Assessment of the Banks' Resilience

The negative economic development will induce the banks to make higher write-downs in the coming years than seen in recent years. Estimates of bank losses over the coming years are subject to great uncertainty, both in terms of the economic outlook and the size of write-downs resulting from a given economic development.

In the baseline scenario, the banking sector as a whole will meet the statutory solvency requirement throughout the period, if the banks exploit the opportunity to receive capital injections under Bank Rescue Package II. The stress test shows that if the economy is hit by more very negative shocks, the banks may need to strengthen their capital base further. The banks' resilience will increase significantly if they exploit the opportunity under Bank Rescue Package II to receive capital injections where the injected capital may be converted into share capital.

The financial markets currently expect the financial strength of the banks to exceed the statutory minimum requirement, highlighting the importance of injected capital of high quality.

Market expectations in relation to the banks, expressed as equity prices and credit spreads, emphasise the uncertainty about the banks' capital adequacy and capital base.

Macro stress test – scenarios

The robustness of the banking sector is tested by means of Danmarks Nationalbank's stress test model, cf. Box 14. The development in the banking sector is modelled in a baseline scenario and two stress scenarios, all of which cover the period from 2009 to 2011.

DANMARKS nationalbank'S stress test model

Box 14

The macro stress test model is described in Financial stability 2008. In comparison with Financial stability 2008, the model is estimated on the basis of quarterly data from the period 1991-2008 for the 14 largest banks. The model architecture is unchanged. Model projections are based on the banks' financial statements for 2008. A few adjustments have been made to the model projections as a result of the financial statements for the 1st quarter.

The data basis for the stress test model includes few observations from periods of intense pressure on the banking sector. Accordingly, it is difficult for the model to map the historically tough economic environment in the scenarios directly onto the banks' financial results. This makes the projection of the banks' losses particularly challenging and, therefore, pure model projections have been adjusted, cf. Box 16. In addition, the model has been adjusted to take Bank Rescue Packages I and II into account: Under the government guarantee issued in October 2008, the banks are not allowed to pay dividends until end-2010, and the government guarantee puts the banks to expense in the form of own contributions to the government guarantee scheme. Moreover, the interbank contagion module has been removed as a result of the government guarantee.

The model results should be interpreted with caution. For example, the stress test model does not take liquidity risk into account, including the fact that, during some periods, the banks may be unable to raise funds in the capital markets. Another factor not taken into account is that the management of a bank can react and change its strategy – for better or for worse – as the scenarios unfold. It is assumed that the development of a bank's foreign branches mirrors the development of its Danish branches. Finally, the model considers the impact of economic development on the banks' credit exposures at sector level. This means that the model cannot take into account differences in the credit quality of different banks, cf. Box 15.

Baseline scenario
The baseline scenario is believed to be the most probable development in the Danish economy and the financial sector. The scenario reflects the expected impacts of the financial crisis, including some degree of credit rationing and the global recession.

In the baseline scenario, GDP contracts by just under 2.5 per cent in 2009, cf. Table 4. From 2010, the economy will begin to pick up again, albeit at a slow pace. GDP growth for 2010 and 2011 is 0.5 and 1.5 per cent, respectively. In the baseline scenario, unemployment is expected to rise sharply, from 1.8 per cent in 2008 to 4.5 per cent in 2009, increasing further to 6.5 per cent in 2010. After 2010, unemployment is expected to stabilise. The banks will see an average annual loss ratio of around 1.3 per cent in each of the three years of the scenario.

Baseline scenario and scenarios from financial stability 2008, 2nd half					Table 4
		Scenarios from Financial stability 2008, 2nd half
	Baseline scenario	Baseline scenario	International recession	Credit crunch	Combination
2009
GDP, per cent, year-on-year	-2.4	-0.0	-1.6	-1.6	-2.4
Unemployment rate, year-end	4.5	3.2	4.5	4.3	5.0
2010
GDP, per cent, year-on-year	0.5	0.4	-0.8	-1.5	-1.8
Unemployment rate, year-end	6.5	4.2	6.5	6.4	7.6
2011
GDP, per cent, year-on-year	1.5	NA	NA	NA	NA
Unemployment rate, year-end	6.3	NA	NA	NA	NA

