make incomplete improvements and muddle through.

Derivatives trades first of all should not just be centrally cleared, but should also be taken out from the government-guaranteed entities, such as commercial banks (or at least we need to impose tight capital requirements on those banks that deal in derivatives). Derivatives traders like doing business with Citibank, because they know the government won’t let Citibank go down. But this puts taxpayers at risk. It would be better to run those trades through an affiliate, not through the bank, so counterparties realize they might not be bailed out if the affiliate failed.

If a banking affiliate’s counterparty is the clearinghouse, then the clearinghouse will have incentives to make sure that the affiliate is well-capitalized. This is particularly so if the clearinghouse won’t get any special priority treatment in a bankruptcy.

Critics of proposals to establish separate bank affiliates for derivatives trading complain about the large amount of capital that would be needed for such affiliates. But the capital that might be needed to buttress a bank affiliate indicates some level of the value (i.e., the taxpayer subsidy) to derivatives players of trading with a too-big-to-fail entity that they know the government will step in to save. They are implicitly getting insurance and should pay for it.

And, since a clearinghouse is itself at risk of being too big to fail, regulators need to police its capital and collateral requirements. If the derivatives market sees the clearinghouse as too big to fail, the potential for derivatives players making overly risky derivatives trades becomes real. Clearinghouses can help manage some systemic risk if they’re run right. If not, they can become the Fannie and Freddies of the next financial meltdown.

Mr. Roe is a professor at Harvard Law School, where he teaches bankruptcy and corporate law.