The portfolio of the major trading banks : an econometric analysis

Abstract

The Australian trading banks are prominent financial intermediaries of the monetary sector and their portfolio behaviour has implications for monetary policy. The purpose of this study is to develop and estimate a short-run (monthly) model of bank portfolio choice. Since the pioneering article by Markowitz and Tobin, mean-variance analysis has been the primary theory of portfolio selection. The resultant optimum portfolio, though desired, need not be attainable in the short-run and the generalised-adjustment mechanism became a popular way of explaining observed portfolios. While the mean-variance generalised-adjustment framework consti tutes the skeleton of the portfolio model for the Major Trading Banks, it is limited in its applicability, as it obviously fails to take into account the institutional features of the Australian banking system. A s ynthesis of portfolio and banking theories has been attempted. An interesting feature of the model proposed is that it is a priori triangular. This ralses the question of recursivity and the Lagrange Multiplier test for recursivity is discussed in the context of a triangular model with lagged endogeneous variables and autocorrelated errors. A variety of methods has been suggested for the estimation of portfolio models and in general iterative techniques are needed to handle the complications caused by the presence of lagged endogeneous variables, autocorrelated errors and non-linear restrictions on the parameter set. The problem with iterative (greater than two steps) methods is that convergence is not guaranteed. A Hatanaka (portfolio) estimator is proposed. Empirical results support the bank portfolio management model developed and the study concluded with a discussion on portfolio adjustment, the money supply and monetary policy.