Again and again the bank stress tests

 Remember those 15 months ago, July 23, 2010

The objective was whether in an unfavorable macroeconomic and financial scenario, a number of European banks could continue to show a solvency ratio above 6%. 6% was not really ambitious, since it will be recalled that with the Bale 3 this regulatory minimum will reach 7.5%: minimum ratio of core capital of 4.5% and a mandatory capital conservation buffer of 2.5%, intended to absorb losses in a situation of high economic stress

Everyone could go on 2010 summer vacation with peace of mind since only seven of the 91 bank stress failed. No luck, it did not take long to realize that these results were not great value. In fact, during the implementation of the 85 billion € Irish bailout in November 2010, 35 billion were intended to recapitalize the four main Irish banks who have yet managed to pass without incident the summer 2010 stress tests (it was Allied Irish Bank, Bank of Ireland, EBS and Irish Life).

 Closer to us, the European Banking Authority published new bank stress tests on July 15 2011. The perimeter tested included 91 banks representing 65% of banking assets in Europe.

Reviewing the bank stress tests – July 2011

As was seen during the stress tests carried out in July 2010, the latest ones published by the European Banking Authority on the 15th of July 2011 do not include a proper measure of market systemic risk. This hinders their (...)

Read article

2 scenarios were developed: a baseline scenario with the main macroeconomic forecasts implemented, and especially an adverse scenario retaining theoretical assumptions of degradation of the economy: 0.5% GDP decline in the euro area, 15% decline of European stock markets and housing markets crash, rising cost of interbank repo market, however no sovereign default considered

Again everyone could go (or rather from thought) quietly in 2011 summer vacation, as the results of the macroeconomic and financial stress tests were very reassuring, since only 8 banks had failed the tests imposed on 91 European banks (ie they showed under stress conditions a core tier 1 ratio below 5%). For cons, the markets were quickly overtaken by history with a disastrous August for banking stocks and the bankruptcy of Dexia early October (although it had passed the stress tests)

 They have today announced a third round of stress tests

This time to gain credibility and to defeat a large number of banks, the European Banking Authority (EBA) would set the minimum capital adequacy ratio to 9%. Let’s remember what is this solvency ratio: it is the ratio of regulatory capital and capital consumption for the activities of credit and capital markets; to get the ratio, we multiply this ratio by 8%. In other words the lower your numerator (lower regulatory capital due to decreases in performance), the more you degrade this ratio; as far as that goes the higher the denominator (rise in capital consumed as a result of market losses, for example), the more you degrade this ratio.

According to Goldman Sachs, 5 European banks could fail the EBA tests

In a research note that was published on Friday, the American bank estimated 50 European banks out of 91 could fail the new tests prepared by the European Banking Authority (EBE)…

Read article

Besides Goldman Sachs recently estimated that the new stress tests could lead to recapitalization needs of 298 billion Euros and on the 91 selected European banks, 68 would fail the new stress tests, considering a 9% minimum standard of solvency. Analysts at the U.S bank relied on the following assumptions of discount on sovereign debt : 60% of the value of Greek securities, 40% of Irish and Portuguese titles, and finally to 20% on Italian and Spanish securities.

Certainly Goldman Sachs stress tests go further than anyone imagined so far as we assume the partial default on sovereign debt of some sovereign countries leading banks to pass provisions in their accounts and thus to recapitalize. So it’s more than just stress the portfolios of sovereign classified for accounting purposes as trading book and thus directly impacting the income statements (these portfolios represent only 15% to 20% of banks positions in sovereign debt); it is also about stressing all sovereign debts that that are not classified as trading but in what is called the banking book (80% to 85% government bonds held) and whose decreased value in terms of IFRS accounting standards are irrelevant to the income statement. Unless one considers that the discount is large enough to override these accounting standards and proceed with the decommissioning of these securities in bad debts. This is what the banks have in their accounts to 30/09/2011 with 21% discount on part of Greek securities held (due before 2020) and what they will do.

