RDP 9202: Some Tests of Competition in the Australian Housing Loan Market 1. Introduction

One of the primary purposes of deregulating the Australian financial system in the 1980s was to make banking a competitive industry. The Campbell committee of inquiry made this quite clear in its final report:

“…the committee is confident that with the adoption of its recommendations, a strongly competitive environment will in due course become firmly established and artificial market imperfections will be of limited significance.” (p. 532)

There has been much recent debate about whether this purpose has in fact been achieved. Milbourne and Cumberworth (1990) argue that entry barriers in branch banking may retard competition and hence lead to above-normal profits. However, they find wholesale margins to have generally fallen since deregulation, due possibly to increased competition from new domestic and foreign banks. Harper (1991) argues that banks still enjoy a degree of oligopoly power in the market for deposits. Valentine (1990), on the other hand, argues that deregulation has made Australian banking competitive, as well as efficient in the technical sense.

The evidence brought to bear by these studies is almost entirely descriptive. In contrast, we present econometric evidence on this question. We do so by estimating a monthly model of conjectural variations for individual Australian banks, i.e. the reactions by each bank to the value of loans made by other banks. This model is estimated for owner-occupied housing loans, the only category for which the relevant data are available, over the period May 1986 through December 1990. The model lets us test three hypotheses: that Australian banking is competitive, collusive, or oligopolistic.

Studies of competition in banking in other countries have been numerous. Recent studies of Canadian banking include Nathan and Neave (1989) and Shaffer (1990). They find evidence in favour of competition. Studies by Evanoff and Fortier (1988) and Shaffer (1989) find evidence of competitive behaviour in US banking, while Berger and Hannan (1989) and Hannan and Liang (1991) find evidence that incumbent banks have market power. Gilbert (1984) surveys the literature on market structure in the US banking industry. Studies which use a conjectural variations framework to estimate the degree of competition in industries other than banking include Sullivan (1985), who investigates the cigarette industry in the United States, and Gollop and Roberts (1979) who examine the coffee roasting industry, also in the United States.

In Section 2 we describe the data that we use in this study, and present some elementary measures of concentration. In Section 3 we motivate the use of the conjectural variations framework. Section 4 presents the results and Section 5 concludes.