Journal of Monetary Economics 9 (1982) 59-71. North-Hollan¢ Pl~'btishing Comp~my ASSET SUBSTITUTABILITY AND MO?:~ETARY POLICY An Alternative Characteril~ fion Carl E. WALSH* Princeton University, Princeton, NJ 08-';$4, US,a. A distance function is defined for a simple portfolio choice problem and then used to express the impact on asset prices of a change in at; asset stock as the str'n of a substitution and a weahh effect. Substitutes and complements a~e defined with referen,:e to quantity rather than price changes. This method of characterizing asset substitutability leads to s!~apler and more easily interpreted results when analyzing monetary policy than doer the standard approach which expresses asset price changes in terms of the price elasticitie~ ,~f the asset demand functions. 1. Introduction In the long running debate during the sixt~o, and seventies between Keynesians and monetarists over the transmission mechanism of monetary policy, it came to be recognized that both groups s!:,ared the same basic view concerning the way in which monetary policy affected the economy. It does so by prod~.cing portfolio adjustments which lead ~:,3changes in asset prices [for surveys see Park (1972), Mayer (1975)]. While accepting this portfofio adjustment framework, different authors :stressed diiff.,~rent types of assets as being involved in the adjustments set off by a monetary disturbance. Differences also arose over the degree of substitutability assumed between money and other assets. Keynes;zns were ,:har~.,zterized as viewing the portfolio adjustment triggered by a change it. the tnoney supply to involve only a fairly narrow range of financial assets a~d as viewing money and short-term assets to be close substitutes. The ability of monetary policy to affect the prices of tong-te:m~ assets, and hence real aggregate demand, wa.,; then viewed to be a function of the substitution ei~ect on the demand for money produced by changes in the prices of mone," substitutes~ The greater tlae extent to which the public viewed money and other short-term financial assets as substitutes, the larger would be these substitution effects and the smaller would be the impact on long-term asset prices of a given change in the money supply [Tobin and Brainard (1963)]. I~1 the standard two asset "I would like to :thank Angus Deaton, Alan Rogers ard an anonymous referee for help'iul zomments. Rest~ngibility for errors remains my own. 0"304--3923/82/O0(OM)lX)O/$02.75 © 1982 North-.Holland