Portfolio behavior of the non-financial private sectors in the major economies

BIS Economic Papers No 17
September 1986


This paper discusses the behaviour of financial asset and liability accumulation and portfolio selection of the personal and company sectors. The principal finding is that the parameters of portfolio demand functions, i.e. the underlying determinants of portfolio choice, are broadly similar between countries. Differences observed in portfolio holdings thus result largely from differing financial conditions.

The analysis uses broadly consistent annual data over the period 1966-84 for stocks of financial assets held by these sectors in the United States, the United Kingdom, Germany and Japan. These data thus cover four of the five principal economies, whose financial systems have widely differing structures and which over this period have experienced various degrees of change, innovation and deregulation. Details of the sectoral and data definitions and sources are given in the appendix.

Portfolio behaviour is introduced in the context of the development of domestic non-financial sector debt in the four major economics. This magnitude has been of considerable interest in the United States recently, because it has been increasing rapidly in relation to GNP since 1982 after a long period of relative constancy. This has coincided with an upsurge of government borrowing and of foreign debt. Various explanations for the behaviour of this aggregate are discussed. However, the conclusion offered is that a more fruitful assessment of such developments in broad aggregates for financial markets requires complementary analyses of the level and determinants of the composition of asset and liability holdings at a sectoral level. This paper analyses two sectors, persons and companies. The public sector is excluded because its actions are better analysed in a framework of economic policy choice than in terms of economic behaviour of atomistic agents, while the financial sector is a holder or transactor of assets and liabilities held ultimately by the private sector, and the foreign sector is only analysed in the flow of funds to the extent that its assets and liabilities impinge on those of the domestic economy.

The analysis commences with a discussion of the behaviour of the size of sectoral portfolios over recent years, then goes on to assess econometrically, subject to the constraints on degrees of freedom imposed by annual data, the determinants of these magnitudes. Are the determinants similar between countries, despite their widely differing experiences? The chosen paradigm is that of a dynamic portfolio model, where asset holdings are determined by relative interest rates, total portfolio size, activity and other variables, but adjustment of actual to desired portfolio holdings takes time. Particular attention is given to the choice between tangible (i.e. fixed capital) and financial assets within the total wealth portfolio, rather than the expenditure-savings decision.

Portfolio distributions are then examined. Firstly, the data are used to assess the significance and universality of certain important trends in financial markets. For example, it is gauged whether there has been a widespread shift of portfolios into interest-bearing assets, from banks to other financial institutions, from depository institutions to "market" assets and a shift within portfolios towards "market" assets held via investing institutions such as pension funds. These are the highlights. However, a complete picture of portfolio distribution may only be obtained by a complete instrument-by-instrument analysis, using flow-of-funds data as described in the data appendix. The paper thus goes on to discuss changes in portfolio shares over time and then analyses the data econometrically using a portfolio approach. The econometrics illustrates the degree to which behaviour is similar between countries and can be analysed in terms of yields and portfolio growth, as well as giving clues to where structural changes have occurred and highlighting certain contrasts in behaviour between countries. The analysis also provides a measure of portfolio instability which can be related to changes in regulations and financial innovation as well as economic stability. From the research shown, broad conclusions may be drawn concerning the behaviour of the private sector in the financial markets and its response to changes which have implications for both the current behaviour and the future development of financial systems.