Monetary Policy Uncertainty, Money Elasticities, and Interest Rates
One of the hypotheses advanced to explain the persistence of high interest rates after 1979 is that the volatility of monetary policy, particularly after the October policy change, increased the riskiness of financial asset ownership. We present a theory in which volatility affects asset choices between risky and riskless assets and is a determinant of changes in the elasticities of money demand and supply. We show that the model is useful in explaining responses to monetary volatility as well as interest rate volatility. In the theory, a change in policy rules that affects volatility causes a structural shift in the coefficients of both money demand and supply, which complicates the task of finding the appropriate setting of instruments in the new regime. This implies that an important prior condition for a change of policy rules should be that the volatility implications are understood.
Three hypotheses are tested: (1) that the elasticities as well as the level of money demand and supply are affected by interest rate volatility, (2) that similar affects are associated with money volatility, and (3) that money volatility increased interest rates in the policy regime between October 1979 and October 1982, and relative stability contributed to lower interest rates after 1982.(1) (1) We use measured volatility differences across policy regimes, but it is not a goal of the paper to explain these differences. For a survey and discussion, see .
Although there are numerous studies showing that shocks from money announcements are associated with volatility in interest rates and in various market prices [4, 7, 8, 10, 15, 19, 22, 28, 30, 31, 33, 35, 39], there are few direct tests of the theory that uncertain monetary policy raises the level of rates for longer than the immediate time period. Long-established theories in economics, however, suggest that this should be the case, if uncertainty in policy generates uncertainty in interest rates. For example, Tobin  shows that greater interest rate risk increases money demand in portfolios, and Turnovsky  relates the demand for money to uncertainty in the future rate of interest through simultaneous portfolio and intertemporal consumption decisions under uncertainty. More recently, Grauer and Litzenberger  found that when money varies stochastically, nominal interest rates are affected through a premium associated with purchasing power risk. Also, uncertainties about interest rates have been shown to affect demand for money through holding costs in an inventory theoretic model .
Among the few empirical investigations of the longer-term connection between interest rate level and volatility, Mascaro and Meltzer , using a IS-LM model incorporating risk, and Klemkosky and Jun , using a capital asset pricing model, find evidence that money variability raises risk premia and hence interest rates. For use in a single equation money demand model, Slovin and Sushka  construct a proxy for interest rate volatility at a given time from the variance of interest rates over the recent past. In their results, volatility affects the level of demand. While they do not test for a monetary policy role in volatility, they suggest that framing monetary policy in terms of monetary aggregates may increase the demand for money if interest rate volatility rises.
In a reduced form macroeconomic model, Peek and Wilcox  show that interest rate reduced form coefficients, associated with structural shift parameters in money supply, change in response to policy regime transitions. Their model with a policy regime variable provides an improved explanation of the high interest rate levels in the early 1980s, although Antoncic  makes a case that April 1980, well after the change in policy in 1979, begins the period of high and volatile real rates.
None of the empirical work cited analyzes the impact of policy-associated changes in volatility on the elasticities of money demand and supply, although theories of the behavioral response to policy argue that elasticities as well as levels could change, for example [17, 24, 40]. …