The relation between the demand for money balances and its determining variables is a fundamental building block in the formulation of modern monetary policy.
In recent policies of using monetary aggregates as a targeted variable, a stable demand function for money is held as a necessary pre-condition.
In mid-70s, monetary policy had regained its popularity; hence, several, empirical studies have been undertaken to examine the money demand relationships during this period.
To comprehend the empirical research on the demand for money in its true perspective, it is essential, to keep in mind the theoretical approaches to the demand for money.
Essentially, the quantity theory of money is a theory of demand for money. In the Fisherian version MV = PT, We may find that V is implicity determined as:
Assuring that V is determined by technological and/or institutional factors, and, therefore, held as a constant factor, we may say that the demand function for money is:
That is to say, M/P, the demand for real balances are proportional to T. So, if T increases the demand for money balances (assuming P to be constant), also increases in the same proportion.
Thus, the Fisherian or classical notion of demand for money is altogether a simple transactions demand for money. Transactions motive, in a broad sense, include precautionary motive.
In fact, the Cambridge version of the quantity theory did recognise transactions and precautionary motives for money balances.
J.M. Keynes, however, regarded by demand for money what amount people 'want' and not 'have' to hold money and Keynes analytically described the idea of liquidity preference.
He introduced the speculative motive for holding money along with the transactions motive contained in the Fisherian equation of exchange or the cash-balance equations of the Cambridge economists.
Keynes regarded money balances and bonds as alternative assets. In this view, bond holding depends on the rate of return on bonds and holding of money balances is inversely related to the rate of interest.
With the introduction of rate of interest as a determining variable in the demand function for money, there is no presumption that velocity or demand for money will be a stable phenomenon.
Post-Keynesian macro-economics, however, developed in several different dimensions with radical ideas.
Prof. Milton Friedman, an ardent monetarist, mooted the idea that money is not held due to liquidity motives but it is demanded like any other asset yielding a flow of services.
He rather emphasized the level of wealth as a major determining factor involved in demand for money. Expected rate of inflation is also supposed to have a bearing on the demand for money. In early empirical studies, Friedman had treated even interest rate as least significant in explaining velocity movements.
In mid-50s, William Baumol and James Tobin, however, stressed the transactions demand for money by stating:
M = (2b T/r) 1
M = demand for money,
r = rate of interest on bond.
T = volume of transactions, and
b = brokerage cost on transactions cost involved in converting bonds into cash.
In 60s, Miller and Orr, extended Baumol-Tobin's analysis of demand for money by allowing for uncertainty it cash flow and provided a new insight that a firm's demand for money corresponds with the variation in if cash flow which is a measure of the uncertainty of the flow of revenue and expenditure.
Moreover, Tobin's portfolio theory of demand for money reformulated Keynes' speculative motive, but actually it undermined the significance of speculative demand for money by showing that since these are risk less assets like savings bank deposits yielding some interest as against zero interest yield on cash balances.
Then people may not hold money for speculative purpose, but may keep balances in savings account and may withdraw as and when needed to grab the market opportunities for bonds/shares.
In post-Keynesian era, economists in their empirical investigations have tried to examine the inverse relationship between the demand for money and the rate of interest.
Some of them have sought to establish a positive correlation between interest rates and velocity. That is to say, with the rise in interest rate, when velocity tends to rise, it implies a decline in the demand for money balances.
Some economists have adopted the approach of bifurcating total demand for money balances into 'active' and 'idle' cash balances and relating the latter with the rate of interest, through a money demand function stated as follows:
M/P = ky + f(r)
y = national income (KMP)
k = idle cash balances measured as M/P - ky
r = rate of interest.
Tobin empirically measured this from 1922-1945 from data about U.S. economy and observed a near unity hyperbolic nature of relationship between idle balances (the speculative demand for money) and the rate of interest.
In seventies, however, a wide range of instabilities in the demand for money has been noticed in the United States.
Modern researchers have, thus, mooted the idea that financial innovation in recent years have contributed to the instability of demand for money in countries like the U.S.A. and elsewhere.
Further, the rate of inflation and rational expectations has also some significant relevance in this matter.
It follows that the money demand function is a part of the complete monetary system, so it cannot be just the function of rate of interest on any other single variable. Apparently, so far there is no satisfactory statistical analysis of demand for money posited by the modern economists.