Journal Issue
Share
Article

Denmark: Selected Issues

Author(s):
International Monetary Fund. European Dept.
Published Date:
January 2013
Share
  • ShareShare
Show Summary Details

Assessing Government Contingent Liabilities From the Financial Sector in Denmark1

Denmark has introduced a number of measures to support the banking system since 2008, including a temporary full guarantee of bank debt, individual bank guarantees, and capital injections. While the associated fiscal revenues have been higher than the costs so far, these actions have enhanced expectations of future bailouts. Indeed, the market perception of an implicit guarantee on systemically important banks persists. Further capital strengthening, especially for the largest financial institutions and setting up a resolution framework for systemic banks would enhance both financial stability and fiscal sustainability.

A. Introduction

1. At 47 percent of GDP, Denmark’s public debt is low by international comparison, and fiscal consolidation is proceeding. Nevertheless, experience from other countries in the wake of the global financial crisis has shown that even strong public finance positions can deteriorate quickly due to explicit or implicit contingent liabilities stemming from the financial sector, resulting in a loss of market confidence. In Ireland, for example, sovereign spreads started to increase after the government extended a full guarantee to the banking system in 2008, and public debt increased from less than 25 percent of GDP in 2007 to over 100 percent of GDP in 2011, as public funds were used to shore up the national financial systems.

2. Since the eruption of the global financial crisis, Denmark has implemented several actions to support and strengthen the banking system. The objective of this paper it to review those measures, assess their fiscal costs, and, more importantly, gauge the potential impact of explicit and implicit contingent liabilities stemming from the financial sector on the fiscal accounts going forward.

B. Recent Government Support to the Financial Sector

3. A number of bank stabilization packages have been enacted in Denmark since 2008.

4. Bank rescue package 1 (October 2008) introduced a government guarantee on all claims of depositors and other unsecured creditors in banking institutions in Denmark (excluding covered bonds). The full guarantee expired on September 30, 2010 and was replaced in 2009 by a scheme allowing banks to apply for individual guarantees, until end-2010 under Bank Rescue Package 2 (see below).

5. Bank rescue package 2 (February 2009) allowed credit institutions to apply until end-June 2009 for state-funded capital injections. A number of banks received hybrid capital instruments that could be redeemed after three years. Forty-three credit institutions received a total of DKR 46 billion in the form of hybrid core capital, with interest rates between approximately 9 and 11.25 percent depending on the individual institution’s risk. DKR 34.8 billion (12 percent of 2011 tier 1 capital) remain on banks’ balance sheets. The package introduced also a state guarantee program on individual bank non-subordinated unsecured bonds, with maturities of up to three years. Several banks joined the scheme—agreeing to restrictions in their activities (e.g., caps on remuneration and dividend payouts), and paying a fee which varied from bank to bank. Individual government guarantees ran for up to three years. Guaranteed bonds of almost DKR 190 billion have been issued since 2009. DKR 80 billion (4.4 percent of GDP) were still outstanding in October 2012. This debt matures in 2012–13.

Security Issuance by Danish Banks

(Billion Krone)

Sources: National authorities and Fund staff calculations.

6. Bank rescue package 3 (October 2010) established a resolution framework for institutions, also envisaging creditor bail-in. This may have reduced the perception of an implicit government guarantee, especially for small banks to which the resolution framework is readily applicable. The framework was first used in February 2011 for the resolution of the 9th largest bank, Amagerbanken, resulting in losses for unsecured creditors.

7. Bank rescue package 4 (August 2011) introduced two different models to create greater incentives for sound banks to take over, in full or in part, the activities of a distressed bank before resolution under Bank rescue package 3 becomes necessary. Specifically, under Model 1, a sound bank willing to take over the entire bank in distress may obtain compensation from the Guarantee Fund for Depositors to cover depositors, and from the Danish government if the distressed bank had received an individual government guarantee. Under Model 2, the Financial Stability Company takes over all parts of a distressed bank, except capital and subordinated debt, and transfers the sound part of the bank to another sound bank. Bank rescue package 4 (model 2) was first used in October 2011, when Max Bank became distressed. The Bank rescue package 4 introduced also the possibility for individual government guarantee with maturity up to three years in connection with mergers before end-2013.

8. Bank rescue package 5 (March 2012) allowed FIH Erhvervsbank to transfer property exposures of DKR 17 billion (1 percent of GDP) to a new company, which will be acquired by the Financial Stability Company, that will manage the loan portfolio.

9. Overall, so far the expenses of government intervention in support of the banking sector have been lower than the associated revenues, with the net surplus from the support totaling nearly DKR 7 billion (0.4 percent of GDP).

