First I want to express my thanks for the invitation to visit the University here in Jönköping and talk about the turbulence we have experienced over the past decade in both the international economy and in Sweden.
What makes this topic particularly interesting is that from time to time the financial unrest, in Sweden as well as elsewhere, has shaped the economic situation in many countries and sometimes also in the world in general.
Early in the 1990s the Nordic countries were beset by severe problems in the financial system. These difficulties were exacerbated by the turbulent situation at that time in Europe’s ERM system. Japan has had to grapple with a crisis for most of the decade. Mexico encountered problems in the mid 1990s, as did countries in Southeast Asia just a few years ago. Last autumn, global financial market unrest arose again when Russia chose to suspend payments on a part of government debt. When this was compounded by problems with an American hedge fund, Long Term Capital Management, the situation in Western financial markets became critical, fortunately only for a time. This year, it is Brazil that has experienced financial difficulties. In addition to the extensive financial worries of this type, there have been relatively abrupt shifts in interest rates, exchange rates and share prices.
In some instances the financial turbulence has stemmed from inconsistent economic policies, as well as from notable shortcomings in the supervision of banks and other institutions. It has not infrequently coincided with and even been the direct cause of far-reaching macroeconomic adjustments in the economies concerned. In Sweden, for instance, GDP fell for three consecutive years, from 1991 to 1993, by a total of 6 per cent and unemployment shot up. Other countries have experienced similar and even greater setbacks. Large loans have been provided by the International Monetary Fund to assist various countries.
In the light of all this, it is hardly remarkable that in recent years matters to do with the global financial system and how it works have been discussed more intensively throughout the world. So my subject today is indeed topical. But let me say at once that I will be considering just a couple of the aspects the title of my talk encompasses. Time is obviously a consideration here but a more important reason is that we do not yet know enough about many of the complex matters in this field.
I shall begin with the crisis in Sweden in the early 1990s because it illustrates some underlying problems that crises in other countries have had in common. If we can understand the fundamental mechanisms involved in earlier periods of turbulence, perhaps it will be possible to avoid or mitigate similar developments in the future. That naturally leads me on to consider what can be done and the role of the authorities.
The crisis in Sweden
Many reasons can probably be found to explain why Sweden in the early 1990s was hit by the gravest crisis since the 1930s. Some of them go back a long way and others probably had more to do with contemporary developments. Macroeconomic factors played a part and so did more structural matters. Many aspects have been thoroughly penetrated in various studies in recent years. Today I shall be concentrating on the financial field.
The deregulation of credit and currency markets in the 1980s was followed by a rapid expansion of credit and strong price increases for various types of asset, including shares and real estate. The GDP ratio for private sector debt rose in five years from 100 to 150 per cent (Diagram 1). Moreover, in the second half of the 1980s asset prices rose in real terms by over 70 per cent.
There are a number of reasons why an expansion of credit leads to rising asset prices and vice versa. The value of an asset can be defined as the present value of the future income the asset is expected to generate. So rising asset prices can be said to mirror expectations of future increases in income, which in turn strengthens both the propensity and the possibility of borrowing. That implies increased demand for credit. On the supply side, rising asset prices increase the availability of credit in that borrowers are in a better position to pledge collateral. The prospect of rising income and cash flows also implies that borrowers will be more able to service and repay debt.
As long as this process is based on reasonable assessments of the future, all is well and good. Problems arise when asset prices and the provision of credit both rise unduly, so that the expectations on which they are based become unrealistic. Looking back, this is evidently what happened in numerous cases. The head of the US Federal Reserve has labelled it ‘irrational exuberance’, a phrase that is now a byword.
A bubble will burst sooner or later. The precipitating factors vary but the event as such is simply a matter of time. The collapse in Sweden began in 1991. Share prices were already declining and property prices soon followed suit.
The real economy is liable to be affected by falling asset prices, just as it is when asset prices rise. Households perceive a loss of wealth that prompts them to restrain their consumption. Another effect comes from the interaction of the credit market and the balance sheets of households and firms: when asset prices and their value as collateral fall, creditors become less willing to provide loans. In addition, falling asset prices may make banks generally less inclined to provide credit, so that even profitable projects suffer. This can ultimately paralyse the supply of credit and lead to the total collapse of the financial system.