danish stress tests vs. us stress tests	Box 15
Macro stress tests distinguish between "bottom-up" and "top-down" calculations. The basis of both methods is a stress scenario, typically a scenario for a stressed economic environment. The scenario may be based on a historical event or a hypothetical future event. In the bottom-up approach, a stress test is conducted on the specific exposures and other risk factors of the banks and the figures are subsequently aggregated. In the top-down approach, a stress test is conducted on the aggregated exposures and other risk factors of the banks and the stressed figures are subsequently allocated to the individual banks. The advantage of the bottom-up approach is that the results best reflect differences in the risk profiles of the individual institutions. The drawback of the approach is that it is highly data-intensive. The stress scenarios are often determined centrally, while the banks themselves perform the analysis of the impact on their exposures and other risk factors. Differences in the calculation methods of the institutions can make aggregation difficult. At the same time, it may be expensive for the institutions. Conversely, the advantage of the top-down approach is that it is relatively easy to implement. The drawback is that aggregated data capture only the effects for the sector as a whole, which can make it difficult to take the risk profiles of the individual institutions into account. Danmarks Nationalbank's stress test model is a top-down model for stress testing the Danish financial sector. This means that only minor allowance is made for differences in the banks' risk profiles; accordingly, the model results should be interpreted as a stress test of the system of banks, rather than a test of the individual banks. Therefore, the model results are not published for the individual banks. The stress test conducted by the Federal Reserve of the 19 largest US banks, as part of the Capital Assessment Program (CAP), is a bottom-up assessment of the effect on each bank in a centrally determined stress scenario. ¹ Methodologically, the stress test was conducted in compliance with the description above. The results were published for each bank on 7 May 2009. A key element of CAP is that banks assessed as needing to augment their capital base have access to government capital injections immediately after the stress test. Had this not been the case, the market response to the publication of the results could have had disastrous consequences – both for the banks performing poorly in the stress test and for the financial system as a whole. The CAP stress test showed that in the stress scenario, the 19 participating banks stood to lose 9.1 per cent of their total lending during the two-year horizon of the scenario. 10 of the 19 banks needed to augment their capital base to meet the requirements specified for the stress test. Most of the banks that failed the stress test requirement failed on account of the non-hybrid core capital requirement. In principle, they had sufficient capital, but the quality of the capital was too poor, cf. Box 17. Allowing for the measures already implemented by the banks in the 1st quarter of 2009, the 10 banks need to increase their core (Tier 1) capital by 75 billion dollars. In Europe, CEBS (Committee of European Banking Supervisors) will conduct a stress test of the European banking system. ² No groups based in Denmark participate in the stress test. The purpose of the stress test is not to establish whether individual banks may need recapitalisation. The result of the test will not be published.
¹ The approach and the two scenarios are described in The Supervisory Capital Assessment Program: Design and Implementation. The results are described in The Supervisory Capital Assessment Program: Overview of Results. Both can be found on the Federal Reserve's website: www.federalreserve.gov. ² Cf. CEBS press release of 12 May 2009.

Compared with the baseline scenario in Financial stability 2008, 2nd half, this represents a substantial downward revision of the forecast of economic growth, cf. Table 4, prompted by a real-economic deterioration in Denmark and internationally. The new baseline scenario includes a downward revision of GDP growth by just under 2.5 percentage points (from zero growth to negative growth of 2.4 per cent) for 2009 relative to the previous baseline scenario. Similarly, unemployment has been revised upwards by around 2 percentage points at end-2010, which was the end of the previous baseline scenario. The current baseline scenario mostly resembles the scenario of international recession from Financial stability 2008, 2nd half.

The results below assume that all banks receive hybrid core capital injections, bringing their Tier 1 ratio to 12 per cent of risk-weighted assets, but ensuring that their Tier 1 capital is increased by at least 3 percentage points. That is, the banks will receive the maximum capital injection under Bank Rescue Package II. It has been assumed that the hybrid core capital injected cannot be converted into share capital. For calculation purposes, it has also been assumed that all banks pay an annual rate of interest of 10 per cent on the government capital received.

