| Sumario: | This article examines the discrepancies in the estimations of demand functions for money in the U.S. Stephen Goldfeld revealed that the received demand functions for demand deposits and M, seriously overpredicted demand during the recent period. In his estimation Goldfeld used data from the Board of Governors of the Federal Reserve System for the period 1952.2-1973.4. The discrepancies in parameter estimates may be attributed to data revisions. However, when the equations are re-estimated over the period 1952.2 through 1976.2, the results are unacceptable in several regards. In the equation for M, the parameter estimates for GNP and RTD are not significant, that for the interest rate on time deposits has the wrong sign, and the speed of adjustment is extremely slow. The decomposition of M, into currency and demand deposits reveals that the currency equation continues to explain demand well over the longer period and confirms Goldfeld's assertion that the demand for demand deposits is the source of the money demand problem. During the longer period the equation for demand deposits is totally unacceptable. The coefficient of GNP is insignificant, the coefficient of RTD has the wrong sign, and the coefficient of the lagged endogenous variable is greater than one.
|
|---|