Minutes of the meeting of the Council for Cooperation on Macroprudential Policy held on 1 October 2013

Swedish banks are fairly resilient but there are still risks in the financial system.

After several years of weak development of the real economy, both in Sweden and abroad, there are now signs that the economic situation is beginning to improve. These early signs of economic recovery and some political progress in Europe have contributed to a decline in stress on the financial markets. At the same time, recent developments on the financial markets have been marked by expectations of tapering in the United States. There is uncertainty over when the tapering will begin and what the scope will be, as well as the potential effects. This uncertainty has recently been reflected in rising long-term interest rates.

The council members observed that, despite a long period of weak real economic growth, the major Swedish banks have shown good results at the same time as their capital and short-term liquidity situations are relatively good. The banks' capital ratios have continued to strengthen and, in an international perspective, the Swedish banks have high capital ratios. One reason for the rising capital ratios is that the banks have moved their assets to exposures with lower risk, which require smaller capital buffers, such as mortgages. This is reflected in the leverage ratio, a measure that does not take risk weights into account, but instead puts the bank's capital in relation to its unweighted assets. In an international comparison the Swedish banks have a relatively low leverage ratio.

Over the past six months the Swedish banks are assessed to have had good access to funding. Moreover, their funding costs have remained stable, even though some volatility arose in connection with the discussion on tapering in the United States.

However, the council members noted that although recent developments have been favourable for the banks, there are still risks in the financial system. Swedish banks' high level of dependence on foreign market funding makes them vulnerable to unfavourable developments on the international financial markets. If the economic crisis in Europe were to worsen, it could have a rapid impact on the Swedish banks. The increase in household indebtedness also comprises a risk here. Over the past 12 months, the price of tenant-owned apartments has increased by almost 11 per cent, while growth in Swedish households' debts shows signs of increasing after slowing down for a period of time. The council returned to this question later on, see below.

The situation for life assurance companies has improved in that there has been a rise on stock markets and in long-term government bond rates in recent months. The average solvency ratio and traffic light ratio(1) have risen over the past three quarters. The reason why the traffic light ratio for life assurance companies is more stable than the solvency ratio is that it includes all of the company's risks, while the solvency ratio mainly refers to the interest-rate risk.

Other macroprudential policy issues

Finansinspektionen's work on implementing the ESRB's recommendations

Finansinspektionen gave an account of the work it has carried out and plans to carry out to meet the requirements in the European Systemic Risk Board's (ESRB) recommendations. The members of the council discussed the ESRB's current assessment of the recommendation on lending in foreign currencies. Finansinspektionen has revised its general guidelines on credit risk management in credit and securities institutions and the regulations are expected to come into force during 2014. This revision means that Sweden complies with the requirements in the recommendation.

With regard to the recommendation on intermediary objectives and instruments of macroprudential policy, Finansinspektionen already uses some of the instruments included. For example, the mortgage cap introduced in 2010, the requirement for a liquidity coverage ratio introduced in January 2013 and the risk weight floor for mortgages introduced in May 2013. Ensuring that macroprudential policy measures can be taken at a national level is one of the requirements made in the ESRB's recommendation on a mandate for macroprudential policy.

Finansinspektionen's work on CRR/CRD IV

Finansinspektionen provided information on its work on implementing the Basel III Accord. Following negotiations at EU level within and between the Council of Ministers, the European Parliament and the European Commission, the capital requirements directive (CRD IV) and capital requirements regulation (CRR) were adopted by the Council of Ministers and the European Parliament on 26 June 2013. The directive and the regulation will begin to apply from 1 January 2014. The Ministry of Finance has given indications that the capital requirements directive will be incorporated into Swedish legislation on 1 July 2014.

Finansinspektionen intends to issue, by 1 January, the necessary regulations that implement national options and transitional regulations in accordance with the CRR. Finansinspektionen is also planning to issue regulations on the parts of the CRD IV that the Government is expected to authorise Finansinspektionen to implement, including regulations regarding capital buffers, regulations on risk control and management and regulations on publication of information. The regulations on CRD IV shall in this case come into force at the same time as the accompanying legislation.

Finansinspektionen's introduction of a risk weight floor

In May this year Finansinspektionen introduced a risk weight floor of 15 per cent for Swedish mortgages. The floor locks in just over SEK 20 billion of Common Equity Tier 1 (CET1) capital in the Swedish banking system. The risk weight floor was introduced into Finansinspektionen's supervisory process immediately after the decision. Finansinspektionen has been able to observe that all of the bank groups have sufficient capital to cover the further capital requirement arising as a result of the risk weight floor. Finansinspektionen will regularly follow up and assess this measure, but now considers that insufficient time has passed since its introduction to be able to draw any firm conclusions regarding the results of the higher floor.

Liquidity risk in SEK

The Council discussed the major Swedish banks' liquidity buffers in Swedish krona.

