Financial Crisis - Experiences from Sweden

  • Date:
  • Speaker: Deputy Governor Lars Heikensten
  • Place: Korea
Let me begin with a thank you to the organisers of this seminar. I am honoured to have the opportunity to share with you some experiences from the Swedish financial crisis in the early 1990s.


1. Introduction

An IMF study from 1996 showed that in the past fifteen years, 132 of the Fund’s 182 member countries had experienced macroeconomic and/or financial crises. Today the figure is even higher. Crises are a recurring phenomenon. In spite of every effort to prevent them, I fear that crises will continue to occur. We should therefore use the opportunities to learn from each other, both in order to avoid making the same mistakes and to find the appropriate correctives sooner.

Needless to say, there are no single “best solutions”. Crisis situations are always specific in some respects, not least with regard to their structural and political environment. Had they not been specific, they would probably never have occurred because they would have been “spotted” in advance. Having said this, however, there are a number of globally accepted “truths” about what constitutes a good response to crises.

Discussions around crisis situations are usually divided into two parts: crisis prevention and crisis management. In this presentation I will be focusing on the latter. To understand the handling of a crisis, it is useful to know which factors generated the problems. Hence, my narrative will include the events that led up to the crisis in Sweden. I will also comment on the developments in Asia, with some emphasis on Korea. Having sketched this background, I will turn to the actual handling of crises, explaining why we found certain types of solutions useful in Sweden—leaving it to you to consider whether the same might apply to some extent to your country.


2. Background

The macroeconomy

First, let me say a few words about the macroeconomic background to the crisis in Sweden. There are both similarities and differences between Sweden and Asia. In the following I will touch on some of them. It should also be remembered that substantial differences also exist between the Asian countries.

One common factor is the fixed exchange rate. In Sweden this led to high levels of foreign borrowing, particularly in the private sector. In addition it made it difficult to counter the gradual overheating of the economy by applying contractive monetary measures—raising interest rates tended to increase capital inflows which, to the extent they were monetised, further contributed to the overheating.

Following the deregulation of the domestic financial market, the crisis in Sweden was also preceded by a rapid expansion of credit; in the course of five years, private borrowing grew from 85 to 135 % of GDP. The credit expansion coincided with a protracted boom and a significant proportion of the borrowed money was spent on speculation, both in real estate and in financial assets, such as equity. The real estate speculation in Sweden culminated in a bubble that burst in 1990-91. Shielded by the economic upswing, there had been many misdirected and overly optimistic investments in various industrial and other projects.

In these aspects I find similarities between Sweden in the late 1980s and the Asian countries during the 1990s. A feature common to Sweden and Korea was, moreover, that the domestic financial market was liberalised long before the external capital outflows. This meant that overheating of the domestic economy could not be adequately balanced by capital outflows for investment abroad.

But there are also some differences. Other factors that contributed to Sweden's acute crisis were a relatively deep international downturn and high interest rates in our part of the world in the wake of German reunion. On top of this, our fiscal position was more vulnerable, entering the 1990s with a public debt-to-GDP ratio of around 40 per cent. Sweden's external borrowing ratio was also high—some 30 per cent of GDP. Besides reflecting the public budget deficit, this had to do with an extremely low private savings ratio. In some years this ratio was actually negative. These weaknesses made Sweden highly vulnerable to shifts in confidence in our economy.


The financial sector

There were also underlying weaknesses in financial systems. Let me now turn to these.

The financial systems had not adapted to the change from a sheltered environment to a deregulated and much more competitive situation. In the earlier environment, banks had had little incentive to develop a sophisticated credit culture, involving comprehensive evaluations of debtors, projects and collateral. (One important similarity between Sweden and Korea was the overoptimistic reliance on collateral and other guarantees, rather than on the viability of the loan project itself). When deregulation was a fact, bank managers saw a chance of gaining market share and increasing revenues by taking new risks, but did not clearly see the dangers in a lack of the necessary expertise.

One of the major faults in credit decisions in Sweden was that banks lent to customers, projects and geographical areas of which they did not have sufficient knowledge. Another major mistake was that banks accepted high-risk concentrations, not just to individual companies but also to economic sectors, primarily real estate, and to geographical areas. At the peak of the crisis, bank loans to real estate, or collateralised by real estate, accounted for more than 60 per cent of all loan losses.

