A Combined Transactions and Portfolio Approach to Asset Demands.

Abstract

This dissertation is an effort to bridge the gap between portfolio selection models and inventory theoretic models of the transactions demand for money. Each of these two classes of models understates the role of money in an investor's portfolio by focusing on only one motive. Transactions models ignore risk aversion, while portfolio models typically do not include cash needs within the holding period and the cost of coverting non-money assets to money in order to meet a cash requirement. The basic model presented here combines the two approaches by assuming a risk averse investor who will be confronted with a stochastic transactions need at a known point in the period. The cash requirement is met by sales of assets from the portfolio, with assets having unequal returns and transfer costs. The amount of each asset held is a function of the bounds of the distribution of the cash need, the rates of return on the assets, the transfer costs associated with each, and the degree of the investor's risk aversion. Extensions to include uncertainty in the timing of the cash need and a stochastic return on the alternative non-money asset indicate that a risk-averse investor will increase his holding of the safe asset when these additional forms of uncertainty are present. Under simplifying assumptions, the combined model collapses to the standard portfolio theory models using mean and variance. The model is also used to examine the effect of financial innovations on the demand for money. The creation of new assets that are either included in the revised M1 definition of money or are close substitutes for money serves to lessen the demand for non-interest-bearing cash and demand deposits. In a similar way, innovations that lower the cost of converting non-money assets to money also induce the investor to hold less money.