Rising sectoral debt/income ratios: a cause for concern?
BIS Economic Papers No 20
Rising debt in relation to income of both the private and public sectors has been of increasing concern to the authorities of several of the major economies in recent years. For example, the ratio of the non-financial sector's gross debt to GNP in the United States rose sharply from around 1.5 in 1981 to approaching 1.8 at the end of 1985. Concern over the consequences of this increase in terms of financial instability has been expressed, inter alia, by Volcker (1986). In other major countries, too, rising debt of the household, corporate or public sectors has often been seen as a problem by the authorities in recent years, for example household sector debt in the United Kingdom and public sector debt in Japan. This paper seeks to analyse the implications of the growth of debt for the stability of the non-financial sectors, and hence indirectly for the financial system. We first offer a broad view of theoretical issues relating debt to stability as well as an overview of historical patterns in sectoral debt and related variables, before narrowing the focus to a direct test of the role of debt in risk pricing and default.
It is concluded from the empirical evidence and from economic theory that under certain conditions, rising debt to income ratios may indeed be a cause for concern. In the case of the private sector, such concern arises from increased risk of default, in the public sector from higher interest rates and the need for higher taxes. These conclusions contradict an important strand of economic theory which argues that methods of finance for the company and public sectors are irrelevant to real economic behaviour. It is suggested that these theories make excessively strong assumptions regarding market efficiency and the rationality of agents. We also reject the view often expressed in the literature that, even if bankruptcy may arise from debt issue, it has no real consequences for the economy but merely redistributes wealth. This view appears to underestimate direct costs of bankruptcy as well as ignoring important external effects on the real economy and the financial system that may arise if the rate of default reaches a critical level.
Nevertheless, it is found that economic theory does offer important insights into the conditions required for debt to lead to economic instability. Most importantly, it shows - for a given level of debt - that the extent to which potential default is realised depends on the behaviour of the other components of the sector's budget constraint, notably income, value of assets and real and nominal interest rates.
Theory also suggests indicators of the current likelihood of widespread default, which may be used to test for the strength and significance of effects of rising debt. One may distinguish between rising debt in a free market equilibrium and disequilibrium increases in debt caused by the loosening of rationing constraints. In the former case, the spread between the interest rate on private debt and a riskless rate provides, in principle, a measure of the market's perception of the riskiness of lending. However, the mechanism may not operate when interest rates do not clear the market; for example, where risk is not easily observable to lenders or interest rates are fixed at non-market-clearing levels by regulation. In such cases changes in default probabilities following rising debt can often only be observed by examination of actual defaults rather than interest rate spreads. The rationing case may have been of particular relevance to household credit until recent years, when a decline in credit rationing has been an important cause of rising household debt.
Graphical analysis of the relationships between debt, income, default, spreads, asset values and interest rate levels illustrates the validity of these conclusions drawn from economic theory. The insights are further borne out by the econometric results, which suggest that private sector debt to income ratios have indeed been an important determinant of both expected and realised defaults. The results also indicate that the magnitude of the effects of debt on default may be estimated when set in a structural econometric specification which allows for the simultaneous effects of the other influences on financial stability. Meanwhile, preliminary results for public debt suggest that growing public sector indebtedness in relation to GNP has tended to increase interest rates, thus increasing pressures on the private sector, though the level of the public sector debt in relation to income apparently has no significant effect on interest rates. This implies that concern with the level should mainly be associated with problems of the higher taxation required to pay future debt interest costs.