Semi-Nonparametric Estimates of the Demand for Money in the United States

Abstract

This paper focuses on the demand for money in the United States in the context of two globally flexible functional forms—the Fourier and the asymptotically ideal model (AIM)—estimated subject to full regularity, using methods suggested over 20 years ago. We provide a comparison in terms of violations of the regularity conditions for consumer maximization and in terms of output in the form of a full set of elasticities. We also provide a policy perspective, using (for the first time) parameter estimates that are consistent with global regularity, in that a very strong case can be made for abandoning the simple-sum approach to monetary aggregation, on the basis of the low elasticities of substitution among the components of the popular M2 aggregate of money.

The Demand for Assets and Optimal Monetary Aggregation

“…Moreover, the majority of assets are own‐price inelastic (), except for currency ( x 1 ) and demand deposits ( x 3 ) that are own‐price elastic ( and ). Looking now at the cross‐price elasticities, economic theory does not predict any signs, but we note that more than half (71%) of the off‐diagonal terms of the second panel of Table are negative (), indicating that the assets taken as a whole are gross complements, which is qualitatively consistent with earlier finding by Serletis and Shahmoradi (, ) based on different flexible functional forms and highly aggregated demand systems.…”

“…As can be seen in the first column of Table , although positivity and monotonicity are satisfied at all sample observations, concavity is violated at all 516 sample observations when the concavity conditions are not imposed. Because regularity has not been attained, we follow the suggestions of Barnett () and, as in Serletis and Shahmoradi (, ), we estimate the NQ expenditure model by imposing concavity following the procedure discussed in Section 4.…”

“…This literature follows the innovative works by Diewert (), Donovan (), and Barnett (, , , ) and utilizes the flexible functional forms approach to investigating the inter‐related problems of estimation of monetary asset demand functions and monetary aggregation. See, for example, Ewis and Fisher (, ), Serletis and Robb (), Serletis (, , ), Barnett, Fisher, and Serletis (), Fisher and Fleissig (, ), Fleissig and Swofford (), Fleissig and Serletis (), Serletis and Shahmoradi (, ), Jones et al. (), Barnett and Serletis (), and Barnett (), among others.…”

Economic assumptions and choice of functional forms: comparison of top down and bottom up approaches

“…A truncated logarithmic polynomial form is applied in establishing the translog function. Barnett and Yue (1988) and Serletis and Shahmoradi (2005) investigated the asymptotically ideal model (AIM). To establish the reciprocal indirect utility function, define x to be the vector of consumption quantities and p to be the vector of corresponding prices.…”

“…Serletis and Shahmoradi (2005) shows that the quasi convexity can be imposed on h(v, h) by modifying the optimization problem as follows:…”