Fiscal Costs of Government Support to the Banking Sector, 2009–2012(Billions of DKK)
Government sponsored hybrid bonds in bank’s capitalIndividual bank guarantees
Interest revenues9.5Revenues from fees3.6
Losses2.9Losses3.3
Net fiscal cost−6.6Net fiscal cost−0.3
Sources: Denmark’s Ministry of Business and Growth and Fund staff calculations.

C. Assessing the Market Value of Government Guarantees for Systemic Banks

10. While the Danish authorities have taken steps to reduce the expectation of an implicit government guarantee on the banking sector, notably with the enactment and implementation of bank rescue package 3, the perception of an implicit guarantee may persist for large banks, considered too big to fail. This section tries to assess whether this is the case. The analysis is conducted using the contingent claim approach, originally developed by Gray, Merton, and Bodix (2007), starting from Merton’s (1974) seminal work. First, the section discusses the concept of Fair Value CDS spreads, which is an indicator of what CDS spreads would be if the markets were ruling out the possibility of any government support. Then Fair Value CDS spreads are compared with market CDS to derive the market value of government guarantees for systemic banks.

11. The government guarantee schemes and the state funded capital injections are likely to have enhanced the perception of the existence of an implicit government guarantee on the financial sector. Indeed, the CDS spreads of Danske—the largest Danish bank—fell by almost 90 basis points (bps) within two weeks after the full government guarantee was introduced in early October 2008. Concurrently, sovereign CDSs spreads jumped up by 40 bps, and remained elevated compared to historical values until early 2009, suggesting that banks’ risk had spilled over sovereign risks. At end 2008 and early 2009, Danske’s and sovereign CDS spreads were in fact very close, and the bank’s CDS spreads climbed again after the expiration of the full guarantee in September 2010.

Danske and Danish Sovereign CDS Spreads

(Basis points)

Sources: Bloomberg LP., Datastream, and Fund staff calculations.

12. Market expectations about the value of the government implicit guarantee stemming from the banking sector can be gauged using the contingent claim analysis (Gray, Merton, and Bodix, 2007).2 This approach is based on the assumption that market indicators (such as equity prices and CDS spreads) contain information about a firm value and viability. More specifically, in this framework a firm’s equity can be valued as an option on the asset value of the firm, and default occurs when the value of the firm’s assets is insufficient to allow the firm to meet its contractual obligations. In turn, the unobservable value of the firm is inferred from equity prices, together with the company’s capital and debt structure (see Appendix for a technical discussion). Within this framework, it is possible to construct an indicator of firm riskiness, called Fair Value CDS spreads (FVCDS) using equity prices. Those can then be compared with CDS spreads from credit markets—which reflect the assessment of credit risks by credit investors, factoring in expectations about government support. Since recent government support to the banking sector has primarily benefited credit investors rather than equity investors, the difference between FVCDS and market CDS spreads can be interpreted as market expectation about government support, i.e., market assessment of the government guarantee.

13. According to this metric3, market expectations about the government’s guarantee on Danske increased sharply in the Fall of 2008, as the Danish government introduced a full guarantee on bank debt, and fell only toward the end of the full-guarantee period. In recent months, as strains in the European financial markets intensified, FVCDS spreads have picked up significantly and remain elevated, while market CDS spreads have increased more gradually, suggesting market expectation of an implicit government guarantee on this “too big to fail” banking institution.

14. For Nordea—the Swedish banks with an important presence in Denmark4—, the gap between FVCDS spreads and market spreads—the “too-big-too-fail” guarantee premium—widened first at the peak of the global financial crisis in late 2008-early 2009 and, more dramatically, in 2012. The recent sharp rise in the FVCDS probably reflects market concerns about Nordea’s exposure to core European markets.

Danske’s Fair Value and Market CDS Spreads

(Basis points)

Sources: Bloomberg, Moody’s KMV, and Fund staff calculations.

Nordea’s Fair Value and Market CDS Spreads

(Basis points)

15. FVCDS and market CDS can be used in a simple zero-coupon model to quantify the market assessment of the value of the implicit government guarantee. Specifically, for one unit of zero coupon debt of duration t, the market evaluation of the implicit government guarantee (GG) can be obtained as follows:

Where r is the assumed risk-free rate, CDS is the market CDS spread, and premium is the differential between FVCDS and market CDS spreads (the too-big-to-fail guarantee premium). Intuitively, the first term above represents the present discounted value of one unit of debt under the government guarantee, and the second term is the present discounted value of one unit of debt without any government guarantee.

16. The results under the contingent claim approach suggest that markets expected in late 2012 that the government would cover about 20 percent of the value of the defaulting debt of the Danske and Nordea groups—two “too big to fail” institutions and 80 percent of the loss would be borne by bank creditors. The estimated contingent liability for Nordea and Danske however, lower than for other large banks in the euro area, possibly reflecting higher capital levels in the Nordic banks.

