Testing the Banks' Resilience

Danish banks are generally found to be robust. Stress tests show that most Danish banks would record losses if exposed to tough economic scenarios, but that they would generally avoid solvency problems.

A static sensitivity analysis, based on the banks' earnings and capital structure at end-2007, shows virtually unchanged resilience compared with 2006. The banks have become more exposed to rising financing costs and increased losses on credit portfolios.

The recent financial turmoil is reflected in market expectations towards the banks. The expectations have generally been adjusted downwards, but the market's expectations of the Nordic banks are still higher than market expectations of other European banks and US banks.

METHODS FOR TESTING THE BANKS' RESILIENCE

In this chapter, the resilience of the Danish banks is tested in dynamic macro stress tests – where the results reflect the interaction between macroeconomic conditions, financial conditions and banking conditions over a prolonged period – and in static sensitivity analyses based solely on the banks' financial statements.

Three macroeconomic and financial stress test scenarios have been constructed on the basis of the risks described in the previous chapter. The effect, over a 3-year horizon, on the banks' earnings and solvency of these extreme scenarios is examined. The purpose is to provide an estimate of the implications for the banking sector of an extreme shock to the economy in general or the financial sector specifically.

A static sensitivity analysis is performed in which selected accounting items are changed according to a ceteris paribus approach, based on the latest annual profit/loss statements and capital structures. The earnings parameters are changed one by one, and for each scenario the effect on the banks' profit and capital structure is calculated. The results are compared with corresponding scenarios for the preceding year.

Finally, the market's assessment of the Nordic banking groups is analysed. The first step is to examine CDS (credit default swap) spreads, reflecting the market's assessment of the estimated failure rate for the groups. The second step is to estimate how much the market value of the Nordic banking groups' assets may fluctuate and still be accommodated by the liquidity reserves.

MACRO STRESS TEST – DANISH BANKS FOUND TO BE ROBUST

The robustness of the banking sector is tested by means of Danmarks Nationalbank's stress test model, cf. the chapter Stress Testing of the Financial System. Three scenarios have been constructed for testing the robustness of the Danish financial system, cf. Box 6:

• Scenario 1: The subprime crisis continues and leads to a prolonged recession in the USA. The price of interbank financing rises sharply. The increase is partially passed on to the customers. Growth in the US economy is negative for eight quarters.
• Scenario 2: Increases in commodity prices.Commodity prices, especially oil prices, rise sharply, and interest rates are raised to keep inflation at bay.
• Scenario 3: Property price drop.Interest rates and unemployment increase, while property prices and the value of assets pledged as collateral for bank loans decrease.

The stress test scenarios are compared with a baseline scenario that is considered to be the most likely development in the Danish economy and the financial sector.^[1] The time horizon for each scenario is three years.

In order to illustrate the results of the stress test model, the banks have been divided into three categories for each scenario, i.e. green, yellow and red. The green banks post profits in all three scenario years. The yellow banks post a loss in at least one of the scenario years. The red banks are unable to meet the statutory capital requirement over the three years.

None of the analysed banks will record losses over the horizon of the baseline scenario, and the return on the banks' core capital is stable in the period, cf. Charts 30 and 31. The earnings of 13 of the 16 banks become negative in scenario 2, and one of the banks experiences solvency problems within the scenario horizon, cf. Chart 30. In scenarios 2 and 3, 10 and 7 banks, respectively, record losses in at least one of the three years.

The scenario with higher oil prices and interest rates has a relatively strong impact on the larger banks, while the scenario with a property price drop and higher loss ratios, especially on building-related sectors, has the strongest effect on the banks in group 2. The subprime crisis scenario affects the banks particularly through higher financing costs, while the economic slowdown as a consequence of the US recession has only a minor effect. Correspondingly, the effect of rising oil prices is moderate when viewed in isolation, but relatively strong if accompanied by rapidly increasing interest rates. The timing of the impact on the banks' financial results of the stress test scenarios also varies, cf. Chart 31.

Overall, the results of the stress test model show that the banks are generally robust to the very tough shocks in the constructed scenarios.

SENSITIVITY ANALYSIS – UNCHANGED RESILIENCE IN 2007

The banks' resilience is virtually unchanged compared with 2006, cf. Chart 32. The banks are slightly more exposed to rising financing costs and increasing losses on lending portfolios. An increase by 1 percentage point in losses on loans for the sector as a whole would have resulted in eight banks recording losses, i.e. twice as many as in 2006. More banks would have recorded losses in the event that the largest counterparty bank in the Danish uncollateralised day-to-day money market failed, while the effect of losses amounting to 10 per cent of large exposures would have been virtually unchanged in 2007 compared with 2006.

A corresponding test of the Nordic groups shows that none of the groups would have had solvency problems in 2007 if loan losses had increased by 1 percentage point, although one group would have posted losses. An increase by 2.5 percentage points would have caused solvency problems for several of the groups. Compared with 2006, the Nordic groups have become more exposed to both increasing losses and rising financing costs.

THE MARKET ASSESSMENT OF THE NORDIC GROUPS HAS DETERIORATED

Substantial widening of credit default swap spreads for the banks^[2]
A credit default swap, CDS, is a financial instrument that is used e.g. to hedge the credit risk on a bank. The price development, typically measured as interest-rate spreads, on a bank's CDSs thus reflects the market's assessment of the probability that the bank will fail within a given period. All other things being equal, a wider CDS spread indicates market expectations of a higher failure rate for the bank.

As the US subprime crisis has evolved, the CDS spreads for financial companies have widened to record levels. This development can be attributed to such factors as subprime-related losses in the financial sector, concerns about the macroeconomic development in the USA and unrest associated with several monoline bond insurance companies.

The spreads for European banks also widened considerably. Thus, the market's assessment of the probability of the European banks failing within a 5-year period has risen substantially in 2008, cf. Chart 33. This shows that problems in the financial sector increasingly tend to spread due to the higher degree of financial integration.

Nordic banks have also been affected, and the CDS spreads have widened strongly in 2008, cf. Chart 33. In terms of CDS spreads for financial companies, the Nordic banks are still at the narrow end of the scale, however. One reason is that Nordic banks' direct exposure to the subprime market has been smaller than that of other banks, so the Nordic banks have initially steered clear of large losses.

Since the beginning of 2008, CDS spreads associated with the two large Icelandic banks have widened substantially, by more than 1,000 basis points.

The market assessment is that Nordic groups are more exposed
On the basis of equity prices and accounting data, Danmarks Nationalbank has estimated a market-based risk measure, distance to insolvency^[3], for the Nordic groups. The distance to insolvency measures the fluctuations in asset market value that can be accommodated within the banking groups' buffers while still observing the minimum capital requirement of 8 per cent.

The distance to insolvency is measured by the number of standard deviations for fluctuations in the estimated market value of the bank's assets. The greater the distance to insolvency, the more robust are the banks, according to the market assessment. A distance to insolvency of e.g. two standard deviations can be interpreted as a risk of around 2.3 per cent that the asset value will decline so much that the bank becomes unable to meet the statutory minimum capital requirement.^[4]

The distance to insolvency for the individual Nordic groups is 1.5-3 standard deviations, compared to 2.5-4.5 at the beginning of 2007, cf. Chart 34. The market-based risk measure shows that the probability of solvency problems for the Nordic groups is greater today than at the beginning of 2007, according to the market assessment.