Further analysis of the stress test of banks' capital adequacy in Financial Stability 2/2010.

• Author: Havro, Goril Bjerkhol

The adverse scenario in FS 2/10

Since FS 1/09 a downturn in the international economy has been the most important risk factor in the adverse scenarios in Norges Bank's Financial Stability reports (see Table 1). In FS 2/10 production for Norway's trading partners is assumed to weaken broadly in line with that assumed for the euro area countries in the EU-wide stress test exercise conducted by the CEBS (Committee of European Banking Supervisors) in summer. (1) The CEBS applied the assumption that GDP growth would be about 3 percentage points weaker, with a time horizon of two years, than in the benchmark scenario. The adverse scenario in FS 2/10 applies a time period of 3 1/2 years. The difference in growth for trading partners is about 5 1/2 percentage points for the overall period, which implies somewhat stronger international growth than in FS 1/10.

Weaker global growth may trigger a drop in oil prices, an important channel into the Norwegian economy. In the adverse scenario, oil prices fall to about USD 50 per barrel. Oil prices in the adverse scenario are in the lower 5% percentile for futures options prices at the end of the second half of 2011. (2) Since oil prices have been relatively stable recently, this assumption results in a smaller fall compared with previous reports.

A drop in oil prices could, in isolation, lead to a depreciation of the krone. On the other hand, GDP growth among our trading partners is slower than in Norway, and uncertainty abroad is considerable. Furthermore, foreign interest rates are low. As in FS 1/10, it is therefore assumed that the krone exchange rate remains broadly unchanged in relation to the benchmark scenario.

As a result of lower growth and weaker price impulses, the key policy rate in the adverse scenario is reduced. This contributes to a dampening of the decrease in economic growth. A lower key policy rate is somewhat counteracted by the rise in premiums in international money markets. External turbulence spills over to Norwegian money markets, and premiums increase by up to one percentage point. The increase occurs as a result of uncertainty concerning sovereign debt and the international financial sector. It is also assumed that lending margins remain approximately at the current level. Consequently, the fall in interest rates facing households and enterprises is dampened.

These shocks reduce growth in the Norwegian economy and unemployment rises. A downturn in international growth will lead to a reduction in manufacturing production in Norway, and particularly in traditional exports. It will also lead to an increase in household pessimism. The adverse scenario assumes that Norwegian households' expectations concerning their own financial position and the country's economy weaken. Low oil prices will reduce investment in the oil sector and related industries. Increased unemployment, weaker expectations and lower income compared with the benchmark scenario lead to a drop in house prices. Reduced investment and lower house prices also lead to lower debt growth for both households and enterprises. A fall in house prices, which reduces collateral values and thereby loan-financed consumption and investment, amplifies the downturn in the real economy.

[GRAPHIC 1 OMITTED]

Banks' results in the stress test

As a result of negative developments in the real economy, banks' results turn negative in 2012, (see Chart 1). However, banks' results weaken to a somewhat lesser extent than in FS 2/09 and 1/10, primarily because the increase in loan losses is lower (see Chart 2). This partly reflects that the share of problem loans does not increase as much as in the adverse scenario in the previous report as economic developments are somewhat more favourable in this scenario. In addition, this reflects a change in the distribution between losses to households and enterprises in this stress test. While losses in households are higher, losses in enterprises are lower compared with the previous report. For the banks in the stress test (3), which have a large share of lending to the corporate market, overall losses are reduced.

It is assumed that due to competitive conditions banks cannot raise their interest margin during the stress period. Banks' net interest income is thus reduced in the adverse scenario when their funding costs increase. The premium on their total market funding (4) rises by 10 basis points under the adverse scenario. This is a lower increase than in the previous report where the premium rose by 30 basis points in relation to the benchmark scenario.

Banks' return on securities is assumed to be the same under the adverse and benchmark scenario. The sensitivity analyses presented below show banks' vulnerability to variations in the return on their security portfolios. As Norwegian banks have limited holdings of high-risk government bonds, the stress test does not include specific analyses of the effects of shocks to various countries' state finances on banks' security portfolios.

[GRAPHIC 2 OMITTED]

[GRAPHIC 3 OMITTED]

The Tier 1 capital ratio in the adverse scenario remains well above the minimum requirement of 4%. It also remains above 6%, i.e. the new minimum Tier 1 capital requirement set out by the Basel Committee (see Chart 3). The Tier 1 capital ratio falls through the period as banks' lending becomes more risky and thereby increases risk-weighted assets. Since the credit risk associated with lending to enterprises rises less than in the adverse scenarios in the previous reports, the risk weights do not increase as rapidly.

[GRAPHIC 4 OMITTED]

Losses for Norwegian banks in the stress tests for the years 2010-2011 (5) in FS 2/10 are comparable to Swedish and Danish banks' losses in the CEBS stress test (see Chart 4). In the CEBS stress test many of the banks used their own models, i.e. a bottom-up approach, while a top-down analysis for all Norwegian banks was used in FS 2/10. Moreover, the time horizon in our adverse scenario is longer than in the CEBS stress test. Since banks' losses rise in line with the duration of the downturn, the FS 2/10 adverse scenario is stricter than the CEBS adverse scenario. (6) It is therefore difficult to compare the results of these two stress tests directly.

Sensitivity analyses of banks' results and capital adequacy

Scenario analyses estimate the effect on banks' accounts of the materialisation of one or more risk factors. The macro stress scenario's design and bank-specific assumptions have a bearing on banks' results and capital adequacy in the adverse scenario. To provide a more comprehensive picture of Norwegian banks' vulnerability, it is useful to look at the effect of different paths for some key variables.

When the exchange rate is assumed to remain unchanged compared to the benchmark scenario, changes in oil prices have a considerable impact on macroeconomic developments. Even though oil prices are high today, they have varied widely over time. Towards the end of 2001 oil prices dropped to USD 20 per barrel. If we apply the assumption that oil prices move down to USD 20, but let the interest rate and real exchange rate follow the same path as in the initial adverse scenario, the share of problem loans rises by about one and a half percentage points (see Chart 5).

[GRAPHIC 5 OMITTED]

[GRAPHIC 6 OMITTED]

In the FS 2/10 stress test, the real exchange rate is held approximately unchanged in relation to the benchmark scenario. According to the estimated relationship in the macro model used in the stress test, the other assumptions in the adverse scenario could result in a weakening of the krone of about 10%. In this case the share of problem loans is around a half percentage point lower (see Chart 6). With a weakening of the krone to levels observed in autumn 2008, the share of problem loans would have fallen further. A...