While Sweden can be said to have avoided a financial crisis in that sense, both the wealth and the balance-sheet effects probably contributed to the course of the worst economic recession here since the 1930s. Demand in general fell, financial saving rose and credit was rapidly repaid. The steep reduction of the loan stock in the first half of the 1990s is evident from Diagram 1.
However, a severe bank crisis was only one of the factors that hit the Swedish economy. At times in the early 1990s there was also heavy pressure on the krona. In November 1992 the Riksbank was finally obliged to change to a floating exchange rate.
One of several causes of the exchange rate turbulence was the short-term external liabilities that had accumulated in the late 1980s. In that period, households and firms had borrowed the equivalent of around SEK 300 billion in foreign currency through the Swedish bank system (Diagram 2). That corresponded to around 20 per cent of GDP. I mentioned earlier that borrowing from banks rose quite remarkably in the second half of the 1980s; about 20 per cent of this took the form of foreign-currency loans. Much of the borrowing was short-term but the funds were not infrequently used for projects with a longer duration, such as property investment. All went well as long as the loans could be renewed and the banks were able to borrow the corresponding amounts in the international market.
It was not a classic run on the banks, with depositors withdrawing their savings, that triggered the crisis in the Swedish bank system. It was a lack of international confidence and problems with financing in the international market. Growing loan losses made it increasingly difficult for some Swedish banks to obtain the necessary funds abroad. What we saw was a more modern kind of bank run. That was certainly a major reason why the Swedish bank crisis coincided with an exchange rate crisis.
So what induced Swedish households and firms to borrow in foreign currency? The main explanation is perhaps that interest rates abroad were lower than in Sweden. Borrowing in a foreign currency was perceived to be cheaper. While some households and firms may have wondered why and found that the higher Swedish rates were simply a risk premium to cover the possibility of a future devaluation, the consequences of this analysis were generally ignored (a devaluation would have increased the amount of a loan expressed in Swedish kronor, making the burden of repayment heavier). The fixed exchange rate policy was perceived as credible, at least by these borrowers. They were lulled, or lulled themselves, into thinking that the exchange risk, even though it was occasioning higher interest rates than in the rest of the world, would be absorbed by the Riksbank. That meant they could try to benefit from the interest rate differential with impunity. As many of these foreign currency borrowers were players who could not afford loans in Swedish kronor, they had no margin when the liabilities became more costly after the krona’s fall and this accentuated the problems.
What lessons can we learn?
I have now identified three features of the Swedish crisis: rising asset prices, a strong expansion of credit and short-term borrowing in foreign currency. These three factors have in fact preceded almost all the financial crises that have hit various countries in the past decade. This does not apply to last autumn’s turbulence over the Russian suspension of payments and the LTCM hedge-fund problems. I am thinking instead of, for example, Mexico, Indonesia, Thailand, South Korea, Malaysia and Brazil. In these cases, inflated asset prices and both domestic and external debt had an important bearing on the subsequent course of events.
One important lesson is that a floating exchange rate with an explicit inflation target makes it easier to avoid a situation with rapidly rising asset prices, an unduly strong expansion of credit and high short-term borrowing abroad. That does not mean that a fixed, but adjustable, exchange rate is bound to result in problems of this type. Experience from Denmark is a good example.
A clear problem in Sweden’s case, however, was that in the 1980s monetary policy could not be used to counter the tendencies to strong credit growth and rapid asset price increases. Another problem, as I just mentioned, was that the fixed exchange rate policy lulled people into a false sense of security and provided an incentive for short-run borrowing in foreign currency.
With a floating exchange rate, the krona’s fluctuations make the exchange risk more evident (Diagram 2). Such a regime may have other drawbacks but it does show potential borrowers that foreign-currency loans involve a risk. The stock of foreign-currency loans to households and firms is in fact smaller today than it was a decade ago.
