Money and Money Substitutes: A Time Series Analysis of Household Portfolios

Abstract

An important open question persists in monetary theory: To what extent are interest-bearing liquid assets effective substitutes for money (currency outside banks and demand deposits adjusted) in the aggregate portfolio of the household sector? Clearly, such assets may not be substituted directly for money as a medium of exchange. Yet they may be poor or excellent portfolio substitutes, depending on the facility with which the household sector adjusts asset holdings to changes in interest rates. The question is of scientific interest for two distinct, but related, reasons. The first is the well-known assertion that if such adjustments are highly sensitive, procyclical changes in the income velocity of money may occur and may thus attenuate the intended effects of a specific monetary (stock) policy initiated by the central bank (Gurley [14, esp. pp. 9-25], Gurley and Shaw [15]). According to this view, the rapid growth of nonbank financial intermediaries and of bond markets throughout the past three decades may have weakened the effectiveness of monetary policy.1 A second reason has to do with the most appropriate deElnition of money. Friedman and Meiselman [ 1 1 ] and Friedman and Schwartz [12], for example, reason that the U.S. money supply should preferably include not only publicly held currency and demand deposits, but also time deposits in commercial banks. Kane