Dynamic Finance Demand Equations: A Comparison of Nonlinear FIML and GMM Results

Barbara Robles
Joint Committee on Taxation, United States Congress
BRobles@hr.house.gov

Abstract

The demand for finance capital in the form of loans and money balances by firms is a standard applied monetary and financial economics research question. This paper approaches loan and money availability for firms by deriving Euler equations for finance and physical capital from an infinite horizon, expected-cost minimization problem faced by the representative firm. The firm employs variable and quasi-fixed factors in their production problem. The variable factors are materials, energy, and services. The quasi-fixed factors are labor, physical capital and financial capital. It has been established in the literature that labor and physical capital have varying behavioral aspects of quasi-fixity. Financial capital is assumed to be quasi-fixed due to two constraints facing the representative firm: 1) the lack of unlimited credit/financing sources and 2) the presence of transactions costs. The designation of the finance capital factors displaying quasi-fixity is an empirical question. Employing manufacturing data for durables and nondurables at an disaggregated industry level, Generalized Method of Moments and Nonlinear Full Information Maximum Likelihood techniques are applied. If the demand structure of the equations is adequately modeled, the two estimators should not produce markedly different results. This study provides empirical evidence on the degree of fixity of the money and loans factors, and information on the assumption of multivariate normality of the variance-covariance matrix.