The background to this is that the major banks currently publish the Liquidity Cover Ratio, LCR, in euro and dollar, which gives the investors an idea of the banks' buffers in these currencies. However, they do not publish the equivalent information to show liquidity risk in Swedish krona. The participants at the meeting noted that it is a strength that the major Swedish banks can show large liquidity reserves in foreign currencies, but noted at the same time that Swedish banks' liquidity buffers in SEK are low, both in relation to their outflows and in relation to the buffers in EUR and USD. The Riksbank commented that Swedish banks' resilience to potential liquidity problems in SEK could be too low and that this is a question that needs to be investigated further. Both authorities were able to note that it is desirable to have increased transparency regarding the banks' liquidity buffers.

The housing market – the analysis group's report: Risks linked to household indebtedness and amortisation behaviour

At its meeting on 19 February 2013, the Council for Cooperation on Macroprudential Policy decided to appoint a joint analysis group. The group's task was to analyse in greater detail the risks linked to household indebtedness; the risks to individual households and the risks to financial stability.

The analysis group's conclusions were presented at the meeting, including a summary that is attached to these minutes. Afterwards the Council for Cooperation on Macroprudential Policy discussed the analysis group's conclusions and the need for new measures to reduce the risks linked to household indebtedness.

The analysis group's conclusions

The analysis group has made a number of analyses that use micro data on both existing borrowers and new borrowers, on the basis of the observed loan-to-value ratios and amortisation behaviour. These analyses indicate that households generally have a good ability to meet mortgage payments and that the risks are small with regard to most individual households, even in stressed situations. They also indicate that the risk of the banks making major loan losses on mortgages is slight. On the other hand, the analyses of micro data show that there are some groups of households that are very vulnerable to a loss of income and/or a fall in housing prices. All in all, the analyses made solely using micro data indicate that the risks to the financial system are currently limited.

Since the mortgage cap was introduced in 2010, the amortisation periods have reduced with regard to unsecured loans and last mortgages for new borrowers and the trend increase in loan-to-value ratios appears to have been broken. However, the amortisation periods for first mortgages are very long, which contributes to keeping the aggregate debt ratio at a high level. At the same time, new calculations using the mortgage survey's panel data from 2011 and 2012 show that net amortisation (that is, gross amortisation minus new loans) was negative for new borrowers between autumn 2011 and 2012; the loans to households included in the data material for both years increased by 11 per cent, net, during this period. If this behaviour continues with regard to new borrowers, it could mean that today's rate of increase in households' total mortgage debts of around 5 per cent per year risks rising further. This would mean that the aggregate debt ratio probably continues to rise, all else being equal. With regard to the risk to financial stability, it is good that the amortisation periods for unsecured loans and last mortgages are falling. However, the long amortisation periods for first mortgages and the negative net amortisations could mean that risks increase in the coming period.

The analyses of macro data indicate that fundamental real demand and supply factors may explain a large share of the upturn in households' aggregate debt ratio in Sweden during the past 10-15 years; primarily lower interest rates, lower taxes and that more people own their own homes. This could indicate that the risk of a negative macroeconomic performance and of financial instability similar to that seen in a number of other countries, where the aggregate debt ratio has also increased substantially, is less pronounced in Sweden. At the same time, it is not possible to rule out the likelihood of more generous credit-granting norms also contributing to the upswing; the loan-to-value ratio for new loans increased from around 60 to 70 per cent between 2003 and 2010. The average loan-to-value ratio in the stock of mortgages is 65 per cent.

Although fundamental factors are thought to explain the upturn in the debt ratio, there is a risk that these factors could change, or that households have incorrect expectations of these factors, and that this in itself can lead to major problems for the macro economy and the financial system. We could also find ourselves in an economic situation where negative factors conspire and the general confidence in the Swedish economy declines. This could trigger a heavy fall in housing prices, for instance. In such a situation there is a risk that households will reduce their consumption considerably, which could lead to loan losses in the banks' lending to non-financial companies. A large fall in housing prices could have a serious effect on the banks' funding situation. If there were a development in the economy where negative factors conspired, the consequences of a high aggregate debt ratio and continued growth in the debt ratio could be more dramatic. It is difficult to fully capture this situation in the analyses, simulations and calculations made by the analysis group. Therefore, it is misleading to dismiss risks on the basis of fundamental factors being the reason for increase in debt and housing prices.

Any decisions on potential new measures to reduce the risks linked to household indebtedness have to be taken on the basis of an uncertain overall assessment. For reasons of caution, it is a good idea to consider further measures relatively quickly to manage the risks of the currently high (in both historical and international perspectives) aggregate debt ratio, although the basis for the decision, in the form of "definite proof" is limited. The starting point for these measures should then be that the potential social costs of not taking action today are assessed as so great that they outweigh the costs of increased regulation. At the same time, it is important that potential new measures are designed in such a way that when they are introduced they do not contribute to an economic downturn. It is also important to analyse further whether there is reason to take further measures to protect particularly vulnerable household groups, although the risks in lending to these groups, when seen in isolation do not appear to offer any major threat to financial stability.

