April 18, 2016
A Summary of the First Policy Symposium Hosted by the Federal Reserve Bank of Minneapolis April 4, 2016
In its first policy symposium on ending the problem of financial institutions that are too big to fail (TBTF), the Federal Reserve Bank of Minneapolis heard proposals to require large banks to hold much more capital, to limit banks’ size and to consider the benefits and costs when developing new regulations.
The Minneapolis Fed kicked off its Ending Too Big to Fail initiative in February with a speech in which Bank President Neel Kashkari said that some banks are still too big to fail and that their failure would cause unacceptable damage to the American people and the economy.1 Kashkari noted that although progress has been made to address this problem, more is needed, and he announced the Bank’s initiative to present a bold and transformational solution to address TBTF once and for all.
As part of this initiative, the Minneapolis Fed is holding a series of policy symposiums to bring together experts with a wide range of views to discuss specific proposals. The symposiums are being streamed live on the web to allow the public to learn alongside the Minneapolis Fed. Following up on the first symposium described in this policy brief, the Bank will host another symposium May 16 and more events throughout the year. The Bank will use these events, as well as public submissions through its website, to develop its plan to end TBTF. The Minneapolis Fed has committed to delivering that plan by year-end for legislators, policymakers and the public to consider.
The first symposium focused on two specific transformational proposals for ending TBTF—higher capital requirements and limits on bank size—and featured a keynote address by a former Federal Reserve governor, who discussed the appropriate role of weighing both costs and benefits in financial regulation. (Other approaches will be considered in future events.)
The Minneapolis Fed encourages the public not only to observe this process but to provide feedback. The Bank has established a website where anyone can share their ideas for solving TBTF.2
Panel 1: Substantially increasing capital requirements
In the first panel, Anat Admati3 proposed that banks substantially increase their equity to as much as 30 percent, a level comparable to nonbanks. More equity means banks can absorb greater losses on their assets before they become insolvent. Here are her main points:
- The failure of systemically important banks generates significant harm to society that banks do not consider when they decide to grow or take on risks. Higher capital requirements would reduce risk of harm.
- Large banks are very risky institutions—they are opaque and complex, making orderly resolution difficult. Passing a stress test is no guarantee of even medium-term solvency. Society is better off trying to prevent their failure than trying to make their failure “safe.”
- Even under Dodd-Frank Act regulation, capital ratios continue to be tiny, on the order of 4 percent. Moreover, these capital figures are misleading because they rely a great deal on complex accounting done by the banks. Leading up to the failure of Lehman Brothers in 2008, banks that later failed had about the same, if not slightly higher, capital ratios than banks that did not fail.
- Moreover, regulatory capital requirements have been shown in many cases to be poorly designed. For example, sovereign debt held by European banks was considered risk-free (and required no capital be held against it), but turned out to be quite risky.
- Banks argue that they cannot operate with higher capital requirements, because it would be too costly and would reduce their ability to provide loans. But compared with nonbank firms, banks have much less capital and are much more likely to make regular payouts to shareholders. Both have risky, long-term, illiquid assets and can use retained earnings (or new shares) to grow. But nonbanks typically have at least 30 percent equity funding (and often more) as a share of assets and often go long periods of time without making payouts to shareholders. Banks, however, rarely have as much as 6 percent of assets funded by equity and typically make payouts to shareholders unless they fail a stress test.
- Current plans to make banks more able to absorb losses would allow banks to issue debt that converts to equity and count that debt as a buffer to absorb losses. That debt approach is largely untested, but experience with it so far suggests that it will not work. Issuing more equity would be more straightforward and effective.
The views of the panel4 ranged from support for higher capital requirements to opposition:
- Supporters highlighted analysis suggesting that the benefits of increasing capital (to up to two times the current requirements) exceed the costs. They also pointed out the simplicity of higher capital requirements compared with the current system, and argued that the current approach is too complex and will not work in practice.
- Other panelists noted that it was difficult to determine how much capital banks should hold. They argued that measurement of the implied government support that firms might get is needed to set capital requirements, tax rates or premiums to encourage banks to curtail risks.
- Still other panelists argued that current approaches will work. They noted that current capital stress tests are precisely about ensuring that banks do not fail in a bad state of the world. Banks have a lot more capital now than in 2008, and the stress tests have gotten harder since the first one in 2009. They also argued that the costs of capital were real and that higher capital would lead to lower economic activity.