 We learn that the EBA (European Banking Authority) would give the banks until mid-June 2012 to recapitalize with two alternatives

• Either the banks are willing to revalue sovereign debt they hold at market value, in which case they should take their core Tier 1 ratio to 9%

• Either the banks refuse to revalue in their accounts the sovereign debt at market value, in which case the requirements in terms of solvency ratio will be stronger with a minimum level at 9.5%

We already know that banks with an important part of impaired sovereign debt on their balance sheets are rather Greek, Spanish, Portuguese and Italian. So presumably, they will opt for the second alternative and will aim for a higher solvency ratio and recapitalize many more. The challenge ahead, but it’s not really a scoop, it is that banks with the most difficulties to recapitalize (via private funds) have the greatest needs in this area

 We can therefore say that the banks solvency is really stressed by taking into consideration significant discounts on sovereign debt, but that is not enough.

A financial institution can be considered solvent and sufficiently capitalized but fail for another reason: the illiquidity.

Do not simply stress the bank solvency, it is also necessary to stress its liquidity

Mory Doré

And this modeling exercise is extremely difficult to measure because it is about confidence in finance, the banking system in general and different banking institutions. Northern Rock in 2007 and Dexia in 2011 show that today financial crisis is more a confidence crisis than a pure solvency crisis , and thus banks ability to raise liquidity from customers and markets. Recall that the solvency ratios of Northern Rock in 2007 and Dexia were normal but these institutions have fallen because they could no longer refinance short-term assets (market close and drain customer deposits)

Moreover, you should know that most of the crises in financial markets are related to liquidity problems and more specifically the inability of some players to buy financial assets and/or the obligation to others to sell. The reasons for this situation are multiple

Perspective and real solvency of banks

Can a universal bank go bankrupt ? An objective, rigorous and professional answer is built in three stages: credibility of the banks performed stress tests - assessment of prudential regulation to come - understanding the evolution of the banking business (...)

Read article

• Excessive debt that prevents purchase of financial assets

• Prospects for financial assets return too low in relation to the cost of positions funding (including absence of buyers and sellers are present)

• Prudential rules leading to forced sales (compliance with the limits, reaching stop loss)

• Commercial constraints leading again to sales, in order to meet withdrawals of institutional and individual customers, themselves subject to various constraints

• Accounting constraints when controlling the income statement leading to sell healthy assets still with capital gains to fund capital losses on other assets

Certainly there will need to recapitalize and bring back most banks to solvency to restore solvency ratios challenged by write -downs but it will not be enough to solve the financial crisis.

When banks balance sheet are sufficiently sanitized and that transparency and simplicity will be back, the deficit of public confidence vis-a-vis banks and banks vis-a-vis other banks will disappear with a gradual return to normalcy in the functioning of the interbank market.

We are still far today as the ECB continues to replace the interbank market by injecting liquidity in unlimited quantities in the system and setting up extraordinary bidding on abnormally longer maturities for a central bank. It announced in early October a 12 months long-term repo starting in November 2011 and a 13 months long-term repo starting in December 2011 (which will allow banks to pass without too much difficulty in terms of liquidity transition season in 2011 and transition year end 2012)

Just a number, since March 2008, the Bulk of liquidity to European banks allocated each month during the ECB open market operations has averaged around 565 billion Euros. With a very abnormal distribution considering 152 billion during normal refinancing operations periods of 1 week and 413 billion in extraordinary repo operations periods of 1 to 12 months. Evidence of the deep dysfunction of the money market in the Euro zone.

Stress tests can therefore take into account liquidity risk with assumptions about rising cost of liquidity. But it is not enough to stress the price of liquidity, we need to also stress in volume and to assess a number of elements such as

 Reserve in liquid securities, secured and highly rated from the institution and the ability of these to be readily marketable securities, including in situation of market disruption

 Excess reserves held at central bank

 The evaluation of the mobilized collateral wealth from the central bank (securities and private loans eligible for the ECB open market operations; eligible financial assets)

 The quality and importance of mortgages and loans to local government that can be backed by the issuance of bonds (mortgage bonds, asset backed securities, covered bonds)

 The stability and diversification of customer deposits, the average maturity of these resources, contractual terms... On this subject, credible stresses must show whether or not an institution may provide a good part of its credits production to the economy (individuals, small...) in scenarios of stability regarding the collect of resources in the balance sheet and even the flight of customers deposits in particular.

So we know today that there is less risk of stress events that may have a systemic dimension rather than continuing to publish satisfactory stress tests but not very credible considering the tests realized on the solvency and liquidity of banks

We are told that if we stress disaster scenarios, it may enter a downward spiral of self-fulfilling prophecies. Not necessarily if one combines strong responses at the policy level, institutional level and the banks recapitalization and their changing business model.