Dutch banks resilient in European stress test

Press release
Date 23 July 2010

The four Dutch banks that, in cooperation with De Nederlandsche Bank (DNB), participated in the stress test of the Committee of European Banking Supervisors (CEBS), prove to be resilient in this test.

ABN Amro, ING Bank, Rabobank and SNS Bank have calculated the impact of the stress on their solvency positions. The average Tier 1 ratio after two years of stress would amount to 10.3%. The Tier 1 ratio is an indicator of the solvency level of banks which includes only the highest quality of capital. All banks stay well above the benchmark set by CEBS at 6% Tier 1 ratio (the regulatory minimum is 4%). This follows from today’s publication of the results by the banks.

The test has been performed for 91 European banks (representing around 65% of the European banking sector), on request of the European Council of Ministers of Finance. The four Dutch banks represent almost 75% of the Dutch market in terms of assets. In a stress test, banks are asked to calculate the impact of a hypothetical scenario, to assess whether they would be able to withstand losses following from it. Given the extreme developments in recent months and the volatility of financial markets, the results are published on a bank-by-bank basis this time. CEBS set the benchmark Tier 1 ratio required to at 6% (which is two percentage points above the regulatory minimum of 4%), just as in the US stress tests of 2009. The stress test also provides transparency on the so-called sovereign risk of banks. Not only is this risk included in the scenario, exposures on European governments are published by all banks as well.


The stress test comprises two scenarios: a “benchmark”, and an “adverse” scenario. The scenarios have been developed by the ECB and the European Commission in close cooperation.  The macro-economic scenarios include a set of key macro-economic variables, differentiated for EU Member States, the rest of the EEA countries and the US for the next two years. All banks are to use these predefined shocks. The benchmark scenario is based on current expectations about economic growth, and serves as a comparison for the adverse scenario. For the Netherlands, the benchmark scenario assumes first signs of economic recovery, although unemployment still rises. House prices and commercial real estate prices remain stable (see Table 1).

The adverse scenario is the core of the stress test, and presupposes a further economic downturn, leading to substantial stress (see Table 1). The scenario consists of a continuing recession, during which economic growth in the Netherlands declines by 1% in the next two years on top of the already stressed year of 2009. Main reasons for this downturn are a global confidence shock, leading to decreasing world trade and falling asset prices. In this hypothetic scenario, Dutch unemployment reaches a level of 7%. The 3-month interest rate increases over a short time horizon to 3.3% and the 10-year interest to 4.9%. This implies a flattening of the yield curve and decreasing interest rate margins. In addition, banks have to incorporate a fall in house prices of 20% over the coming two years. A similar shock is assumed for commercial real estate. These shocks are nearly as large as the decline in house prices in the beginning of the 1980s. Next to that, equity markets fall by 20% in the adverse scenario. In addition, the scenario prescribes shocks for spread risk and banks’ trading books.

The adverse scenario considers sovereign risk as well. The hypothetical sovereign shocks lead to write-downs on exposures in the banking book and additional haircuts on sovereign bonds in the trading book of banks. It is furthermore assumed that these shocks in equity, credit, and real estate markets occur simultaneously and that the banks’ balance sheets will not increase over the scenario horizon.

Table 1: Scenario’s for the Netherlands

Table 1



Dutch banks have calculated the impact of the scenarios on their consolidated balance sheets. In these figures, the impact of the current credit crisis has been taken into account. The stress test assumes a new shock on top of losses already realised. Banks factored the developments in the first quarter of 2010 into their calculations. The participating banks have used their internal models to calculate the impact of the scenario. The stress test does not allow banks to take new measures that mitigate the impact of the scenario.

DNB has applied a rigorous evaluation of the institutions’ calculations and assumptions and has also monitored the consistency of the results. These are dependent on banks’ internal models and their interpretation of the scenario. DNB assessed whether banks have adequately translated the macro economic shocks of the scenario into loss ratios.


The results show that in the adverse scenario, aggregate impairments for the four Dutch banks amount almost to EUR 21 billion over two years. This is 24% of their Tier 1 capital. Hence, the average Tier 1 ratio for the Dutch banks declines from 12.0% in 2009, to 10.3% in 2011, including sovereign risk. If the impact of the sovereign stress is excluded, the impact is slightly smaller, and the Tier 1 ratio would decline to 10.5%. The impairments are largely accounted for in the profit and loss account. Dutch banks appear to be sensitive to losses in the corporate portfolio. The average loss ratio on corporate loans is 1.3% after two years of the adverse scenario, compared to 0.9% on the retail portfolio.


For all four banks, the Tier 1 ratio remains well above the benchmark of 6% defined by CEBS (against the 4% regulatory minimum), due to the relatively large capital buffer in the reference period. The four Dutch banks prove to be resilient in the stress test. The results of this stress test will be used as input for a broader assessment of banks’ capital position in the Basel II framework.