The analysis group's ambition has been to contribute to increasing knowledge of the risks to individual households and financial stability related to household indebtedness in Sweden. But there is a need for further analysis. Firstly, it is important to be able to better assess the effects on private consumption of falling housing prices and to try to quantify what consequences this in turn may have for the banks' loan losses with regard to loans to non-financial companies. It is important that a continued analysis does not merely look at how debts relate to household incomes or to the market value of their housing, but to how they relate to households' total wealth. Here the lack of individual wealth data is a very large problem. Secondly, it is of central importance to try to assess the effects on the aggregate debt ratio in the long run as a result of demographic changes and urbanisation.

The Council's discussion of the analysis group's work

The members of the Council for Cooperation on Macroprudential Policy essentially shared the analysis group's conclusions and assessments regarding the need for continued analysis.

The Riksbank said that the work of the analysis group had contributed to greater knowledge with regard to household indebtedness, but that ultimately it was a question of subjective assessments with regard to the risks linked to household indebtedness. The Riksbank pointed out that there was uncertainty over the explanatory value of fundamental factors with regard to the upturn in Swedish housing prices and debt. Particularly bearing in mind that the fundamental factors that can explain the upturn in Sweden can also explain 70 per cent of the upturn in the households' aggregate debt ratio in the Netherlands. The Riksbank considered that it would be interesting to supplement the work of the analysis group, which is based on how the situation looks today, with studies of long-term scenarios, where the loan-to-value ratio and indebtedness continue to increase. It would also be interesting to capture various demographic aspects, as the analysis group itself proposes. Such scenarios can mean that the conclusions regarding the need for increased amortisation and the importance of loan-to-value ratios not being too high may change.

Finansinspektionen also considered that the work of the analysis group had provided an important contribution to the understanding of risks linked to household indebtedness and also shared the analysis group's and the Riksbank's assessments regarding the need for continued analysis. With regard to the long-term demographic developments, Finansinspektionen considered that it is, for instance, important to look at the risks that arise if in the future highly-indebted households, because of low amortisation, want to sell their housing when they retire. Finansinspektionen also pointed out that there is considerable uncertainty and that it is ultimately a question of subjective assessments of the risks linked to household indebtedness.

The Council's discussion of the need for new measures

With regard to measures to reduce the risk of household indebtedness, the Riksbank saw a need to implement further measures in the near term. The Riksbank considered that one possible measure is to further raise the risk weights for mortgages. The Riksbank's representatives at the meeting advocated raising the risk weight floor in stages up to 35 per cent. In addition, the Riksbank considered it would be desirable to have a stronger amortisation culture so that households pay off their loans to a greater extent. It would be preferable for the banks and households to resolve the amortisation question on their own. Moreover, the Riksbank pointed out that the countercyclical capital buffers would probably have been activated if the regulations had been in place today. The Riksbank also said that there is a need for the banks to review and harmonise their standard calculations of the loan scope for new mortgage applicants, what are known as discretionary income calculations, so that they better reflect the households' actual economic situation.

Finansinspektionen emphasised the measures it has already taken in the form of a mortgage cap and a risk weight floor, as well as the coming proposal for a requirement that the banks give mortgage customers a proposal for an amortisation plan. Finansinspektionen also pointed out that there is already a review of discretionary income calculations underway in the field of supervision. With regard to the Riksbank's proposal to raise the risk weight floor further, Finansinspektionen pointed out that the risk weight floor already triples the risk weights on mortgages so that they are now above the average for European banks using the IRC method(2) and that there is a need to first evaluate the higher risk weight floor introduced last spring. Finansinspektionen also pointed out that it is not primarily the supply of credit that is the problem, but the demand for credit. And here other measures than a higher risk weight floor may be more effective; amortisation plans and amortisation requirements being two examples.

Both authorities were agreed that, given the uncertainty, it is important to gradually implement several different types of measures. Both the Riksbank and Finansinspektionen noted further that there is a need for a better information base with regard to household wealth. It is not satisfactory that since wealth tax was abolished in 2007, there is no individual data on household wealth and indebtedness for the population as a whole. The Riksbank said that there is a need for more lasting and comprehensive data than the sample from the mortgage survey gathered once a year.

Next meeting

The Government announced at the end of August that Finansinspektionen will have main responsibility for the instruments for financial stability and that a financial stability council will be established, consisting of the Riksbank, Finansinspektionen and the Swedish National Debt Office and led by the Minister of Finance.

The Council for Cooperation on Macroprudential Policy agreed that they would together continue to prepare questions to be managed either within the new financial stability council or at the next meeting of this council.

 1) The traffic light model calculates a capital buffer on the basis of the actual value of both assets and liabilities. After that, stress tests give rise to an overall hypothetical "capital requirement" that is made of the company. The financial part of the stress test in the traffic light model measures the insurance companies' sensitivity to large changes in prices on the financial markets. The insurance risks are subjected to stress by changing the conditions for the assumptions made when calculating the technical provisions. Companies are also required to include an increase in their operating costs in the calculation. The traffic light ratio is the capital buffer divided by the capital requirement. The analysis covers 32 life assurance companies. Sources: Finansinspektionen and Reuters EcoWin

 2) The IRC method means that the bank has the permission of the financial supervisory authority to use the internal risk classification method to calculate risk weights for its credit exposures.