Tougher capital regulation pays off
Banks continue to lobby for weaker financial regulation: capital requirements are excessive, liquidity requirements are overly restrictive, and stress tests are too burdensome. Yes, in the aftermath of the 2007-09 financial crisis, we needed reforms, they say, but Basel III and Dodd-Frank have gone too far.
Unfortunately, these complaints are finding sympathetic ears in a variety of places. U.S. authorities are considering changes that would water down existing standards. In the name of supporting community banks, the Senate has approved a bill to relax the regulation of large banks, in part by raising the asset threshold for a bank considered large. The Treasury wishes to “recalibrate” requirements on the largest U.S. banks, fearing that strict regulation makes them less internationally competitive (see here). And, the Federal Reserve is considering a variety of changes, some of which would reduce the effectiveness of their stress tests (see here).
In Europe, news is not promising either. Just last week, under pressure from Italian authorities (whose banks have spent the better part of the past decade in intensive care with stubbornly high non-performing loans), supervisors at the ECB have weakened requirements on how banks treat bad loans. Once again, national authorities have pushed to reduce the safety of the system.
These moves are not only discouraging, but they are self-defeating. Regular readers of this blog are familiar with our view that capital requirements remain too low. Higher capital clearly improves resilience. And, at current levels of capitalization, it does not limit banks’ ability to support economic activity. Experience suggests the opposite: well-capitalized banks lend more and they lend better, while zombie banks lend to zombie firms (see our previous post).
As it turns out, on this particular subject, there may be less of a discrepancy between private and social interests than is commonly believed. The reason is that investors reward banks in jurisdictions where regulators and supervisors promote social welfare through tougher capital standards.
To see this, we turn to evidence on the ratio of market value to book value of banks—their “price-to-book ratio” (PBR). The following figure plots the (asset-weighted) average PBRs for a sample of large U.S. and euro-area banks. From 2000 to 2016, data are quarterly. The dot at the end is an approximation we have constructed for mid-year 2017. (For a discussion of PBRs, including a study of their determinants, see Bogdanova, Fender and Takáts (2018).)
Bank price-to-book ratios for the euro area and United States (quarterly), 2000-June 2017
Source: Bogdanova, Fender and Takáts (2018), FDIC Global Capital Index, and authors’ calculations. The mid-2017 dots are for G-SIBs only.
We can draw a number of interesting conclusions from these data. In the run-up to the crisis, prior to 2007, PBRs were well in excess of one. Then, when the crisis hit, since banks mark only a fraction of their portfolio to market, the ratios plunged as investors looked through the accounting veil. Until about 2011, the data suggest that banks in the euro area (in black) and the United States (in red) followed a similar pattern. While U.S. banks enjoyed a persistently higher PBR, the gap with the European banks remained about the same size.
Since 2012, however, the gap has averaged half again as large as in 2008, with U.S. banks’ PBRs rising more rapidly. We can think of several explanations for this. The first concerns stress tests. While these started at the height of the crisis with the Supervisory Capital Assessment Program (SCAP) in spring 2009, and the Fed completed the first Comprehensive Capital Assessment Review (CCAR) in March 2011, they did not become truly systematic for several more years. Over time, the U.S. tests have become increasingly effective, partly because bank managers learned that those who fail risk incurring the wrath of their shareholders. Finally, in 2013, the Dodd-Frank Act stress tests (DFAST) complemented what had become a well-established CCAR process.
Meanwhile, in Europe, things proceeded quite differently. While the European Banking Authority (EBA) (or its predecessor, the Committee of European Bank Supervisors) reported the results of stress tests in September 2009, July 2010 and July 2011, few observers really believed them. Famously, Belgian and French authorities had to rescue Dexia in October 2011, a mere three months after it received a clean bill of health. Looking closely at the chart above, we can see that between the first and second half of 2011, the PBR for the largest euro-area banks plunged from 0.75 to 0.45!
It is worth noting that the evolution of SRISK—the NYU Stern School Volatility Lab’s measure of capital shortfall—is consistent with the view that the early version of the EBA stress tests were insufficiently stressful. In their study of these episodes, Archarya, Engle and Pierret suggest that the failure of the euro-area tests to stress sovereign exposures adequately bears some responsibility for subsequent difficulties.
Responding to these strains, and understanding that further financial integration was essential for the success of monetary union, euro-area authorities set out to unify banking supervision. The result is the Single Supervisory Mechanism (SSM), which resides inside of the European Central Bank (ECB). In anticipation of taking over responsibility for supervising the bulk of the euro-area banking system, the ECB and EBA undertook a comprehensive assessment of bank health, including a detailed review of the quality of the assets of 130 major banks (see here.) Completed in October 2014, this review led to a modest improvement in the conditions of the banking system: euro area banks’ PBRs rose to roughly 0.85 by the beginning of 2015.
Europe is now undertaking stress tests on a two-year cycle (with one having started earlier this year). We questioned the rigor of these tests in the past, and remain concerned. As far as we can tell, European authorities continue to lag in ensuring their banks are sufficiently resilient.
The following chart plots a measure of banking system leverage ratios—the unweighted ratio of capital to assets—in the euro area (black) and the United States (red). Specifically, we take the numbers from the FDIC Global Capital Index for the global systemically important banks (G-SIBs) and compute an asset-weighted average of their (IFRS-based) leverage ratios. For the eight largest American banks, this average has risen over the past five years from an abysmally low 3.7% in June 2012 to a less dismal 6.6% on its most recent mid-2017 reading (a range of analyses points to a desirable leverage ratio more than twice as high). By contrast, for the nine largest euro-area banks, the ratio has climbed only from 3.0% to 4.2%. Not only that, but as we noted earlier, authorities remain hesitant to require the weakest banks to realize losses on non-performing loans.
Leverage ratios: largest euro-area and U.S. banks (percent of tangible assets financed with equity, semi-annual), 2012-June 2017
Source: FDIC Global Capital Index and authors’ calculations. Leverage ratios are IFRS asset-weighted averages of the ratio of tangible equity to tangible assets (where equity and assets are adjusted by subtracting goodwill, other intangibles and deferred tax assets).
Throughout this episode, European authorities exhibited substantial reluctance to requiring their banks to recapitalize. Aware that their banks tend to raise leverage ratios by shedding assets (see here), supervisors may worry that, by being tough on banks, they would limit the supply of credit and smother any nascent recovery. So, this line of thinking continues, it is better to wait until production and employment are strong before cracking down on the banks. Well, a decade is a long time to wait. Not only that, but evidence from elsewhere, like the United States in 2009, is that the faster you fix the problem, the better. And, by requiring banks to issue equity rather than shed assets—as the Fed did following its 2009 stress tests—it is possible to minimize the short-run impact on lending.
Our conclusion is simple: effective supervision is a public good. Just like disclosure, it helps overcome information asymmetries, reducing uncertainty and the cost of financing. When investors trust banks’ valuations of their assets, so that they can confirm there is sufficient capital, they reward them with higher equity prices. Similarly, seeking certification of their creditworthiness, high-quality borrowers will gravitate toward well-capitalized banks that have the capacity to lend.
Banking industry lobbying aimed at diluting regulation and weakening stress tests, complaining about narrow definitions of capital, expansive classification of assets, and rigorous treatment of risk, is shortsighted. At current levels of capitalization, more capital and effective supervision are good both for the financial system (making it more resilient to shocks and less prone to collapse) and for the banks themselves, raising their market value.
Acknowledgment: We thank Bilyana Bogdanova, Ingo Fender and Előd Takáts for sharing their data.