As the banks receive capital injections, their excess capital adequacy, in general, will be higher at end-2009 than a year earlier, cf. Chart 41 illustrating the development in the banks' excess capital adequacy in the baseline scenario with and without Bank Rescue Package II. In 2010 and 2011, the banks' excess capital adequacy shows a declining trend. If the banks do not receive capital injections, approximately half of them will have problems meeting the solvency requirement during the scenario period. This highlights the importance for the banks of strengthening their capital base, e.g. through Bank Rescue Package II.¹

excess capital adequacy in the baseline scenario – with and without bank rescue package II	chart 41

Source: Danish Financial Supervisory Authority, banks' financial statements and own calculations.

The macro stress test model assumes that the banks do not actively adapt their business models to macroeconomic developments beyond what is implied in the model through the estimated relations. However, there are recent indications that the banks have been able to significantly increase their interest margins, bolstering their ability to absorb losses. Similarly, other things being equal, it will be easier for the banks to meet the solvency requirement if they reduce their risk-weighted assets. One way to reduce risk-weighted assets is by reducing lending. Contraction of bank lending is known from previous crises. Total lending e.g. dropped by approximately 10 per cent from 1993 to 1994 towards the end of the Nordic banking crisis.

Actually, the financial markets of today very rarely accept that a bank only just meets the statutory solvency requirement. Before getting to this point, the bank's counterparties will reduce its credit and trading lines, which will, in practice, put a stop to the bank's operations. Consequently, most banks are well advised to stay comfortably above the statutory minimum requirement.

Stress scenarios
In addition to the baseline scenario, two stress scenarios have been constructed. These scenarios are deemed to be very improbable and have primarily been constructed to test the resilience of the banks. Unemployment developments in these scenarios are illustrated by Chart 42, while selected key ratios are shown in Table 5. The rise in unemployment in the stress scenarios is much steeper than previously seen and GDP growth is at a 50-year low. For both stress scenarios, the main reason for the pressure on the banks' financial results is the impact of economic development on the banks' write-downs.

development in unemployment	chart 42

Note: The most recent actual observation is 2008. Source: Statistics Denmark and own calculations.

The two scenarios envisage a negative shock to the Danish economy and a long, deep recession, an "L scenario", combining the negative shock to the Danish economy with a negative shock to the international economy. The economic situation and uncertainty as to how the various stimulus packages will affect the economy imply that the scenarios and their impact on the banking sector are subject to unusually high uncertainty.

Scenario 1: Negative shock to the Danish economy
The Danish economy is hit by a further negative shock, e.g. in the form of a severe credit crunch, causing the banks to cut back further on lending rela tive to the baseline scenario. This effect is reinforced by widespread pessimism among households and companies. The trend of the international econ omy mirrors the baseline scenario, thus the interest-rate developments also mirror the baseline scenario. Compared with scenario 2, cf. below, interest rates thus put further pressure on the domestic economy, in cluding housing prices. The scenario operates with an unemployment rate of 9.6 per cent at end-2011, and the average loss ratio is approximately 2.2 per cent of total loans and guarantees for each of the three years.

development in selected key ratios in the scenarios			Table 5
	Baseline scenario	1: Danish shock	2: L scenario
2009
GDP, per cent, year-on-year	-2.4	-4.0	-4.5
Unemployment rate, year-end	4.5	5.6	6.1
Average bond yield, per cent p.a.	4.1	4.1	3.9
2010
GDP, per cent, year-on-year	0.5	-1.9	-2.9
Unemployment rate, year-end	6.5	8.9	10.2
Average bond yield, per cent p.a.	4.6	4.6	3.2
2011
GDP, per cent, year-on-year	1.5	-0.0	-0.6
Unemployment rate, year-end	6.3	9.6	11.8
Average bond yield, per cent p.a.	5.3	5.3	2.8

Scenario 2: Long, deep recession
While Denmark is hit by a domestic shock, as envisaged by scenario 1, the financial crisis also leads to an international recession that is longer and deeper than expected. Thus, the export markets fail to make a positive growth contribution to the Danish economy. Monetary policy is eased further, where possible. The effects of an international and a domestic shock are mutually reinforcing, providing for a far worse economic environment than envisaged by scenario 1. The result is a very negative economic development for all three years of the scenario. However, falling interest rates will help to cushion the impact on the banks' financial results. In this scenario, the rate of unemployment is 11.8 per cent at end-2011, and the average loss ratio is approximately 2.9 per cent of total loans and guarantees for each of the three years of the scenario.

