Divisia Monetary Aggregates

If macroeconomists agree on anything, the most likely candidate would be the long-run relationship between the aggregate price level and the quantity of money or, alternatively, the long-run rates of money growth and inflation. But even this relationship is clouded. Even the relatively close relationships in the 1960s and 1970s, were disturbed occasionally by large shocks to the relative prices of energy and other primary commodities that had disproportionate effects on the price level (see, for example, Tatom, 1981). By the 1980s, however, some as-yet-unknown influence caused the trend rate of velocity to deviate sharply from the nearly constant 3 per cent rate it had exhibited over the previous thirty-five years and undermined virtually all attempts to ‘fix’ the aggregate price equation with further empirical efforts along these traditional lines. As such, the growth rate of the M1 aggregate in the USA (and similar measures abroad) became a very poor indicator of future inflation and led to some large overpredictions of inflation until at least the middle of the decade.

At the conference in Oxford, Mississippi, it was tempting to locate the theme of the conference in the work of Oxford’s greatest son, and, indeed, one of the greatest literary figures of the twentieth century. William Faulkner. It is clear that Faulkner was a student of the economies of inflation … Why else would he have written a book on ‘Soldiers’ Pay’? It could also be that he anticipated the demise of monetarism in ‘As I Lay Dying’, though this more likely to be an allusion to the Keynesian medium term (since we know what happens to Keynesians in the long run!). No doubt he had Divisia in mind as the ‘Intruder in the Dust’. As far as this chapter is concerned, we shall anticipate the comments of our discussants by suggesting the potential parallel between this and ‘The Sound and the Fury’. This is a story told through the eyes of several different characters, one of whom is a congenital idiot. We shall leave it to others to identify which of us is Benji. Our learned discussants would, no doubt, wish to complete the line in William Shakespeare’s ‘Macbeth’, from which Faulkner drew his title: … ‘it is a tale told by an idiot, full of sound and fury, signifying nothing’.

Financial innovations and regulatory changes are often associated with the deterioration of previously stable monetary relationships in several countries. ‘Missing money’ and ‘velocity puzzles’ have led to frequent redefinition of monetary aggregates and, finally, particularly in the USA and the UK, to the abolishment of monetary targeting strategies. In these circumstances it is not surprising that research on alternative monetary aggregates, based on sound microfoundations and consistent with economic aggregation and index theory, has been intensified.

This paper is divided into four parts. Section 5.2 describes the Swiss National Bank’s (SNB) internally used monetary indicator and reporting system, of which the Divisia aggregates are part. Through financial innovations and the increasing importance of non-cash payments, the SNB is currently revising the definitions of its simple-sum as well as of its Divisia monetary aggregates. We sketch the main aspects of the revision in Section 5.3. In Section 5.4, we present different empirical tests, evaluating the relative performance of the present and newly defined aggregates. To determine the short-run predictive information content of money, traditional Granger-causality tests within low dimension vector auto-regressive (VAR) systems in terms of first differenced variables are carried out. Vector error correction models (VECMs) are also used, to shed light on the relative importance of different monetary aggregates, when possible long-run relationships between money, prices, interest rates and output are explicitly taken into account. Tests of additional predictive information of lagged money in a structural Keynesian price equation supplement the VAR-Granger tests. As a final exercise, a series of Engle-Granger (1987) cointegration regressions of income velocity on the yield of government bonds are carried out. Section 5.5 concludes. The main findings of the investigations are as follows: for most of the measures studied, money does — as expected — help to predict prices and output in the short run as well as in the long run.

In the 1970s, the heyday of monetarism, ‘money matters’ gradually became incorporated into mainstream macroeconomics. It was generally recognised that shocks to the money supply formed an important source of business-cycle fluctuations, and that excessive money growth caused inflation in the intermediate run. As a consequence, many central banks switched to a policy of targeting growth rates of monetary aggregates to control inflation. Prime examples are the USA, Switzerland and Germany. Over the years since then, however, most central banks in the industrialised world have again abandoned monetary targeting, although inflation control – or even price stability – remains their dominant policy objective.

An application of the Divisia aggregation theory to Japanese monetary data was first made by Ishida (1984). This paper argued for the use of Divisia monetary aggregates in the case of broadly-defined money (M2 + CDs, M3 + CDs), whereas for narrowly defined money (M1) it was found that the aggregation method made no significant difference in the chosen empirical applications.

Since the mid-1970s, the Korean financial markets have experienced vast innovations and various regulatory policy changes, even though the speed of change has been more moderate than in many advanced industrial nations. Since the end of the 1980s, these financial innovations and regulatory changes have resulted in the rapid growth of non-bank financial institutions and the widespread introduction of various new financial assets in Korean financial markets. These developments have led to large fluctuations in the M2 measure of the money supply, which has been used as the official monetary target variable since 1979. Through portfolio shifts from components of M2 to other highly liquid assets with higher interest rates, which are offered mainly by non-bank financial institutions, higher volatility in M2 growth has raised questions about the appropriateness of M2 as an intermediate target or indicator of monetary policy. In order to overcome this situation, several researchers and practitioners have suggested the use of broader monetary aggregates such as M2B and M3 as replacements for M2.¹

Taiwan has implemented major financial liberalisation measures since the late 1980s. In July 1987, trade-related foreign exchange controls were abolished and capital-flow-related foreign exchange controls greatly relaxed. The entry of new securities firms was permitted in January 1988, with the result that the number of securities firms increased from 60 to 150 within the first year. The limit on daily fluctuations in stock prices was raised from 3 to 5 per cent in 1988 and to 7 per cent in the following year. In December 1990, foreign institutional investors were allowed to invest directly in the local stock market. In respect of the banking sector, a revised Banking Law was promulgated in September 1989, leading to bank interest rates on deposits and loans being completely liberalised, and new private commercial banks were allowed to be established. As of the end of 1993, sixteen new private banks had begun operating.

December 1983 represents a watershed in the development of the Australian financial system. This point in time coincides with the floating of the Australian dollar on world financial markets, and marks the culmination of a period of rapid financial deregulation and innovation. Some of the major changes included the removal of interest-rate ceilings and quantitative controls, and the granting of licences to foreign banks.

One of the alternative methods for constructing monetary aggregates that has received much attention in the literature is the method proposed by Barnett (1980), which uses statistical index number theory. His approach makes use of aggregation theory to compute indices of financial assets that reflect the total utility, relative to some base period, attributable to the monetary services obtained from these assets.

Central banks continue to publish simple-sum measures of the money stock and draw policy inferences from their behaviour even though it has been demonstrated conclusively that these data violate basic principles of economic and index number theory. As such, any in-sample results based on simple sum data must be spurious. Furthermore, the absence of any statistical properties in these data preclude their use in making out-of-sample forecasts.¹ Nonetheless, some research, which acknowledges the conceptual error of simple-sum measures, has defended their use on practical grounds.² Generally speaking, the reasoning has been that index number theory raises some interesting and potentially important issues for the construction and use of measures of the money stock, but that measurement has turned out to be unimportant empirically in real-world applications.