A point to bear in mind is that the problem in the late 1980s was not the strong asset price rise as such but the combination of this and heavy borrowing in Sweden and abroad. The wealth effect tends to be symmetric, in the sense that economic activity is stimulated by rising asset prices to much the same extent as it is checked when asset prices fall. Neither are wealth effects usually all that great. Problems arise when they coincide with an expansion of credit. The effects may then be asymmetric in that they are more extensive when asset prices fall. The balance-sheet effects lead to loan losses and, in more serious cases, to the suspension of payments. In the worst case there will be a financial crisis.
The situation today does not resemble what happened in the second half of the 1980s, as can be seen from Diagram 1. Asset prices have indeed risen strongly in recent years but credit has not grown to the same extent. It seems that at least to date in the present upswing, Swedish banks and their customers have had another approach to the provision of credit and to borrowing.
A pause may be called for here to take a look at monetary policy’s attitude in principle to asset markets compared with markets for goods and services. The inflation targeting policy that has been introduced in Sweden and other countries in the past decade has in many cases had the purpose of mastering recurrent overheating in product and labour markets. It might be said, somewhat improperly, that the policy aims to avoid the ‘bubbles’ that have tended to arise in these markets from time to time. To date, the inflation targeting policy must be said to have done well on the whole in this respect; today, the inflation target seems to function as a credible benchmark, a nominal anchor, for price and wage formation. That wage formation does not yet seem to have been fully adjusted—unemployment in Sweden remains high—is another matter. It probably has more to do with structural factors.
For asset markets, however, there is no such clear benchmark. So neither is there anything to prevent psychological factors—irrational exuberance, if you like—from continuing to inflate various asset prices beyond what is reasonable in relation to underlying economic conditions. The path of the American stock exchange in recent years demonstrates that low inflation is often not sufficient to ensure a restrained development of asset prices. That this can be seen as something of a dilemma is understandable.
Does this imply that monetary policy should in fact be used more than hitherto for actively and directly countering sizeable price increases in asset markets? There is no obvious answer to that question but the dominant opinion at present is that the answer should be no. This is because central banks would find it difficult to identify true bubbles in the development of asset prices and judge the appropriate measures correctly. Still, I would like to leave the door ajar. Situations may arise that are hard to foresee today and where monetary policy may need to play a part. If risks of a new financial crisis were to mount up in some distant future, a central bank could not stand idly by.
The type of monetary policy regime with an explicit inflation target that applies in Sweden and other countries does, however, provide ways of coming to terms with problems from fluctuating asset prices. For one thing, the regime means that in the event of asset market instability, the monetary stance is adjusted automatically in a stabilising direction. As asset prices affect aggregate demand and this is an important mechanism for inflation, interest rates will be raised when asset prices move up and vice versa. This can be expected to exert some stabilising effect on asset markets. For another thing, macroeconomic stability as such should have a stabilising influence on financial markets.
Focus on the provision of credit
While a credible macroeconomic policy is needed to prevent financial crises, it is probably not sufficient by itself. It is also necessary to oversee and supervise the financial system, which includes having appropriate rules for the operations of banks and other credit institutions.
In the final analysis, it was the provision of credit that lay at the heart of the events in the 1990s to which I have referred. Extensive credit risks constituted the problem and it seems that banks made similar mistakes time and again. This applies to the generosity of domestic banks in financing asset investments of private players, as well as to the propensity of international banks to place foreign currency at the disposal of domestic counterparties. Lending to financial institutions tends to be highly concentrated and this exacerbates the situation in times of financial turbulence. It may also be asked whether credit assessments were as thorough as they should have been.
What are the reasons for this? One is probably the difficulties for both banks and supervisors in judging the risks in credit commitments. Credit risks are not simple to handle, neither is it easy to obtain a clear picture of different players’ loan portfolios. So genuine difficulties in making assessments are one explanation. Another is that in certain cases, instead of attempting to assess risks independently, banks appear to have simply followed the herd.
A recurrent theme in the decade’s crises accordingly seems to be the underestimation of credit risks. It is therefore up to the banks to establish a better credit policy in their institutions. This credit policy must pervade the whole establishment, from individual credit managers up to the more comprehensive strategies of executives and boardrooms. But that is not sufficient, or rather it probably will not happen unless measures are also taken by the regulatory and supervisory authorities. This is because in their lending and general behaviour, banks are influenced by their expectations of how authorities will act in the event of a crisis.