Results of the stress scenarios

In all of the scenarios, the negative development of the economy impacts the loss ratio of the banks, cf. Chart 43. In the baseline scenario, losses are expected to rise further in 2009, to 1.34 per cent of total loans and guarantees. The high loss ratio will be maintained throughout the period. The scenario is underpinned by rising failure rates observed in Danmarks Nationalbank's failure-rate model, cf. Chart 27, and an increase in the banks' credit-risk measures, cf. Chart 29.

Banking sector's overall loss ratios	chart 43

Source: Danish Financial Supervisory Authority, banks' financial statements and own calculations.

In both of the stress scenarios, the pace of increase in the loss ratio is unusual. It should be noted, however, that some uncertainty prevails as to the level and timing of write-downs, cf. Box 16. In the Danish shock scenario, the loss ratio increases to the level seen during the Nordic banking crisis in the early 1990s. The high loss ratio will be maintained throughout the period. In the scenario with a long, deep recession, loss ratios reach a level of around 3 per cent, or approximately 0.5 percentage points above the level seen in Denmark during the Nordic banking crisis. The loss ratio will remain at the historically high level throughout the period.

bank losses in the stress test model				Box 16
The credit-risk module of Danmarks Nationalbank's stress test model is continuously developed and updated. Under the current circumstances, the credit-risk model provides unrealistically low estimates of the development in the banks' losses – both in terms of the level and timing of losses. The data basis for the estimations includes only few observations of the banking sector under intense pressure, all of which originate from the Nordic banking crisis in the early 1990s. At the beginning of the 1990s, unemployment rose from an already high level, interest rates were high and bank losses were heavy. Now interest and unemployment rates are very low, keeping estimated losses down. The estimated loss level is significantly lower than the loss levels seen in the 4th quarter of 2008 and the 1st quarter of 2009. Part of the explanation could be that the model applied attaches too much weight to the levels of both interest and unemployment rates relative to other economic variables, including the change in the variables in question which do not form part of the estimation. Moreover, accounting rules were changed in 2005, meaning that write-downs were to be booked at a later date. The result has been a number of years with very low write-downs. In terms of the timing of bank write-downs, the absence of a history of high losses under the current accounting rules also complicates matters. In the context of the above, the estimated loss ratios of the credit-risk module have been manually adjusted, based on an assessment of the appropriate weighting of the various elements of the crisis relative to future losses. This approach has been chosen because it provides the best opportunities for taking into account the elements specific to the ongoing crisis. Other, more model-based approaches could have been chosen. The table below illustrates the estimated losses, along with the losses under three alternative projection approaches. The first alternative approach (basic estimation) is based on "pure" estimated loss ratios from the credit-risk module. This estimation provides the basis for the manually-adjusted loss ratios applied in the stress test. The second alternative approach (level-shift estimation) is based on an addition to the loss ratios of the basic estimation to enable them to recreate the level of loss ratios in the 1st quarter of 2009. In the third alternative approach (estimation without interest-rate effect), the credit-risk module has been re-estimated without using the interest rate as an explanatory variable. For the baseline scenario, both the basic estimation and the estimation without interest-rate effect are significantly below the level of the 1st quarter of 2009 – about 140 basis points on an annual basis. The greatest weakness of the level-shift estimation is that the differences between the scenarios seem too small compared with other banking crises.
losses under different projection approaches
Basis points	2009	2010	2011	Total
Baseline scenario
Estimates applied	135	146	115	396
Basic estimation¹	33	58	50	141
Level shift to 2009 Q1²	179	204	198	581
Estimation w/o interest-rate effect³	27	54	52	133
Scenario 1: Danish shock
Estimates applied	194	246	228	667
Basic estimation¹	45	128	125	298
Level shift to 2009 Q1²	190	274	273	737
Estimation w/o interest-rate effect³	42	161	218	421
Scenario 2: Long, deep recession
Estimates applied	258	314	283	855
Basic estimation¹	50	137	117	304
Level shift to 2009 Q1²	194	282	265	741
Estimation w/o interest-rate effect³	47	249	454	750
Source: Own calculations. ¹In the basic estimation, the banks' (transformed) loss ratios are regressed on a number of explanatory variables. The loss ratios of different sectors are modelled separately. Explanatory variables include developments in unemployment, interest rates and housing prices. ²The loss ratios of the basic estimation have been level shifted, so that the adjusted loss ratio is equivalent to the observed aggregated loss ratio of the 1st quarter of 2009. ³Modelled as the basic estimation. All relationships are re-estimated without interest rates in the set of potential explanatory variables.