Market Assessment of Government Guarantee per Unit of Bank Debt 1/

Sources: Moody’s KMV and Fund staff calculations.

1/ Assumes risk-free rate at 1 percent and 5 year debt duration.

Market Assessment of Government Guarantee per Unit of Bank Debt 1/

Sources: Moody’s KMV and Fund staff calculations.

1/ Assumes risk-free rate at 1 percent and 5 year debt duration.

17. This approach has the advantage of providing an estimate on the market assessment of government guarantee based on a well-developed analytical framework and easily available market variables. Results are not very sensitive to the assumptions about key parameters (e.g., the risk free rate). However, during periods of high market volatility, CDS and FVCDS spreads—and the associated assessment of a government guarantee—may exhibit temporary and large movements. Furthermore, the approach is based on the assumption that markets are able to evaluate and price a firm riskiness well. In practice, markets may actually be wrong about that, as well as about the true willingness of the government to support banks.

D. Estimating Expected Losses

18. The approach to assess the market expectations about implicit government guarantee described above cannot be used for small and medium sized for which liquid credit market CDS spreads are not available. The contingent claim analysis can be applied, though, to estimate bank expected losses. Specifically, the contingent claim analysis can be used to assess the default probability, where default is defined to occur when the value of the bank’s assets falls below the value of its contractual obligations (see Appendix for details). Given the estimated probability of default, expected bank losses can be computed as follows:

Expected loss= Expected probability of default*Loss Given Default*Bank debt

The expected probability of default within one year estimated using the contingent claim analysis framework has increased dramatically for small banks (Group 3) in recent months, and is markedly higher for this bank group than for the others, even though default probabilities vary significantly among Group 3 banks.5

19. Danske’s expected losses within one year are about 3.2 percent of GDP, even though the expected probability of default is low, given Danske’s large size. Expected losses in Jyske and Sydbank are low at 0.3 percent of GDP, due to the two banks’ low expected probability of default. For group 2 banks, expected losses are also small (about 0.1 Percent of GDP). Total expected losses within one year in group 3 banks are low due to their small size.6

Expected Losses Within 1Yearby Bank Group 1/

(Percent of GDP)

Sources: Moody’s KMV and Fund staff calculations.

1/ Assumed loss given default is 60 percent.

2/ Sum of expected losses in Spar Nord, Vestjysk, Alm Brand, Ringkjobing Lando bank.

3/ Sum of expected losses in Sparekassen Lolland, Sparekassen Himmerland, Nordjyske Bank, Sparekassen Faaborg, Diba Bank, Sparekassen Hobro, Svendborg Sparekasse, Totalbanken, Salling Bank, Kreditbanken.

Expected Losses Within 1 Year: Danske 1/

(Percent of GDP)

E. The Impact of Banking Sector Contingent Liabilities on Debt Sustainability

20. This section presents a number of scenarios on the potential impact of contingent liabilities from the financial sector on debt sustainability (Figure 1).

Figure 1.Denmark: Impact of Banking Sector Continget Liabilities on Public Debt Sustainability:

(Percent of GDP) 1/

Citation: 2013, 23; 10.5089/9781475527322.002.A005

Sources: International Monetary Fund, country desk data, and Fund staff calculatons.

1/ Shaded areas represent actual data.

21. Outstanding explicit government guarantees on individual bank debt are unlikely to pose a threat to debt sustainability, given Denmark’s relatively low debt level. Even in the most pessimistic scenario where all guarantees are called in 2013, debt would jump to 52 percent of GDP and fall to less than 50 percent of GDP in 2017. Expected losses from small and medium sized banks 3 banks are assessed to have a small impact on fiscal sustainability, assuming public support.

22. The estimated government contingent liability from Danske is, instead, quite sizable, and could bring the public debt to GDP ratio up to over 70 percent by 2017, assuming that, in line with market expectations, 20 percent of Danske’s debt (excluding subordinated debt and covered bonds) were taken by the government in the event of bank default. If Nordea were to default, part of the fiscal cost could fall on Denmark. Assuming that the Danish government would cover 20 percent of Nordea Danmark’s liabilities, Denmark’s public debt could increase to about 53 percent in 2017.

F. Policy Implications

23. Since 2008 Denmark has introduced a number of measures to support the banking system, including a temporary full guarantee of bank debt, an individual bank guarantee and recapitalization bonds. While the associated fiscal revenues have been higher than the costs so far, they have enhanced expectations of future bailouts. Going forward, the provision of new government guarantees on bank debt could create expectations of further bailouts, distort market pricing mechanisms, may induce moral hazard and delay needed bank restructuring. Bank funding strains could instead be addressed through central bank refinancing facilities if the bank is solvent and has adequate collateral. If emergency lending assistance is provided, it should support rather than delay necessary restructuring.