A further weakness was the inability of the banks, as well as of the authorities, to register the growing prevalence of indirect concentration risks. For instance, bank borrowers demanded loans denominated in foreign currencies. Interest rates on these loans were lower than on krona loans and as the krona was fixed within a fairly narrow band, there was presumed to be no exchange risk. The banks borrowed the foreign currencies abroad and on-lent them to customers, avoiding open exchange risk exposures. However, when the krona was attacked and the fixed exchange rate subsequently had to be abandoned, banks were doubly affected. Foreign counterparties, perceiving the banks' weakened position, reduced or withdrew credit lines. This forced the banks to demand either early repayment from customers or loan conversion into domestic currency. This led, in turn, to sharply increased loan losses because, at short notice, borrowers were unable to raise the larger amounts in krona that were needed to repay the foreign currency loans, which the krona's depreciation had made much more expensive.

Neither had the supervisory authority, the Bank Inspectorate, adjusted. It continued its traditional formal supervision, ensuring that reports, permits et cetera were formally correct, rather than supervising the actual risks. The authority was also somewhat complacent about risks (to be frank, the Central Bank and the Ministry of Finance should share the blame for this) since Sweden had not experienced a financial crisis since the 1930s. Lacking relevant crisis experience, the authorities did not perceive that the 1980s could pave the way to a new financial crisis. The decade was, after all, characterised by sustained economic growth, rising bank profits and extremely low loan losses.

The financial weaknesses I have described so far are quite similar in Sweden and Korea. However, some differences should also be noted:

- In Sweden, banks had built up good capital reserves. Before the onset of the crisis, all the banks fulfilled or surpassed the 8 per cent Basle requirement for capital. So, although non-performing and other impaired loans amounted to 15 per cent of GDP in Sweden (a figure that is rather similar to recent estimates for Korea), the original cost to the government of rescuing the banking system amounted to only some 4 per cent of GDP; since then, moreover, most of this has in fact been recovered.

- Sweden applied restrictive rules to central bank involvement in the banks. Only liquidity support was provided, inter alia in the form of foreign currency deposits to stem the outflow from the banks and to reduce the risk of a credit crunch. This did not involve the Riksbank in any risks because the State guarantee for banks' repayments also applied to us as a lender.


3. Triggers

Now, let me turn to the actual crisis. It is well known that the event (or events) triggering a crisis is often different from the crisis' underlying causes. In the absence of a triggering factor, countries with substantial underlying problems from time to time have also been able to avoid an acute crisis. The event triggering a crisis is sometimes rather dramatic, such as the assassination of a politician—as in the Mexican case—or a sudden loss of confidence, e.g. due to the presentation of a disappointingly weak fiscal budget. Opinions may differ about what were the most important “triggers” of the acute crises we are discussing today. Again, I see both differences and similarities between Sweden and the Asian countries.

In our case, a policy shift to the low-inflation regime in the rest of Europe, combined with structural changes and a more hostile international environment (with higher interest rates and lower growth, as I mentioned before), made the existing situation unsustainable and triggered the crisis.

In the early 1990s the government refrained from accommodating the high costs and inflationary pressures that had been building up in the 1980s. Under similar circumstances earlier in the 1970s and 1980s we had devalued several times. Now there was broad agreement that we should stick to the fixed exchange rate in order to bring inflation down; in this respect, the policy was indeed successful: inflation fell from 8-10 per cent to 2-3 per cent in less than two years. One effect of this, however, was that real interest rates shot up. Structural reforms were also implemented, particularly in the tax system but also in housing and other areas. One aim of these reforms was to create more market-related incentives for savings and to reduce subsidies. While these changes were long overdue, they now came at the worst possible time from a macroeconomic point of view.

In this context, it is relevant to mention the importance of the sequencing of policy changes. In our case, some things, such as financial deregulation, were achieved during the 1980s (although obviously not enough) and supported demand when this was already strong. Later on, when the economy was on its way down, tax changes were implemented that encouraged saving. To this should be added the mistake I have already mentioned, of not liberalising outward capital flows at the same time as the domestic credit market is deregulated.

The crisis culminated in the fall of 1992, when for a few days repo rates were raised to 500 per cent in the vain hope of stemming currency outflows. Interest rate hikes and other contractive measures for defending the fixed exchange rate in the midst of European currency turmoil further exacerbated the economic downturn and added to the banks' loan losses. All in all, in the three years from 1991 to 1993, Sweden's GDP fell by a total of around 6 per cent. Before the financial crisis had even been recognised in the summer of 1992, production had probably already fallen by 2-3 per cent.

It is my impression that the process in these respects has been quite different in Asia, perhaps apart from Thailand. Growth did not drop much before the crisis. More important were the direct contagion effects, particularly via financial markets. Other “triggers” were political, such as uncertainty caused by upcoming elections, and perceived balance of payments problems. In many respects the process—once it had started—was also more rapid in Asia. Perhaps markets have - after all - learnt something from previous crises in Europe and Latin America. They did not see the crisis coming but once it started, they reacted more rapidly.