The steep increase in loss ratios causes bank profitability to drop sharply, leading to a general decline in the banks' excess capital adequacy, cf. Chart 44. In the Danish shock scenario, a few banks will have problems meeting the solvency requirement during 2010, while just under half of the banks will have problems meeting the requirement in the third year of the scenario. In the scenario with a long, deep recession, in which the banks annually lose almost 3 per cent of their total loans and guarantees, the banking sector is hit hard. More than half of the banks analysed will find it difficult to meet the solvency requirement in the course of 2010, and few banks will meet the solvency requirement by the end of 2011. Above it is assumed that the capital injections cannot be converted into share capital.

banks' excess capital adequacy in the scenarios – with bank rescue package ii	chart 44

Source: Danish Financial Supervisory Authority, banks' financial statements and own calculations.

Quality of capital injections
As previously stated, the calculations above assume that the capital injections are not converted into share capital. If, on the other hand, it is assumed that the capital injections are converted into share capital, the banking sector performs significantly better, cf. Chart 45. The main reason is that share capital is regarded as being of higher quality than hybrid core capital in the calculation of the banks' solvency, cf. Box 17. Moreover, the banks' core earnings are positively impacted by a conversion, as interest is payable on hybrid core capital, while this is not the case for share capital. In the Danish shock scenario, the banking sector as a whole pulls through the entire scenario. In the long, deep recession scenario, the banking sector will not have problems meeting the solvency requirement until 2011.

Banks' excess capital adequacy – with bank rescue package II – with and without conversion of hybrid core capital into share capital	chart 45

Note: In the calculations for the scenario with conversion, it has been assumed that the entire capital injection will be converted in 2009, meaning that no interest is payable on the capital injection. Source: Danish Financial Supervisory Authority, banks' financial statements and own calculations.

capital requirements and solvency problems in the stress test model	Box 17
The solvency ratio of a bank is calculated on the basis of its capital base, made up of three elements: non-hybrid core capital (roughly equivalent to the share capital), hybrid core capital and supplementary capital. Use of the three types of capital is subject to two restrictions. Supplementary capital must not exceed 50 per cent of the capital base. Hybrid core capital must not exceed 50, 35 or 15 per cent of the Tier 1 capital (the sum of non-hybrid and hybrid core capital), depending inter alia on the bank's opportunity to convert hybrid core capital into share capital. This means that the composition of the capital base may have an impact on when a bank will have problems meeting the statutory solvency requirement. In the Chart below, the bank in the example needs a capital base of at least 300 to be solvent. Assume that the bank has a capital base of 400, its supplementary capital accounts for 50 per cent of the capital base, and the hybrid core capital accounts for 35 per cent of the Tier 1 capital (and its quality is such that it must not exceed 35 per cent of the Tier 1 capital), as illustrated by example 1. Assume that the bank suffers a loss of 65, bringing its non-hybrid core capital down to 65. As both limits are binding, the hybrid core capital and the supplementary capital can be included in the solvency calculation only to an extent equivalent to 35 and 100, respectively, as in example 1b. In this case, the bank no longer meets the statutory solvency requirement. Now consider example 2 in the Chart. In this case, the bank also has a capital base of 400, but the composition is different: 195 is non-hybrid core capital, 105 is hybrid core capital and 100 is supplementary capital. Assume, as in example 1, that the bank loses 65. Because the composition of the capital base is different, there is no reduction in the extent to which supplementary capital can be included in the solvency calculation. On the contrary, the bank can now include 35 of its hybrid core capital, which can no longer be included in its hybrid core capital (as the hybrid core capital must not exceed 35 per cent of the Tier 1 capital) in its supplementary capital. Even though the capital base and the loss are, from the outset, the same, in example 1 the bank does not meet the solvency requirement after losses, while it does meet the requirement in example 2. The explanation is that the capital in example 1 is of poorer quality to that in example 2.
Examples of capital base composition