24. Systemic institutions, such as Danske, create sizable contingent liabilities for the public sector. Therefore, further capital strengthening especially for the largest financial institutions, through rights issuance and earnings retention, would enhance both financial stability and fiscal sustainability. Other important steps to that end could include setting up a resolution framework for systemic institutions as well as establishing a specific agreement with other Nordic countries on burden sharing in the event that a cross-border bank needs to be supported or resolved.

25. Banks with a sizable share of government sponsored hybrid bonds in their balance sheet could develop plans to raise alternative forms of core capital. Indeed, these expensive instruments hinder their profitability and in view of forthcoming Basel III regulations. Addressing vulnerabilities in a number of small banks with high estimated default probabilities, including through resolution, would foster financial stability and minimize potential fiscal costs.

Appendix: Contingent Claim Analysis Methodology

The contingent-claims approach (CCA) provides a methodology to combine a firm’s balance sheet information with widely used finance and risk management tools to construct marked-to-market measures of balance sheets items that reflect underlying risk. This methodology have been developed and made operational by Moody’s in their Moody’s KMV model.

The overall level of risk facing a firm balance sheet, i.e., the default probability, depends on the value of total assets and their volatility. Those variables, however, typically cannot be measured easily, given that many assets (e.g., bank loans) are not traded. In contrast, liability items (e.g., equities) are often traded, and thus can be used to assess the value and volatility of a firm’s assets. Merton’s (1974) key insight in option pricing theory was that liabilities are contingent claims on total assets, with each liability having a different priority and maturity structure. The most junior liability on the balance sheet can be valued as an implicit call option on total assets. Indeed, the limited liability feature of equity means that the equity holders have the right, but not the obligation, to pay off the debt holders and take over the remaining assets of the firm. Hence, equity is the same as a call option on the firm’s assets with a strike price equal to the book value of the firm’s liabilities.

In this framework the default probability of a firm is determined in three steps:

1. Estimate asset value and volatility using the market value and volatility of equity and the book value of liabilities. If the market price of equity is available, the market value and volatility of assets can be determined directly using an options pricing based approach, which recognizes equity as a call option on the underlying assets of the firm.

2. Calculate the distance-to-default from the asset value and asset volatility (estimated in the first step) and the book value of liabilities. The distance from default measures how many standard deviations the firm is from having to default on its debt, where default is postulated to occur when the value of the assets falls below the value of liabilities.1

3. Calculate the default probability: The default probability is determined from the distance-to-default and historically observed default rates for given levels of distance-to-default.

The methodology to compute fair value CDS spreads 2 takes information from equity prices and creates an estimate of what the spread on debt would be using the characteristics of the debt and aggregate information on comparable firms.

References

Prepared by Edda Zoli.

A thorough illustration and discussion of the contingent claim analysis is presented in Gray and Malone (2008).

FVCDS data are computed using Moody’s KMV’s methodology.

Nordea Bank Danmark is the second largest bank in Denmark, accounting for 15 percent of the Danish banking system assets.

Group 2 includes banks with working capital between DKR 12 and 65 billion; group 3 comprises banks with working capital between DKR 250 million and 12 billion. Danske’s assets represent almost 60 percent of the Danish banking system assets; Jyske and Sydabank account together for about 10 percent; group 2 banks included in the sample (Spar Nord, Vestjysk, Alm Brand, Ringkjobing Lando bank) represent almost 4 percent of the banking system; and group 3 banks in the sample (Sparekassen Lolland, Sparekassen Himmerland, Nordjyske Bank, Sparekassen Faaborg, Diba Bank, Sparekassen Hobro, Svendborg Sparekasse, Totalbanken, Salling Bank, Kreditbanken) account together for nearly 1.5 percent of banking system assets. The selection of banks to be included in the sample was driven by data availability

The loss given default is very conservatively assumed to be 60 percent, even though in recent episodes of bank resolution in Denmark haircuts have been much smaller. Consistent with the Moody’s KMV contingent claim methodology, default is assumed to occur when the value of a bank’s assets fall below the value of total short-term debt plus half of long-term debt.

In Moody’s KMV framework, default is assumed to occur when the value of a firm’s assets fall below the value of its short-term debt plus half of long-term debt, consistent with empirical evidence on a large number of default episodes and bankruptcies.

Fair-value, or mark-to-market, accounting, refers to the accounting standard of assigning a value to a position held in a financial instrument based on the current fair market price for the instrument or similar instruments.

Other Resources Citing This Publication