Is the Riksbank right about Swedish mortgages posing a threat to financial stability?

New Ekonomistas post (in Swedish). Here is an English translation:

The Riksbank is fighting an uphill battle to ex post justify the monetary policy that, with too high a policy rate, has led to an inflation far below the target and an unemployment far above a reasonable long-run sustainable rate. It is not going very well. The Riksbank has, for instance, stated that (1) households’ expectations about future mortgage rates are too low. It has also stated that (2) higher debt has led to a larger fall in consumption and rise in unemployment, in the countries that were hit by a fall in housing prices during the recent crisis. But it is difficult to find any foundation for these two statements, which can be seen in this and this post. Furthermore, even if the statements were true, it does not follow that tighter monetary policy is the best response or that it would even improve the situation. For instance, a lower inflation than expected has led till a higher debt burden (not lower). The Riksbank has also stated that (3) a fall in housing prices might make banks’ financing of the mortgages more difficult and this way threaten financial stability. Is there any foundation for this third statement? 

Swedish banks to a large extent finance their mortgages by emitting so-called covered bonds, with the banks’ mortgages as collateral. About a third of the covered bonds are bought by foreign investors. The Riksbank is worried that, if housing prices would fall, foreign investors might start doubting that Swedish mortgages are safe and therefore may no longer want to invest in the banks’ covered bonds. Then, according to the Riksbank, the banks’ financing and thereby the financial stability would be threatened (see, for instance, this recent speech by Executive Board member Kerstin af Jochnick).

The risk for losses from the Swedish mortgages are small, however, according to FI (Finansinspektionen, the Swedish Financial Supervisory Authority):

In FI’s view the banks in general take higher interest rates into consideration and make sure that households are able to amortise their mortgages. This means that the banks in general ensure that households have a satisfactory repayment ability and resilience to any adverse events. FI therefore assesses the risk of direct credit losses for the banks to be small. (Finansinspektionen’s Mortgage Market Report 2013, p. 13.)

A most important fact in this context is that Swedish mortgages are “full recourse,” that is, borrowers are personally responsible for the repayment of the debt, regardless of the value of the housing. This means that in practice the main “real” collateral for Swedish mortgages is not the value of the housing but the borrowers’ repayment capacity, including the borrowers’ “human capital,” the present value of the borrowers’ disposable income (with a deduction for necessary minimum expenses). The present value of 25 years of disposable income is about 22 times current disposable income, under reasonable assumptions. A fall in housing prices does not directly affect the large repayment capacity from the borrowers’ human capital. Not even during the great crisis of the 1990s in Sweden, with a large fall in housing prices and a large increase in unemployment, were there any credit losses to speak of on mortgages. The large credit losses came from commercial property.

Of the loan losses in the four major bank groups at the height of the bank crisis in 1992, only 6 per cent (about SEK 4.5 billion) came from the household sector. (Sveriges Riksbank Financial Market Report 1/1998, p. 15-16.)

If foreign investors – in spite of this – would start having doubts about the safety of Swedish mortgages and not want to invest the Swedish covered bonds, this would thereby be a liquidity problem for the Swedish banks, not a solidity problem. The concern among foreign investors about not getting their money back would thus be unfounded. It is precisely in such situations that it is justified to give liquidity support of the kind that the Riksbank and the National Debt Office gave during the latest crisis, when the Riksbank lent money to the banks’ against various collateral, not least covered bonds, and the National Debt Office exchanged government bonds for covered bonds. Another powerful instrument in such a situation is a government guarantee, according to which the government guarantees the covered bonds. (Since the Swedish mortgages are nominated in SEK, it is in this case liquidity support in SEK that is relevant, not liquidity support in foreign exchange. (The need for liquidity support in foreign exchange and the size of the Riksbank’s foreign exchange reserves have been examined in a Government Commission of Inquiry by professor Harry Flam, this post (Swedish only)).

To act when solid banks have financing problems is a classic task for a central bank. That the central bank in this way acts as a “lender of last resort,” that is, supplies loans against collateral in the form of solid but temporarily illiquid assets, can be seen as a response to a market failure. Market failures are often due to information problems. In this case, it is a matter of foreign investors having doubts about Swedish banks in spite of credit risks in mortgages being small.

But information problems should of course also be prevented before they occur. With better information and knowledge about the Swedish mortgage market, the risks that investors suddenly start having doubts about them should reasonably be reduced.  (For instance, perhaps all U.S. investors in Swedish covered bonds are not aware of the big differences between mortgages in Sweden, with personal responsibility for repayment (“full recourse”) and thorough credit assessments of borrowers, and mortgages in the U.S, with in many states the housing being the only collateral (“no recourse”) and credit assessments of borrowers previously in many cases severely deficient.) For anyone worried about investors suddenly having doubts, an information campaign would be appropriate. Naturally, the banks have the main responsibility to inform about their activities, but the message should be disseminated more effectively if responsible authorities would give the same picture.

The strange thing is that when the discussion is about financial stability, the Riksbank mainly signs up to the view that the risks are small that banks shall be hit by large losses from mortgages: “[T]he risks of the banks being affected by severe loan losses directly from mortgage lending may seem small at present.” (Financial Stability Report 2/2013, p. 13.) The scenario where lending to households threaten the financial stability instead shows up mainly when the Riksbank tries to justify its monetary policy. The interpretation that the Riksbank ex post is looking for ways of justifying its previously chosen policy, rather than ex ante doing analysis to provide better policy decisions, is unfortunately close at hand.

The conclusion is that neither does this argument for a tight monetary policy stand up to scrutiny. If Swedish banks would get problems with the financing of their mortgages, this would be a liquidity problem, a problem that can be handled effectively with the above-mentioned existing policy instruments.