In the stress test model, a bank is considered to have problems meeting the statutory solvency requirement if it fails to comply with the conditions specified above. The model calculations assume that the capital injections under Bank Rescue Package II are in the form of hybrid core capital of a quality that enables it to account for 35 per cent of the Tier 1 capital. In the example, it is assumed that the hybrid capital is not converted into share capital. However, Danish legislation allows banks to apply for the injected hybrid core capital to be convertible into share capital, cf. Box 3.

Sensitivity of the results in terms of the loss ratio
The estimates of the development in bank losses under different economic assumptions are subject to considerable uncertainty. Therefore, it can be relevant to decouple the relationship between economic development and bank losses in the stress test model and instead perform sensitivity calculations of the development in the banks' solvency subject to different loss ratios.

Chart 46 illustrates the development in the banks' excess capital adequacy, assuming that each bank in each of the three years of the scenario has a loss ratio of 1.5 and 2.5 per cent, respectively. The banks will suffer a loss each year under both assumptions, putting pressure on their excess capital adequacy. If the loss ratio is approximately 1.5 per cent, the banking sector will get relatively safely through the crisis. If, on the other hand, the loss ratio is approximately 2.5 per cent, which was the case during the crisis of the early 1990s, approximately half of the banks will need to further strengthen their capital base.

Banks' excess capital adequacy, sensitivity relative to the loss ratio – with bank rescue package II	chart 46

Note: Calculations are based on core and market earnings from the baseline scenario. Source: Danish Financial Supervisory Authority, banks' financial statements and own calculations.

As before, if the capital injections are convertible, the banks can also strengthen their capital base in this situation. If the banks have the option of converting capital injections, most of them will get through the scenario with an annual loss ratio of 2.5 per cent.

Summary of the stress test
Overall, it is the assessment of Danmarks Nationalbank that, with capital injections under Bank Rescue Package II, Danish banks are well positioned to withstand the expected economic developments.

Due to the economic uncertainty, the banks should carefully consider the option of being able to convert capital injections into share capital.

If the economic situation deteriorates markedly in relation to current expectations, further capital injections may be required.

Market assessment of the banks

Large, systematic fluctuations in equity prices
A systemic shock has rippled through the financial system, affecting the valuation of all assets. The focus has been on the system as such, rather than on individual banks. At the same time, risk aversion has been increasing rapidly.

Since the collapse of the investment bank Lehman Brothers in mid-September 2008, daily returns of the Nordic groups, positive as well as negative, have been significantly higher than experienced during the initial stages of the crisis from August 2007 to September 2008, cf. Chart 47. The covariation between equities has also been rising. For a period beginning in mid-September 2008, the equity-price covariation of the Nordic groups was over 80 per cent. Subsequently, the trend of covariation has been declining, but remains high. Fluctuations in market valuations of the Nordic groups continue to be strongly affected by the development of the financial system as a whole.

return and covariation for equity prices of the nordic groups	chart 47

Note: Highest and lowest returns are calculated on the basis of daily logarithm returns. The proportion of covariation is calculated as the proportion of the total variation attributable to the first principal component over a rolling window of 125 trading days, equivalent to about six months. Source: Bloomberg and own calculations.

Banks' CDS spreads remain high
A credit default swap, CDS, is a financial instrument used to hedge the credit risk e.g. on a bank. The price development, typically measured as interest-rate spreads, of a bank's CDS reflects the market's assessment of the probability that the bank in question will fail within a given period. The spread is affected not only by credit risk, but also by other factors. A key factor is the value of the bank's debt if it fails.

Lower expectations of the value of the debt will cause the spread to widen. The spread is also affected by investor requirements in terms of the risk/return ratio. Currently, the credit risk on a bank is impacted strongly by risks to the financial system as a whole. Thus, hedging achieved through a CDS on a bank also provides some measure of protection against the financial crisis as a whole. Other things being equal, increasing covariation between the banks, cf. Chart 47, will therefore prompt investors to demand a higher premium for hedging the credit risk on banks.

