The EU Stress Tests and the Experience in Spainby Angel Ubide | July 29th, 2010 | 09:00 am
In the days following publication by the European Union of the long-awaited stress tests of its banking sector, the market reaction—judged by the evolution of the euro, bank stocks, and peripheral spreads—has looked positive. This makes sense. The hundreds of pages of analysis of the impact of the stress tests have focused on individual banks, criticized the scenarios used in the stress test as overly harsh or not harsh enough, and as more or less adequate or stressful.
Certainly the stress tests, while reasonably well done, could have been done—and especially communicated—better. But the critical issue that the stress tests were supposed to address was a macro question: Is the European Union, Spain in particular, in danger of a vicious circle that could trigger a banking crisis that could lead to a sovereign default? Beyond the travails of Greece, heavy doubts hovering over markets attest to this question. And the answer from the tests is a strong no. Therefore, the doubts in some quarters about the stability of the euro should be greatly reduced after this exercise.
The first important piece of the news from the exercise is the sharp increase in transparency. We now know more than we knew before, and that’s positive. My colleague Jacob Funk Kirkegaard has done an excellent job of uncovering some of the conclusions hiding in the massive amounts of data disclosed by the European authorities after the test results were unveiled on July 23. Two important messages come across: First, banks were already very well capitalized going into the stress test, as the levels of Tier 1 capital had increased significantly over 2009. Second, even if Greece were to default—and thus banks had to take losses in their banking books—the European banking system (outside Greece, certainly) would be sound. It is important to realize that even if the headline number of “failures” and capital needs looks low, the underlying losses from the exercise are not—over half a trillion euros. It just happens that banks had a large capital cushion to start with.
To this assessment at the European level one must add the superb job that the Spanish supervisors delivered with their own stress test. We should not forget that it was the Spanish leadership, pushing for the full disclosure of the results at the individual level against German resistance that produced this event. The exercise carried out in Spain was on an order of magnitude superior to the rest of the European Union—see here, especially the excellent PowerPoint presentation summarizing the many differences between the Spanish tests and the rest of the European Union. These differences are indeed striking: Spanish authorities tested more than 90 percent of the banking sector, compared to barely 50 percent in many countries. They adopted much more stressful scenarios, where projected losses were greater than those suffered in comparable financial crises, such as Finland, Sweden, or Norway. In fact the Spanish tests show that, back-testing the results using 2007 as a starting point, the stressed probabilities of default are higher than those observed in 2008–09. And the result of the exercise is likely to culminate in a process of consolidation and recapitalization of the banking sector. In other words, after the process of mergers of the savings banks and the capital injections from the FROB (the fund for the orderly recapitalization of the banking sector, which has already injected over €10 billion into the savings banks and is authorized to €99 billion), the Spanish banking sector is ready to withstand a very stressful scenario. The recent changes in the Spanish laws governing the savings banks, to allow them to become banks or quasi-banks and thus be able to raise funds in capital markets, will further reinforce the soundness of the banking sector and allow the recapitalization to take place, at least partially, in the market, alleviating the pressure on rates for sovereign debt. One of the savings banks that failed the stress tests, Banca Civica, has already raised funds from JC Flowers, the US private equity group. Some investors are already giving a vote of confidence with their money.
The stress test record also serves as a useful reminder of two supervisory aspects. The first one is that nonperforming loans (NPLs) in the Spanish mortgage sector are low, barely 3 percent despite 20 percent unemployment, and are expected to remain low even under a stressful scenario. The reason is quite simple: very good underwriting standards. The average loan-to-value (LTV) of the existing book of mortgages is barely above 60 percent, with more than 80 percent of mortgages having an LTV below 80 percent. In other words, despite a very rapid appreciation of house prices, mortgage lending standards remained high. The second one is that, thanks to the Spanish system of dynamic provisioning, Spanish banks have had a very large buffer to meet the expected (and realized) losses. Recall that in 2007, going into the crisis, the Spanish banking sector had a provisioning coverage equivalent to about 200 percent of expected losses, compared to about 50 to 60 percent in the European Union or the United States. Spanish banks complained bitterly when this system was put in place in the late 1990s, arguing that it would put them at a competitive disadvantage. But the supervisor prevailed over the (with hindsight, misguided) lobbying efforts, and now the benefits are clear. These are clear lessons for the future.
So to answer the question: Despite the question marks over the details of the stress tests exercise, should we be more confident today that the EU banking sector is relatively sound, with significantly reduced probabilities assigned to a sovereign event in the European Union? The answer should be yes. Once the European Financial Stability Fund is finally up and running, with its AAA rating assigned, the answer should be an even stronger yes, as an explicit insurance mechanism will be in place. The politics of achieving a European solution are difficult and convoluted, but the skills to deliver are not missing.