Gerard Caprio, Jr.
Senior Fellow
Banking and Finance and Public Policy
Dr. Gerard Caprio, Jr., is the William Brough professor of economics at Williams College and chair of the Center for Development Economics there. He has taught at Trinity College, Dublin (as a Fulbright scholar), and visited at George Washington University.
read bio
Moutusi Sau
Senior Associate, Program Research Analyst, Center for Financial Markets
Moutusi Sau is a senior associate at the Milken Institute's Center for Financial Markets in Washington, D.C. She conducts research and helps execute programmatic activities on Milken Institute projects involving new tools for access to capital, strengthening capital markets in developing countries, and housing finance reform.
read bio

European banks' stress test: A three-part series

By: Gerard Caprio, Jr. and Moutusi Sau
November 04, 2014

What stress tests really stress: Supervise the supervisors!

By Gerard Caprio, Jr.

Last week, the European Central Bank, which this year took over as supervisor of the 150 largest euro area banks, published the results of its first stress-testing exercise, a war game of sorts to test the strength of banks’ balance sheets. Headlines stress the fact that (only) 25 banks were found deficient of capital, and 13 have yet to raise the additional capital mandated by the tests. However, the real news reported by the Financial Times (“Asset Quality Review: Check the Fine Print,” October 26) was found in the small print: 80 percent of the banks tested were underprovisioned for loan losses.

When banks make loans, they know that even in good times, some, presumably small, percentage of loans go bad, and the loans end up producing losses for banks. Even for loans that are not written off, while the borrowers continue to make some payments, banks set aside provisions for the losses they reasonably expect to see. Bank capital is there to cover unexpected losses; provisions are there to absorb expected losses. Banks that do not provision fully to cover the expected loss in the value of loans are lying about their capital because the losses not covered by provisions will be deducted from their capital. And while banks are responsible to their shareholders to show an accurate balance sheet, the banking supervisors’ job is to police the banks.

So, in revealing that so many banks were inadequately provisioned, what the ECB really revealed is that supervisors have not been doing their jobs. Since the ECB is just this year taking over the role of supervisor of larger banks from national agencies, it was safer for it to reveal this shortcoming sooner than later.

How could it be that supervisors were not doing their job? Actually, underenforcement of regulations is the rule in banking, not the exception. As Jim Barth, Ross Levine, and I point out in “Misdiagnosis: Incomplete Cures of Financial Regulatory Failures” (, in the run-up to the global financial crisis, regulators around the world failed to diagnose what went wrong, did not design reforms to address these defects, and did not implement corrective reforms. Moreover, since that paper was written, research by the nonprofit organization ProPublica, which blew the whistle loudly on many practices at the heart of the crisis, published a report ( broadcast in the NPR series “This American Life” ( saying that well after the crisis, the New York Fed systematically failed to enforce its own regulations on Goldman Sachs.

And in October, the Federal Reserve Board’s own inspector general ( published a report criticizing the N.Y. Fed for its failings in the 2013 “London Whale” episode at JPMorgan Chase. Some wondered at the time whether the supervisors had not been aware of the massive risk that Morgan was taking on, or if they knew but failed to act. Now we know: It was the latter.

So welcome aboard, ECB, to your new job as a bank supervisor. We don’t yet know how you will perform. But what is known is that supervisors get captured by the banks they oversee. For decades, economists have discussed the “revolving door” form of capture—regulators being lenient in enforcement in hopes of a future high-paying job. Although this is plausible in the world of finance, we know many supervisors who do the best they can and we suspect that a far more common form of capture is the psychological bias described in Guardians of Finance: Making Regulators Work for Us( Just as sports’ umpires and referees tend to rule in favor of home teams because they are biased by the fans who turn out for the game—like everyone, they want to be liked—regulators are influenced by the banks, who not only represent the players but also the fans in the box seats, whose views the regulators hear loud and clear. The public, sitting in the “nosebleed” section, can’t even see what is happening and have little influence with regulators. In short, regulators work so closely with banks that they are influenced by them, and their oversight is often ineffective. We apparently expect the guardians to guard themselves.

It is not surprising that regulators’ enforcement of the rules favors banks, rather than protecting the public interest, but it is surprising that voters keep tolerating it. In most sports, some form of instant replay or “electronic eye” has reduced the home-field advantage, suggesting that, as proposed in the Guardians volume, some form of oversight of the regulators themselves is needed, such as by disclosing more information to the public and disclosing all information they gather to a watchdog group of experts.

What the recent stress tests stress is that supervisors keep failing, and the public keeps failing to hold them accountable. If we don’t supervise the supervisors, what else can we expect?

Will the stress tests restore faith?

By Moutusi Sau

The European Central Bank (ECB) and the European Banking Authority (EBA) recently published the results of a yearlong review of finances of the top 130 banks in the European Union. Among the 25 that were earmarked as weak during the review process, 13 failed the stress test. The remaining dozen banks had managed to raise capital or make other arrangements to strengthen their positions by the time the final results were published. The ECB found there was an overall capital shortfall of €25 billion within the 25 earmarked banks. The results revealed that the biggest shortfall was in Italy, with four state-listed banks in trouble, followed by Greece and Slovenia, with two banks each. The hope is that the test results will bolster lending to corporations that have moved away from traditional borrowing toward bond issuance.

