RDP 8301: Financial Innovations and Monetary Policy, A Preliminary Survey III Shifts of the LM Curve

Some consequences of financial innovation act to shift the money-supply or money-demand curve (and therefore also the LM curve). First, the development of new transactions technologies has considerably reduced the demand for transactions balances. Moreover, since a greater range of financial “cross-products”, offering some transactions and some investment services, is now available the volume of purely transactions balances can be decreased, thereby reducing the demand for narrow money. This is the observation made by Cagan and Schwartz (1975) from their empirical work in an M1 context. All but transactions balances are transferred to higher-yielding assets outside Ml. As a result, the demand function for M1 shifts to the left[6] and the corresponding LM schedule shifts to the right,[7] producing higher levels of income (for any given IS curve) and a higher velocity of circulation of money (M1).

As far as M3 is concerned, the picture becomes a little more complicated by a second effect. Provided that the present division in the payments system between banks and non-banks remains, then it could be argued that, in an accounting sense, funds will only disappear from M3 if they are put into government securities or are transferred overseas. Otherwise the money which shifts from M1 according to the first shift effect must find its way back into the banking aggregates, in which case an LM curve based on-M3 would hardly shift at all. Consider, for example, funds which are removed from a bank term deposit to be placed with a building society, which in turn puts the funds into its cheque account with a trading bank. Although M1 increases, M3 is unaffected. What this means is that the growth of a range of close substitutes for bank deposits need not lead to a fall in M3. However, despite this accounting constancy of the volume of M3, the velocity of M3 might be expected to rise due to the impact of innovations on the effective money supply. People could base spending decisions on the easy availability of the close substitutes for bank deposits (a form of liquidity) and so a higher level of income could be supported from a given base volume of money by encouraging spending in this manner.

The U.S. experience provides more direct evidence of the impact of financial innovations on the supply of money. New types of transactions media – for example, repurchase agreements (RPs), overnight Eurodollar deposits, sweep accounts and daylight overdrafts – have been designed precisely to provide intra-day funds and consequently to avoid the money-supply reports compiled at the end of the day. In these instances, the supply of M1 available for daily operating purposes is obviously greater than appears in the statistics or is subject to controls.

With these two points, a case begins to form that innovations act to oppose decreases in the supply of money. This factor might produce an outward shift of the money-supply function – at least an inward shift is resisted.

The third consequence to consider is this: it might also be the case in the longer term that, as successive innovations improve the operating efficiency of the whole financial system, the public will need to hold fewer financial assets to fund a constant level of productive activity. The resources freed from this function will be re-invested for the most rewarding purpose which need not necessarily be with another financial institution. Therefore, we might expect that, as a behavioural response to a more efficient financial system, the size of the public's total portfolio of financial assets, and hence the demand for money, will decline, shifting the corresponding LM curve to the right.[8]

The combined effect of the three shift factors pushes the LM curve to the right, thereby working against monetary restraint. Furthermore, it is likely that, for any given increase of the IS curve, innovations enable the LM curve to move accommodatingly to minimise the increase in interest rates and maximise the change in income.


It also becomes steeper. When only the bare minimum necessary amount of transactions balances remains, the relevant demand curve should be vertical since the demand for transactions balances is presumed not to be interest-sensitive. A fourth Influence on the slope of the corresponding LM curve can be discerned here. [6]

Such an exodus could, ofcourse, be stemmed if a market-related rate of interest were paid on current accounts. [7]

In the U.S. context, this effect has also been suggested by Porter, Simpson and Mauskop? (1979). Over the last two decades in Australia, the velocities of M1, M2 and M3 have increased, but for a very broad definition of money – total borrowing by all financial institutions – the velocity has declined slightly. [8]