Asset Prices and the Real Economy

In macroeconomic as in microeconomic analysis, it is important to distinguish stocks from flows. The accounts of a company can be used to illustrate the difference: a balance sheet is about stocks; a trading account is about flows. The same distinction also occurs in national accounts. The Keynesian income-expenditure tradition is about the economics of flows, as in Monetarism with attention focused on changes in the rate of growth of the money supply rather than on its level. The city University Debt-Deflation Conference held in April 1994, by contrast, was about balance sheets; that is, it was about stock economics. This is a subject which, although not ignored, has received much less attention than flow economics.

The recession which many countries experienced in the early 1990s had certain unusual aspects. Most notably, and common to all countries, was the behaviour of asset prices relative to the general price level. Before the recession, the prices of both real and financial assets had risen very substantially. This rise was reversed with great speed while the general price level, in contrast, slowed in its rate of increase but fell nowhere. Also much discussed was the fact that most economies entered the recession with an unusually high level of debt relative to income. When the recession came it was in some countries — such as Britain — long lasting, and in some — such as Sweden — unusually severe. In consequence reasons were sought to explain these (and other, noted below) special characteristics of the recession. As a result of the behaviour of asset prices attention turned to ‘debt-deflation theories’ associated in different forms with J.M. Keynes and Irving Fisher.

The Swedish economy has undergone spectacular convulsions since the mid-1980s. Several aspects of the development are common to many industrialised countries: an unusually long period of economic expansion, involving financial liberalisation and characterised by low real rates of interest, rising asset prices, rapid growth in private consumption, declining household saving rates and rising corporate and household indebtedness. Eventually this period of expansion was followed by downturns of varying severity, with increasing real rates of interest accompanied by falling asset prices, balance sheet consolidation, financial fragility , slow or negative income growth, rising unemployment, and expanding public deficits. This cycle is prominent in countries such as the USA, Japan and the UK.² In Sweden, the turnaround was particularly dramatic.³ Therefore, the Swedish case is of special interest when studying the anatomy of the process.

Recent research, both historical and contemporary, has broadened existing analyses of the connections between financial markets and macroeconomic conditions to encompass a broader menu of debt, credit and intermediation linkages between real and nominal variables. It is useful to distinguish two categories of contributions to this literature. In the first, which we label ‘bank failure’ explanations of cyclical fluctuations, one finds research linking bank failures, bank runs and other disturbances to the operation of financial intermediaries to fluctuations in output and employment. Bernanke’ s 1983 article on non-monetary effects of the financial crisis in the propagation of the Great Depression, emphasising the role of bank failures in disrupting financial intermediation and worsening the US depression, is an influential member of this school. ² In the second category, which we label ‘debt-deflation’ theories, one finds studies seeking to establish the relevance for the business cycle of downward movements in asset and commodity prices, movements which, by affecting the net worth of non-financial borrowers, alter spending by households and firms . Calomiris and Hubbard’s 1989 article on the real effects of price-level movements in the postbellum USA is a leading example of this genre.³

The period since 1989 has witnessed a dramatic decline in bank lending in the UK despite a sharp fall in interest rates. This follows several years of exceptionally high volumes of bank lending with corporate and personal borrowers’ debt-income and gearing ratios rising to unprecedented levels. This itself suggests a possible line of enquiry: that the recent sharp deceleration is a reaction to previous ‘excessive’ borrowing (unsustainable gearing ratios) and lending (recent Joan-Joss experience of banks). Perhaps what has recently been happening is a stock-adjustment correction following an earlier stock-adjustment and over-shooting of long-run sustainable positions, and a return to more normal balance sheet positions by both borrowers and lenders.

’Debt-deflation’ refers to one of the ways in which the cumulative past of the economy matters for its future evolution. In a debt-deflation, existing debt contracts become a burden on producers. consumers, or intermediaries, and prevent them from achieving levels of activity that they otherwise would have achieved had their earlier contracting decisions been different. Adverse shocks to real income or a lower than anticipated price level effectively raise the burclen or value of debt relative to equity, and the amount of debt payments relative to cash flow, which in turn reduce agents’ credit-worthiness and force them to limit their activities. As Keynes ( 1931) and Fisher (1933) argued, debt-deflations that constrain expenditures of firms and individuals, as well as lending by intermediaries, can produce debt-deflation spirals of ever-worsening economic conditions. Much research on the Great Depression and other historical depressions has focused on debt-deflation as one of the ways in which shocks to demand produced persistent declines in aggregate economic activity by undermining the credit-worthiness of borrowers and banks (Mishkin, 1978; Hunter, 1982; Bernanke, 1983; Eichengreen and Sachs, 1985, 1986; Calomiris and Hubbard, 1989; Temin, 1989; Bernanke and James, 1991; Calomiris, 1993).

Over the years 1989-92 many of the major industrialised countries have experienced protracted periods of below-trend growth, and some the longest recession since the 1930s. And the most severe recessions occurred in those countries which had experienced the largest increases in private debt burdens. Figure 5.1 illustrates this correlation for ten major countries, the G7 countries together with Australia, Norway and Sweden. It plots the difference between the actual annual growth rate between 1989 and 1992 and the trend growth rate (measured by the average growth rate of GDP between 1974 and 1989), against the change in the ratio of household debt to GDP from 1984 to the end of 1988. ² The larger the increase in debt over the preceding five years, the greater the shortfall in output relative to its trend level. It is not surprising, therefore, that the phrase ‘debt-deflation’, coined by Irving Fisher some 60 years ago, has been rediscovered by the economics profession. Is the coincidence of a rise in debt burdens and the prolonged nature of the recent recession an accident of history, or does it reflect deeper forces affecting the behaviour of market economies?)

The current Japanese recession, which started in the spring of 1991, can be characterised by three features.