The results of the European Union-wide stress testing of banks were announced recently, with 91 banks (including 27 in Spain) tested. Seven (including five Spanish cajas) failed, requiring a combined bank capitalisation of €3.5bn. The purpose of the tests was both to create more confidence in the soundness of European banks, with the hope of freeing up the interbank lending market, in particular, in relation to Spain, and to create reassurance generally in the capital markets to boost banks’ share prices. When the US tested its top 19 banks last year, confidence in the financial system appeared to improve and it was hoped that the same effect would occur here.
It is too early to say whether the interbank lending market has been freed up by the EU stress tests because this requires banks to have confidence in each other. Spain’s added detail disclosure of its high risk property lending, together with the fact that it tested 27 of its banks, however, should certainly assist in its efforts to increase confidence.
It does appear that the results, together with the watering down of some of the more controversial Basel proposals for reform, have raised the share price of banks. Barclays, for instance, which had the fifth highest stressed tier one ratio of 13.7 per cent, has seen a rally in its share price. The combination of these announcements appears to have created some investor confidence.
There have, however, been some criticisms of the test. The main criticisms are:
- Most concern has been raised in relation to the sovereign debt holdings of banks, but it has not been adequately dealt with in the tests. The banks (apart from six of the 14 banks tested in Germany which have seen their share prices fall as a result and may be pressured into disclosing details) disclosed their sovereign debt holdings; however, they only stress tested the trading book position and not the entire holding by banks. This was on the basis that it is not plausible that a country would default. Any EU aid provided to governments cannot last forever, however, and once it is withdrawn the question of whether governments can self fund themselves does become a scenario which should be considered.
- The wider definition of tier capital was used in the tests rather than the narrower core tier one definition which only permits shares and retained earnings. The wider test included hybrid investments, some of which were discredited during the crisis. This decision may have been taken because if the core tier one test had been used, Germany and Spain would have had issues. Germany’s Landesbanken, the state banks, for example, are provided with support in the form of hybrid instruments, whereby the state does not become an owner of the banks and, as such, it would not have been able to include those instruments in the narrower test.
- The fact that only seven banks failed has led to a criticism that the scenarios were benign and the tests engineered to ensure a pass mark by virtually all banks.
- Although stress scenarios were applied consistently across the EU, differing accounting standards, in particular, in Germany, and differing interpretations of what counts as tier one capital have distorted cross-border comparisons.
- A large number of the banks stress tested are receiving some government aid in the form of guarantees and liquidity support, but tests were not run against this aid being withdrawn and the ability of banks to self fund themselves.
The criticisms are valid. The greatest concern regarding the tests, in my view, however, is that they may be being viewed generally as indicating the safety of banks. The stress test was, and should be seen, simply as a test of how well banks can meet the regulatory capital requirements (tested against a tier one ratio of six per cent) against a series of adverse stress scenarios for 2010 and 2011, without needing bank capitalisation. It is incorrect to say that they indicate the safety of the banks.
Banks which went into liquidation or required government support as a result of the crisis were meeting their capital requirements prior to the crisis. This included Northern Rock, which was about to distribute its excess capital to shareholders just before it ran into difficulties. It is correct to say, however, that there was not sufficient adverse stress testing of the capital requirements by Northern Rock against the fact that its lending portfolio was risky. If there had been sufficient adverse stress testing, Northern Rock may have decided or been required by the Financial Services Authority to hold the capital against that risk and the incident would not have happened.
The main point is that if a country defaults there will be another banking crisis, due largely to banks’ sovereign debt holdings. It does not take a stress test to work that out. In addition, there is a danger that the banks’ concentration on passing the stress tests has led to them shoring up capital and reducing their lending, which will not assist in attempts to boost the economy.