4. Crisis management

I will now turn to the issue of how to resolve a crisis, once it has erupted. It is obviously necessary to tackle macro and micro problems in parallel since the two are interlinked.

As I stated earlier, I believe that the origin of the Swedish crisis was primarily macroeconomic, due to policy mistakes in the 1980s and early 1990s, and triggered by the policy shift to a low inflation regime. (Although there were also structural weaknesses in the financial sector). The solutions were also primarily macroeconomic in kind. Once the krona peg had been abandoned, demand got a stimulus. The task confronting policy makers was to restore the credibility of economic policy. This required in our case, given the weak fiscal position, very strong fiscal measures (some 8 per cent of GDP 1994-98). But it also called for a consistent monetary policy. However, and this is important to underline in the present context, if the financial problems had not been managed properly, they could easily have been a drag on the macro economy. For example, by taking prompt action it seems that we largely avoided a "credit crunch", which was a possibility that worried people a lot at the time. Credit volumes obviously fell, but this could be explained by the reduction in demand.

Briefly, the important components of the solution for the financial sector were as follows:

1. The Government, later confirmed by Parliament, issued an unlimited guarantee stating that no depositors or other counterparties to Swedish credit institutions would suffer any losses. This was a fundamental measure in order to restore confidence in the Swedish banks both from domestic and from foreign depositors and other counterparties. The results showed themselves rather quickly, e.g., in the form of renewed interbank credit lines with foreign banks. Also the Swedish public remained calm and there were no runs on the banks. However, the government (Parliament) guarantee and the other measures must also be made known to all interested parties.
2. The mandate to implement the support policies was given to a new agency, the Bank Support Authority, which was at arm's length from the political sphere but accountable to the Government. To avoid conflicts of interest, it was found appropriate not to have the Support Authority in either the Ministry of Finance, the Central Bank or the Financial Supervisory Authority.

3. A common framework of measures was constructed for support of the banking system. A strategy for deciding which banks to reconstruct and which to liquidate was developed and explained to the general public. The measures were designed to minimise costs for the Government and the risk of moral hazard. Consequently, shareholders were not covered by the government guarantee and would lose their equity investments to the same degree as the Government had to provide support for their banks. This provided an incentive to the owners to avoid applying for state support unless it was absolutely necessary, to minimise the amount of support and to reduce the time-period of receiving support. We used a very simple, but clear and easy-to-explain, method of identifying banks which should be given support in order to survive and those which should be liquidated or merged. First you write-down the bank’s bad loans. Then you test the bank in a micro- and macroeconomic model. If the model indicates that the bank will be adequately profitable again in the medium-term future it should be given support to survive. But if the model indicates that the bank will never be profitable again the bank will not be of economic benefit to the society and should thus be closed or merged in an orderly manner.

4. Strict valuation rules were used from the onset of the crisis. This was deemed necessary in order to restore confidence but the policy was not without risks. All bank assets had to be marked-to-market even if the market, such as that for real estate, was exceptionally weak. This led to higher bank losses and a weaker financial position than if a somewhat more flexible valuation rule had been adopted. The Bank Support Authority hired experts on valuation of various types of assets to ensure that the banks did not over- or underestimate the value of their assets with the aim to motivate requirements for more support. In addition, the rules on when to identify a loan as a non-performing loan and how much to write-off were made stricter and with less room for the banks’ own judgement.

5. The Swedish government did not take any specific measures to rescue or reconstruct the non-financial companies, neither did it instruct the banks on how to use the support they had received. All support was provided via the banks and the state-owned asset management corporations, leaving them to negotiate the best terms for themselves and their non-financial customers. There was a wave of bankruptcies and restructuring of non-financial companies, primarily in the domestic sector due to weak domestic demand.

6. The extensive use of Asset Management Corporations (AMCs) in Sweden has become well known abroad. AMCs had, in fact, been established in other countries before the Swedish crisis, so this was nothing new. The special feature in our case is the emphasis on splitting an ailing bank into a good bank and a ”bad part” and transferring the bad assets to the AMC at carefully assessed market values. Once in the AMC, the assets are regrouped and improved in order to become attractive to potential buyers. As soon as a reasonable price can be obtained, the asset was sold. This was sometimes a rather time-consuming process but in the end the results were probably better than if there had been a quick “fire-sale” of the assets. It is my opinion that the use of AMCs is superior to having the bad assets remain in the problem banks. The bank staffs are no experts in the field of handling bad assets and may not obtain the best sales values. Also, the bank management needs to tackle other problems and to work hard on future strategic planning for the bank – they do not have time enough for the often very large number of complex decisions to be taken on the bad assets.