The stress tests conducted in the European Union involved many bodies, including the European Systemic Risk Board, the ECB, and the European Commission. The EBA coordinated the EU-wide stress tests this time around. Overall, there have been four rounds of EU-wide stress testing in 2009, 2010, 2011, and most recently 2014. In 2009 and 2010, the Committee of European Banking Supervisors (CEBS), predecessor to the EBA, coordinated the tests. The EBA coordinated the latter two.

There have been many comparisons between the stress tests conducted by the U.S. Federal Reserve and its European counterparts. Many have said that the tests performed by the Fed have been stricter than those by the EBA. The Fed conducted its first stress tests in 2009 and failed 10 of 13 banks that year. In contrast, the CEBS conducted its first tests that same year but did not fail any European banks.

The latest tests conducted by the EBA have been more extensive than earlier tests, covering 123 banks overall. A fundamental difference in test design makes the American approach more robust. Under the American design, banks cannot meet the necessary capital ratios by reducing lending. In contrast, under the European design, European banks reduced lending in anticipation of the tests to increase their capital ratios as evaluated by the EBA.

A bank stress test is primarily an examination of top banks’ balance sheets using hypothetical situations to check how well the bank can manage amid economic stress. Performance is measured under both baseline and adverse scenarios, which casts some light on potential weaknesses. The need for these tests was highlighted in the wake of the worldwide contagion effect observed in 2008. The most important ratio used in testing was common equity Tier 1 capital (CET1). According to a helpfulinfographic by the EBA, the capital ratio, or CET1 ratio, measures funds readily available to the bank, expressed as a percentage of its general exposure. The capital ratios measure a bank’s ability to finance operations with its own capital rather than relying on short-term borrowed funds. A bank fails the test when its capital ratio falls below a threshold, set this year at 5.5 percent.

Overall, the tests revealed a relatively stable position for the European banks. Previously, the stress tests have run into trouble, dubbed too easy by the capital markets. This year, however, they seem to have struck the right balance, with the results going a long way to restoring confidence.


Sources: Thomson Reuters, ECB, Milken Institute.

Bullets for Chart of the Week

  • Of the 25 banks earmarked as weak during the yearlong review process, 13 failed the stress test with an aggregate capital shortage of €25 billion.
  • Among the 13 banks that fared worst, four were based in Italy.
  • Italy’s Monte dei Paschi di Siena, the oldest bank in the world, was found to have the largest capital shortfall at €2.11 billion.


Europe: Passing a stress test does not guarantee health

By Keith Savard

The headline takeaway from the European Central Bank’s comprehensive assessment of 130 European banks is that the vast majority are solid enough to survive a crisis. Indeed, according to the 170-page report released last week, only 25 of the 130 banks examined had a shortfall in capital, totaling €24.6 billion at the end of 2013. Taking into account capital raised so far this year, the present shortfall amounts to just €9.5 billion. The banks included in this exercise raised more than €200 billion between 2008 and 2013, as well as an additional €57 billion in the first nine months of 2014.

With Europe’s banks now deemed to have healthier balance sheets, it is hoped they will lend more to companies and households. Since the financial crisis in 2008, bank lending to companies has fallen by nearly €555 billion, or 11 percent, according to ECB data. This decline has been partly offset by companies’ bond issuance, which totaled more than €432 billion in the last six years.

However, while there could be some marginal pickup in lending, the likelihood of it gathering momentum seems slim given that about half of the 130 banks tested still have thin capital buffers, according to banking analysts. The weak position of some of these banks could lead to consolidation in the industry. Moreover, demand for loans is weak, despite signs of easing in credit conditions. With the economic outlook for the euro zone deteriorating, including an uptick in unemployment, we suspect that even if banks are in a position to lend, companies and households will remain reluctant to borrow.

The ECB generally has been given high marks by banking analysts and economists for the quality and thoroughness of its nearly yearlong examination of the banks’ resilience and positions. This follows the previous shoddy exercise conducted by national authorities in 2011. Nevertheless, there are a few points worth mentioning to encourage further improvement in future asset quality and stress-test exercises.

While agreement was reached on a common definition for nonperforming exposures (NPEs) across countries (contributing to a €136 billion increase in NPEs to €879 billion), the same due diligence was not applied to the definition of capital. In particular, many countries still allow their banks to use accounting gimmicks to make their capital position more robust. Greece, Italy, Portugal, and Spain, among others, allow their banks to count future tax refunds as capital. The ECB in its report expressed confidence that this and other shortcomings will be corrected soon.

Perhaps the area most in need of further improvement involves the assumptions underlying the adverse scenario portion of the stress test. With a growing list of countries in the euro zone already experiencing deflation, the document mentions deflation only one time. According to the ECB, “While the adverse scenario does not strictly embody a prolonged deflationary environment, it does entail material downward pressures on inflation.” Nonetheless, in the adverse scenario, yearly inflation for the euro zone in 2015 averages 0.6 percent and in 2016 it averages 0.3 percent.

While the ECB has been pressured this year to engage in full-fledged quantitative easing, it seems a bit surprising that the central bank did not include outright deflation in its adverse scenario. Although the ECB leadership always has maintained that deflation is unlikely to engulf the region, problems with the credibility of the previous stress test, combined with the overall fragility of these economies, would suggest erring on the side of caution.

In any event, the ECB should be congratulated for its efforts, which hopefully will help to underpin the basic aspects of lending in the banking system as well as encourage the restructuring of Europe’s fragmented banks. Bank equity prices so far have not responded well to the release of the ECB report, but it is premature in our view to underestimate the impact this report will have on rebuilding confidence in the European banking system.


No one has commented on this page yet.

Post your comment