Financial rule-makers must learn to use their calculators
Believe it or not, U.S. financial regulators don’t have to calculate the economic impact of the rules they write. It’s an omission they should correct before it becomes a serious obstacle to fixing the financial system.
The term “cost-benefit analysis” is gaining prominence as the battle over the 2010 Dodd-Frank financial-reform act moves into the courts. Lawyers working for corporate lobbies have challenged several rules on the grounds that the authors failed to fulfill adequately a legal requirement: Regulators must evaluate or consider — not necessarily quantify — the costs and benefits of new rules.
Courts have already struck down two Dodd-Frank rules, one designed to curb speculation in derivatives markets and another aimed at giving shareholders more say in the selection of corporate directors. The next targets could be higher-profile items, including the forthcoming Volcker rule, which seeks to prohibit short-term speculative trading at federally insured banks.
It’s easy to portray the legal challenges as typical financial-industry tactics. Problem is, they’re effective because they have a valid argument. Regulators’ cost-benefit assessments have largely failed to meet the low standards set by the law. And even if they succeeded, they would still fall short of what common sense demands.
Consider the level of analysis that has so far accompanied a crucial element of financial reform: New capital requirements that would make banks more resilient in times of crisis. In the proposed rule issued by the Federal Reserve and other agencies, the cost estimate amounts to a calculation of how many hours bank employees will spend on compliance paperwork. No effort is made to quantify the benefits. As a result, opponents of higher capital requirements have been able to dominate the conversation with specious warnings that the new rules will harm economic growth.
Financial regulators protest that forecasting the effect of an untested new rule can be difficult, particularly when one must weigh relatively certain costs against hard-to-quantify reductions in the likelihood of crises. Somehow, though, other agencies manage to do it. When, for example, the National Highway Traffic Safety Administration issued new emissions and fuel-efficiency standards last year, it included more than 50 pages with tables quantifying everything from the added cost to manufacture specific automobile brands to the benefit of reduced highway fatalities due to lighter cars.
Agencies that specialize in finance should be able to do at least as good a job as those that deal with health and safety. In a recent paper, two University of Chicago professors — economist Glen Weyl and legal scholar Eric Posner — suggest a place to start: Develop a standard statistical cost of a financial crisis, much like the statistical value of a human life that agencies such as the NHTSA use to estimate the benefits of safety rules. Such a measure would allow regulators, with an educated guess at how much a new rule would reduce the probability of a crisis, to produce a ballpark benefit estimate that could be weighed against a rule’s costs.
In a 2011 analysis of bank capital levels, economists at the Bank of England and the Bank for International Settlements showed how the calculation can be done. Looking back at nearly 200 years of data, they estimated how much each added percentage-point increase in capital ratios would lower the likelihood of systemic crises. They also assessed the effect of increased capital on banks’ funding costs, interest rates and economic growth. Using the Bank of England’s own estimate of the cost of a crisis — 10 percent of economic output, with a quarter of the effect being permanent — they concluded that the optimal ratio of capital to risk-weighted assets would be about 20 percent, or more than double the level required by the latest iteration of international banking rules.
To be sure, regulators won’t always have enough data to produce an empirical estimate of how much a given rule will reduce the odds of a crisis. That said, even a ballpark guess has value. It puts a limit on what regulators can claim and makes them think harder about how different rules might overlap and interact. If, for example, estimates of how much all the rules in Dodd-Frank would lower the probability of a crisis add up to more than 100 percent, that would be a clear signal to go back to the drawing board.
In short, regulators can let cost-benefit analysis stymie their efforts to build a better financial system, or they can use it to their advantage. Need we recommend which course of action to take?