Role of the authorities
Oversight in Sweden is the responsibility of the Riksbank. We are required by law to ‘promote a safe and efficient payment system’. The payment system is such a large component of the financial system that our role can really be described as being to promote the latter’s general safety and efficiency. Supervision, however, is the function of the Financial Supervisory Authority.
What this means is that the Riksbank primarily oversees developments in a system or a market, while the Financial Supervisory Authority, besides monitoring the stability of the entire system, focuses on the situation in each institution. In practice, however, these two roles are difficult to distinguish clearly, so close cooperation between the two authorities is important. It is indeed a hallmark of our relationship, from the executive level down to managers and their personnel.
Still, the Riksbank is in a position to detect more direct problems at an early stage, since they are likely to show up in the payment system, for which we are responsible. Our possibilities of affecting the course of events then lie in talking with bank executives and informing the Financial Supervisory Authority. Moreover, our series of Financial Stability Reports can no doubt exert a more continuous influence in this respect. In the longer run the Riksbank can also contribute to stability through international cooperation, as well as by participating in Sweden in work on new legislation and the promotion of structural changes in various systems.
I should like to stress the importance of international cooperation between authorities in different fields, particularly now that problems in one country tend to spread to other countries more quickly than before. The Riksbank accordingly participates in the work of the Group of Ten, the Bank for International Settlements, the International Monetary Fund and so on.
The recurrent crises in financial markets clearly show that work on strengthening systems is a continuous endeavour and that much remains to be done. Banks and other financial institutions have admittedly done a great deal to improve risk management, not least as regards loan portfolios. But the pace of developments in the financial sector is so rapid that neither financial institutions nor authorities can rest on their laurels.
It should also be realised that the extensive work in progress on reinforcing the financial system does not mean that in future the system will be completely immune to problems.
Moreover, our knowledge of phenomena of this type is limited. It may be asked whether it will ever be possible to predict financial crises. The models with leading indicators that are being tested at present tend to be too blunt. They sound the alarm about things that never happen and are silent about those that actually occur.
The authorities must therefore focus on doing everything so that banks and other financial systems have such good systems for risk management that the probability of financial crises is reduced to a minimum. We must also work to mitigate any effects and try to ensure that crises which do occur are not as extensive and economically costly as those we have experienced in recent years.
The objective, in other words, must be to combine the necessary regulation and supervision of the markets with the benefits that a well developed financial system can confer. The regulation and supervision of financial firms are necessary for various reasons. But these functions must be carried out in such a way that they do not throttle the potential that can contribute to greater prosperity in the global economy.
Conclusion
Many countries have deregulated their financial systems in the past two decades. For the first time since the outbreak of World War One, capital is again more or less free to move across national borders. Moreover, major changes in communications and information technology have provided new methods and instruments for risk management. As a result of all this, today the financial system is more extensive and globalised than ever before. This holds even though capital flows relative to GDP were also large at the turn of the nineteenth century. At the time, however, it was other countries that received the flows, which also took a somewhat different form.
The rapid pace of financial developments and internationalisation is really a logical consequence of the globalised production of goods and services, with the strong growth of cross-border trade in the world economy. These developments have accentuated the need of new financial solutions for firms and households. Expanding foreign trade and globalisation of the production of goods and services call for the development and internationalisation of the financial system so that the opportunities of greater prosperity can be fully utilised. A well-developed financial system is particularly important, moreover, when advances in tele- and data-communications are beginning to have more far-reaching consequences for production processes.
At the same time, the transition to a more global financial system has not been straightforward. Developments have often been so rapid that systems have been ‘left behind’. But this does not mean that the problems lie in the deregulated financial system or in capital mobility as such. Turning back the clock would be difficult but the notion is worth a thought to indicate the consequences that a return to a highly regulated financial sector would have in terms of growth and employment.
Here I have tried instead to show how the system itself, as well as the authorities around the world, are endeavouring in various ways to contribute to greater financial stability at the same time as the positive characteristics of the new financial system are allowed to contribute to the growth of prosperity in the world.
