(c) 2013, Bloomberg News.
(c) 2013, Bloomberg News.
The following editorial appears on Bloomberg View:
It's been almost a year since we caused a stir by pointing out that the largest U.S banks received a taxpayer subsidy worth an estimated $83 billion a year. What's changed since then?
Better banking rules are coming into force, and if they work they'll reduce the subsidy. It would be good to know exactly how much. Unfortunately, measuring the transfer in real time is something regulators aren't well equipped to do.
Policymakers and legislators have largely come to accept that the subsidy is there. The biggest and most systemically important banks can borrow more cheaply than they otherwise would, because creditors expect the government to rescue them in an emergency. This is an unfair and unintended transfer of wealth to bank shareholders and executives, and it weakens market discipline by desensitizing banks to risk. In effect, banks are being rewarded for presenting a threat to the economy.
Federal Reserve Chairman Ben Bernanke and his likely successor, Janet Yellen, have both acknowledged the too-big-to- fail subsidy as an issue that needs to be addressed. Regulators are putting in place new capital requirements to ensure that the biggest banks are less likely to fail; the Federal Deposit Insurance Corp. is working on ways to handle those that do. The soon-to-be-published Volcker rule is explicitly intended to cut the subsidy for speculative trading.
Banks say these reforms are enough. Others think there's a long way to go. Research and experience strongly suggest that the economy would benefit from capital levels much higher than what regulators have proposed. On the subsidy, the evidence is mixed: One recent study finds that the Dodd-Frank Act may have had some effect, another says it hasn't. Settling this issue will require better measurement.
The subsidy has always been difficult to estimate. To know how much banks are benefiting from investors' expectation of government bailouts, you need to know what their borrowing would cost without it. This cannot be observed, so researchers use different tricks to tease it out.
The economists whose work we cited in our $83 billion estimate — Kenichi Ueda of the International Monetary Fund and Beatrice Weder di Mauro of the University of Mainz — focused on the credit-rating "uplift" that banks receive due to the expectation of government support. By looking at long-term data on borrowing costs at different ratings, they could infer the typical annual value of the taxpayer subsidy.
The rating method, though, is less useful in monitoring the effect of reforms. Studies have shown that rating companies can succumb to pressure from banks, which are among their biggest customers. In the immediate post-crisis year of 2009, on which Ueda and Weder di Mauro focused, the raters were bound to recognize the value of government support. In calmer times, maybe not.
Moody's Investors Service recently removed all uplift related to government support from its ratings of the largest U.S. banks. Its reasoning was that the FDIC's new resolution mechanism means there'll be no more bailouts — even though the mechanism isn't yet complete and the FDIC's vice chairman, Thomas Hoenig, says it couldn't handle a crisis like the one in 2008.
Researchers from New York University, Virginia Tech University and Syracuse University have adopted what seems a better way to measure progress. They use small banks, which don't enjoy too-big-to-fail status, as the borrowing-cost benchmark (with statistical controls to strip out other features that make big and small banks different). This approach suggests that a multibillion-dollar subsidy persisted at least through 2011.
Economists are working on other methods, such as building a mathematical model of a subsidy-free big bank. No one approach is likely to be perfect. The best solution, suggested by Minneapolis Fed President Narayana Kocherlakota in a recent speech, is to develop several measures and use them all — just as Fed officials do, for instance, when judging the state of the labor market.
Government can help move this along. At the behest of Congress, the Government Accountability Office is analyzing the too-big-to-fail subsidy, work that might be used to develop real-time indicators. The Office of Financial Research, set up by the Dodd-Frank Act to give regulators more information, could act as a clearinghouse, vetting the available research.
In the end, though, the estimated size of the subsidy, which can vary with market conditions and the general level of interest rates, is less important than its presence. As long as it exists, banks will want to take on too much risk. Under international accounting standards, the six largest U.S. banks have amassed more than $14 trillion in assets, equivalent to nearly a full year of economic output. Only about 4 percent of that is financed with loss-absorbing equity. A drop in asset value of only about 4 percent could be enough to make them insolvent.
It would be good to know that financial reforms have eliminated the subsidy. Without better evidence, that call would be a leap of faith.