Money Supply
Multipliers

The money multiplier defines the relationship between the money supply and the monetary base, and is usually defined as the ratio of M3 to M0. The behaviour of the money multiplier has changed dramatically in recent years. For the hundred years starting in 1870, it was relatively stable, fluctuating within fairly narrow limits until about 1970. Since then it has more than doubled in magnitude and, in the 1980s, M3 has continued to outgrow M0 by a large margin. During the last ten years the demand for cash by the public has fallen and the demand for bank deposits has increased. The private sector has made greater use of the banks because they have offered interest rates on sight deposits in order to attract business. As intermediaries, banks have expanded their assets, and bank lending has increased because of shifts in supply rather than in demand. At the same time, banks have been able to lower their cash reserves and expand their lending. Thus, for a given volume of base money (M0), broad monetary aggregates have grown disproportionately because the money multiplier has increased.
In Discussion Paper No. 106, Research Fellow Michael Beenstock and Kam-Fai Chan argue that monetary growth can be usefully analysed in terms of the money multiplier and the monetary base, and they present an econometric model of the behaviour of the money multiplier in the United Kingdom over the period 1950-85.

The change in the money supply can be defined in terms of changes in the money multiplier and the monetary base M0. This approach may be contrasted with the 'conventional approach', which explains the change in M3 through the identity which links it to the public sector borrowing requirement and changes in other financial assets. Beenstock and Chan note that there can be no logical inconsistency between these approaches, but they argue that the money multiplier approach may in practice provide a better explanation of changes in the money supply than the conventional approach.

Beenstock and Chan decompose the money multiplier in terms of three components which, they argue, can be regarded as key decision variables by the private sector. These are (1) the currency ratio, the ratio of cash held by the non-bank private sector to their sight deposits with banks; (2) the time-deposit ratio, the ratio of time deposits to sight deposits; and (3) the reserve ratio, the ratio of banks' reserves to their liabilities.

Beenstock and Chan formulate behavioural hypotheses about the determinants of these ratios and test them by means of regression analysis, using annual data for the period 1950-85. The regression estimates suggest that interest rates and the level of economic activity influence the currency ratio as well as the time-deposit ratio. Interest rates, the Bank of England's discount market policies and reserve requirements are found to be the main influences on the reserve ratio.
Beenstock and Chan use their estimated model of the money multiplier to simulate the effects of various disturbances on the supply of money for a given level of M0. They begin by considering what would have happened to the multiplier if banks had not begun to offer interest on sight deposits in the mid- 1970s. The simulations suggest that the practice of paying interest on sight deposits has been responsible for about half of the growth in the multiplier over 1980-1985. The money supply in 1984 would have been about 20% lower than it was (ceteris paribus) had banks continued to refuse to pay interest on current account deposits.

In another simulation Beenstock and Chan consider, for a given level of M0, the effects on the supply of money of an increase in the level of interest rates. According to conventional theory, the demand for money is likely to fall, but less attention has been paid to the effects of interest rates on the supply of money. The authors assume that interest rates are permanently raised from 1965. The currency and reserve ratios both fall but the time-deposit ratio rises. The net effect is for the multiplier to rise. By 1984 the multiplier increased by about 9%, suggesting that the supply of money is quite elastic with respect to the rate of interest. The authors note that these simulations are to some extent artificial, because interest rates have been varied in the simulation, which is only appropriate if they are exogenous to the system. Nevertheless, the results suggest that interest rate shocks have a positive effect on the supply of money.

Finally, Beenstock and Chan simulate the effects on the supply of money of an increase in the level of GDP. The estimation results suggest that this will lower the currency ratio but raise the time-deposit ratio. In their simulation GDP is raised by 1.0% from 1970, and this eventually raises the value of the multiplier by about 0.9%. This suggests that for a given level of M0 the supply of money rises slightly less than proportionately when GDP rises.

Beenstock and Chan argue that money multiplier analysis is now more relevant because M0 is an intermediate target which forms part of the Medium Term Financial Strategy. Assuming this target is approximately attained (as has been the case to date), a model of the money multiplier would provide the basis for projections of the broad money supply, conditional upon a particular level of M0. The authors note, however, that a more comprehensive theory of the supply of money must explain the joint determination of the money multiplier and the monetary base.


The Determinants of the Money Multiplier in the United Kingdom
Michael Artis and Kam-Fai Chan

Discussion Paper No. 106, April 1986 (IM)