CDS spreads for financial companies widened considerably in early 2009, reaching a new record in mid-March, cf. Chart 48. During the last two months, the spreads have fallen back somewhat, but still remain high. The level may probably be attributed, in part, to market expectations that the value of the debt will be very low in case of failure compared with the value prior to the financial crisis.

CDS spread for selected nordic and european banking groups – 5-year senior debt	chart 48

Note: European banks are comprised of Deutsche Bank, ABN Amro, BNP Paribas, Barclays, HSBC, Royal Bank of Scotland and UBS, while Nordic banks are comprised of Danske Bank, DnB Nor, Nordea, SEB and Handelsbanken. The Chart shows the development in the range between the maximum and the minimum CDS spread, respectively, for each category. Source: Fitch.

The market expects further losses and a high risk of solvency problems
Overall, the banks have seen a significant decline in the "price/book value" since the onset of the crisis, cf. Chart 49. When the markets were at their lowest in mid-March, both global, Nordic and Danish banks were priced, on average, at around 40 per cent of their book value. However, this figure covers significant variation, and in mid-March a few banks were trading at below 15 per cent of their book value. Since then, the banks have seen significant increases in equity prices and global and Nordic banks are now trading at around 85 per cent of their book value, while Danish banks in groups 1 and 2 are trading slightly lower. Pricing somewhat below book value indicates that the market still expects further write-downs and losses over the coming years.

price/book value for Global, Nordic and danish banks, respectively	chart 49

Note: The selection of the 20 largest European and US banks complies with Box 2 of Financial stability 2008. Merrill Lynch, HBOS and Fortis were removed from the list of the 20 largest European and US banks on 15, 18, and 29 September, respectively. Roskilde Bank and Forstædernes Bank were removed from listed banks in groups 1 and 2 on 10 July and 15 September 2008, respectively, while Fionia was removed on 23 February 2009. Price/book value is calculated as a simple average for each of the three groups. Source: Bloomberg and financial statements.

On the basis of equity prices and accounting data, Danmarks Nationalbank has estimated a market-based risk measure, distance to insolvency, for the Nordic groups. The distance to insolvency measures the fluctuations in the market valuation of assets that can be tolerated by a group, while still meeting the minimum capital requirement of 8 per cent.

The distance to insolvency is a forward-looking risk measure, assessing the risk of solvency problems over a time horizon of one year. The distance to insolvency is measured by the number of standard deviations for the estimated market value of the banks' assets. The shorter the distance to insolvency, the more exposed a bank is, according to the market assessment. A distance to insolvency of zero can be interpreted as the market's assessment, in a risk-neutral world, that the probability of the group encountering solvency problems is 50 per cent. Chart 50 shows that, since October 2008, the market has priced some of the Nordic groups at an implied probability of solvency problems in excess of 50 per cent.

distance to insolvency for nordic groups	chart 50

Source: Financial statements, Bloomberg and own calculations.

It should be noted, however, that if risk aversion prevails, the assumption of risk neutrality always means that derived probabilities of solvency problems are higher than objective, but unobservable probabilities. In the current situation, the results of the model should thus be interpreted with caution, as they must, to a great extent, be assumed to reflect increased risk aversion in the market.

Summary of market assessment of the banks
The analysis of market assessment of the banks shows widespread uncertainty about the banks' capital adequacy and capital base. Taking into account that the risk aversion in the market is higher than usual and that the value of a bank's debt drops dramatically the moment it no longer meets the solvency requirement, the market assessment of the banks is believed to underpin the results of the stress test.

¹ 11 of the 14 banks analysed have indicated that they will apply for capital injections under Bank Rescue Package II. The three banks that have not indicated an intention to apply for capital are Nordea, Arbejdernes Landsbank and Ringkjøbing Landbobank.

Publication in PDF-format.

PC: Press the right mouse-button, choose "Save Link As", then choose where to save the file.

MAC: Hold down the mouse-button, choose "Save Link", then choose where to save the file.

Download
Acrobat Reader here: