US monetary aggregates, income velocity and the euro-dollar market

BIS Economic Papers No 2
April 1980

Introduction

The basic premise of monetarist doctrine is the proposition, lent support by extensive empirical research, that the income velocity of money behaves in an essentially stable way. That is, technically expressed as a "money demand function", the demand for money has been found in a number of countries to be closely, and principally, related to changes in interest rates and the level of income. Broadly speaking, rising interest rates, at short and/or long term, tend to reduce money demand by inducing a shift towards alternative financial assets, while rising income leads to an increase in the demand for cash balances for transactions purposes. In the monetarist view, control over the money stock is therefore seen as the most effective way, making allowance for variable and sometimes lengthy lags, of controlling final demand and, in particular, the price level. Moreover, the monetary authorities in various countries, though rather more sceptical of the stability of money demand, have used estimates of money demand functions, together with projections of actual and desired levels of final demand, in formulating policies with respect to money growth.

In the inflationary environment of the 1970s the monetarist approach to economic stabilisation would appear to have found increasingly wide support. Yet its basic theoretical premise - the stable demand for- money function - has seemed at times, in different countries, to rest on shifting sands. Previous demand-for-money relationships have tended to break down as a result of rapid institutional and technological change and the growing sophistication of both bank and nonbank market participants in cushioning themselves against the impact of monetary control provisions or circumventing such measures. Problems have arisen not only in the choice of appropriate operating and intermediate target variables but also with regard to the technical control instruments themselves. At the same time, new questions have been posed concerning the appropriate definition and identification of monetary aggregates for purposes of monetary control.

This broad set of issues also has an important international dimension. It is widely accepted that claims held in the Euro-currency market, to the extent that they are not already counted in national money supplies, may serve as a substitute for domestic liquid balances. It is asserted, in particular, that a certain proportion of Euro-currency assets should, to all intents and purposes, be viewed as the equivalent of domestic liquidity. The failure to treat it as such statistically (and to exercise some control over it) may mean that income velocity measured in terms of the domestic aggregates can at times increase faster than would otherwise be the case. One clear expression of this view has been given by Governor Wallich of the Federal Reserve Board, who has estimated that the monetary-type volume of Eurodollar claims which should be added to the US monetary aggregates amounts to about $50 billion and is growing at the rate of about 25 per cent, a year. He concludes that

"... if monetary authorities focus exclusively on the growth of domestic aggregates, ignoring the effects of the more rapid growth of liabilities to non-banks that is occurring in the Euro-currency market, they may facilitate more expansionary and more inflationary conditions than they intend, or may be aware of. Indeed, there is a risk that, over time, as the Euro-currency market expands relative to domestic markets, control over the aggregate volume of money may increasingly slip from the hands of central banks."

These broad issues, domestic and international, pose analytical problems of kinds which cannot easily be taken into account in the usual demand-for-money analysis. On the international side, there is no easy answer to the question whether individual countries should include some portion of Euro-currency claims in their national money supplies and, if so, to what extent. Moreover, on the domestic side it would appear that factors relating to competition and equity have, independently of interest rates as such, significantly influenced the changing pattern of financial intermediation. This would seem to be true both of changes in the regulatory framework and of the development of new financial instruments and payments practices. In some countries, particularly the United States, the banks' recourse to "liabilities management", with a view to minimising the cost and increasing the availability of funds, is a special case in point.

For these reasons, I have found it helpful, as an alternative to the conventional demand-for-money approach, to use a broader, if less rigorous, income-velocity framework - one which can be more directly related to changing patterns of financial intermediation. The paper draws primarily on the experience of the United States, for which comprehensive data on sectoral flows of funds are available. Among other things, it examines some of the links between the Euro-dollar market and US domestic monetary conditions.

In terms of its theoretical foundations, the analysis in this paper leans towards those views which place emphasis on the demand for credit as distinct from the demand for money. Basically, it sides with the Gurley/Shaw "new view" of financial markets, which stresses the need for a financial policy designed to enhance credit creation over financial markets as a whole instead of a monetary policy focusing on the control of specific banking-sector monetary liabilities. On the banking-sector level, the approach is consonant with the "European" (and IMF) view of money creation, which underlines the relative exogeneity of changes on the assets side of the banks' balance sheet: credit to the private sector, credit to the public sector and net foreign assets. A closely related view, of course, is the "monetary approach to the balance of payments", which highlights the role of external flows in equating money-supply creation, as it derives from domestic credit expansion, to the actual demand for money.

The evidence presented in this paper suggests that since the 1960s, in contrast to earlier years, changes in the income velocity of M, in the United States can be ascribed largely to variations in the growth of total domestic credit-market debt in relation to the money stock. In behavioural terms it would appear historically that the income velocity of total credit, i.e. the relationship between financial wealth and income, has become increasingly stable. From the early 1960s onwards the ratio of total credit-market debt to gross national product fluctuated fairly narrowly around a zero trend, even declining slightly during later years when the Euro-dollar market was growing very rapidly. With regard to the domestic expenditure effects of the Eurodollar market, this behaviour suggests one of two things. On the one hand, viewed independently of US credit-market developments, it could mean that any marginal influence that increased non-bank holdings of Euro-dollar claims have been up to now at the domestic level deflationary rather than inflationary. On the other hand, viewed in conjunction with US credit-market developments, it could mean that US monetary conditions have given sufficient encouragement to domestic credit creation and net expenditure abroad to outweigh any domestic expenditure effects deriving from the growth of non-bank Euro-dollar activity.