Panel 2: Altering the organizational structure of financial institutions
In the second panel, Simon Johnson5 proposed that the size of banks, as measured by “total exposure” reported to the Federal Reserve System, be capped at 2 percent of GDP, or about $350 billion. For institutions with exposures above this threshold, he proposed imposing stringent capital requirements. He defended his 2 percent threshold by noting that:
- Lehman Brothers had assets in excess of $600 billion when it failed in September 2008, and Bear Stearns had assets of nearly $400 billion when it was saved. In contrast, smaller firms, such as CIT Group (about $120 billion in total exposures) were allowed to fail without causing damage to the broader financial system. If all banks were small enough, they could fail without the need for bailouts or large costs for society.
- The size cap would impact the following 10 banks, listed here with total exposure in 2014: JP Morgan Chase ($3.7 trillion); Bank of America ($2.8 trillion); Citigroup ($2.8 trillion); Wells Fargo ($2.2 trillion); Goldman Sachs ($1.5 trillion); Morgan Stanley ($1.3 trillion); US Bancorp ($539 billion); PNC ($460 billion); Bank of New York Mellon ($418 billion); and HSBC North America ($417 billion). The framework for resolution of large, complex financial firms is clearer now than in 2008, but there are still substantial residual risks associated with resolving these firms, especially regarding their global footprints.
- Measured in terms of shareholder equity divided by total exposure, bank capital is still quite low even after the post-crisis regulatory changes. Firms could still fail in the event of a large adverse shock.
- Some risk-taking by large banks has been limited by the Volcker rule and other provisions of the Dodd-Frank Act, but the performance of these provisions under stress remains untested.
- These regulations could be implemented under existing legislation.
- Stringent capital requirements for banks above the threshold would lead boards of directors and management to reorganize and break up their firms.
- Large banks are a relatively recent phenomenon, and there is no evidence that their rapid growth over the past two decades has enhanced growth of the U.S. economy. However, the size and willingness of large financial institutions to take on bigger risks did contribute to the severity of the crisis in 2008.
Among the issues raised by the panelists6 and the audience were:
- Large nonfinancial firms, especially multinationals, have banking needs that may be best served by large banks.
- After controlling for risk-taking, analysis suggests significant scale economies in banking, especially for the largest banks. A 2 percent size limit would forfeit these economies.
- A size cap would mean millions of Americans would need to change their banks. Retail customers seem to prefer big banks.
- Breaking up banks into smaller pieces would be difficult and could have unintended consequences, including loss of access for less-profitable customers.
- Banks are prohibited from having a single large equity stakeholder who would have the incentive to provide adequate monitoring. Eliminating this distortion is important, but a size cap would not address it.
- Past financial crises, such as the Great Depression, were not related to the size of banks.
- Part of the reason some banks got so big is that they acquired troubled smaller banks, including during the last financial crisis.
Keynote Lunch: The need to consider both costs and benefits while assessing bank regulations and in addressing sources of financial fragility
Former Federal Reserve Governor Randall Kroszner7 delivered the keynote address over lunch. He discussed the importance of conducting careful examinations that consider both costs and benefits of various bank regulations, while warning against allowing “analysis paralysis” to prevent regulators from acting. Key points:
- Good policy regulation involves several steps: diagnosing the problem via theoretical and empirical analysis; considering both costs and benefits before choosing an action from a set of reasonable alternatives; acting; reviewing the effects after a period of time, say five years, while continuing to refine the options.
- As part of the review process, questions to ask include: Did we achieve our goals? Have there been any unintended consequences? Again, data are needed to answer these questions. The Office of Financial Research is working to collect new data with this in mind.
- Financial institutions exhibit three fragilities: leverage, liquidity and interconnectedness.
- Elected officials must provide guidance regarding how safe a financial system we want. There is a trade-off between safety and growth.
- We do not want to rely too much on any one particular form of regulation.
- Use historical and international comparisons to inform comparisons of costs and benefits.
- Kroszner also agreed that debt that converts to equity may not work in practice when a crisis develops.
The first symposium on ending TBTF brought both transformational options for ending TBTF and a very wide range of views on those options to a national debate. The second symposium on May 16, 2016, will do the same. Both will greatly inform the Minneapolis Fed’s final proposal for ending TBTF.