5. Conclusions

Finally, can we draw any conclusions for Asia in general and Korea in particular from the way the financial crisis was handled in Sweden?

First, I think it is important, as I said before, to recognise the importance of macroeconomic developments. Once the fixed exchange rate had been abandoned, Sweden's economy experienced an export-driven upturn. But it did take a long time for confidence in the Swedish economy in general to return—at least another two years. It is noteworthy that this was so despite the fact that the recovery in growth had started in the summer of 1993. The explanation lies primarily in the fiscal situation, to which the financial sector and related problems had made a substantial contribution—enlarging the budget deficit by perhaps as much as 6% of GDP at the time. On top of this, the krona was depressed for a long time by the repayment of foreign loans.

The financial problems did not have a major impact on economic developments from 1993 onwards. Once the important decisions concerning the financial sector had been taken, it was not long before the financial aspect of the crisis had disappeared from the headlines. And the banks were already recovering in late 1993—just a year after the peak of the crisis. This in itself was obviously a great achievement and it may be worth briefly stating some of the more important lessons that, in my opinion, can be drawn from this episode of financial crisis management:

1. Confidence needs to be restored rapidly. This was achieved in Sweden as far as the financial sector was concerned. If confidence returns fairly soon, the real effects on the economy need not be large. But if, on the other hand, the lack of domestic and/or external confidence were to persist, this could act as a drag on economic development for a long period. It is also my experience from other countries that the longer the period with a lack of confidence, the more drastic the measures that will be needed to restore credibility. Among other things, restoring confidence involves being transparent about the problems and about the measures for dealing with them. It also means that measures are implemented as stated by the responsible authorities, without any deviations in timing or application. Just the suspicion that specific industries or other interest groups may be receiving unjustifiably favourable treatment would quickly erode any confidence.

2. Transparency is important. A severe crisis wipes confidence out—rightly so. We realised that it was up to Sweden to regain it. To do so, all the cards had to be on the table. The extent of the losses and the magnitude of the bad loans had to be revealed in their entirety and valuation had to be prudent. If you hide losses initially or are overly optimistic in valuations, you risk having to apply again to the international investment community or to taxpayers for additional support. This, in my view, might jeopardise confidence and legitimacy. It is also necessary to have highly transparent and fair ”rules of the game” for government support to the financial sector, not least to avoid political bickering. The responsibility for transparency falls primarily on two parties: the authorities and the banks. The authorities must be transparent about the broad framework for the solution of the crisis and the banks must be transparent about their losses and other deficiencies, financial position and need for external support.

3. Someone must take the losses. It is important to realise that once the crisis has occurred, the losses cannot be hidden and must be dealt with in one way or another. In this respect Swedish institutions worked well. Party politics were clearly separated from business realities. And the politicians did a good job, realising that bygones were bygones, setting up a fair framework to sort the crisis out. From a political perspective it was important that the banks' owners were not rescued. People at large had to take substantial cuts in living standards—at least temporarily. Letting the “big guys” go unhurt under these circumstances would have jeopardised the whole operation.

4. The macroeconomic policy mix for crisis management can differ. In our case, there was obviously not much room for fiscal stimulation of the economy. Considering the much lower public debt, there could be more room in some Asian countries but the potential budget effects of the financial losses also need to be taken into account. Monetary policy has a difficult dual role: it has to strike a balance between stimulating the economy so as to ease the burden on the banks and on their borrowers and at the same time ensuring the capital inflows that are needed to rebuild the depleted foreign currency reserves.

5. On top of this—and perhaps the most important factor— there was political consensus. The government sought and received backing from the main opposition party. That party was also allowed to appoint Board members of the Bank Support Authority so that it could be continuously informed and influence decisions. That was important both for political legitimacy at home and for credibility abroad—investors could be sure that the system we had created would be upheld even if there were a change in government.

Let me end on a positive note. Economically, the crisis in Korea and in many of the other Asian countries should be manageable although it will probably take time before the results are clearly visible. Korea is in far better shape than many other countries, including my own in the early 1990s, given the high rates of potential growth and that the fiscal balances are in reasonable shape. However, what actually happens will depend on how the crisis is managed.

That is now precisely the question.


Last reviewed

Content expert

Contact content expert

Fill in the information

To minimize automated spam, please answer the question in the box below.

7 + 4 ?