Banks as Multioutput Oligopolies: An Empirical Evaluation of the Retail and Corporate Banking Markets
区分零售与公司银行市场，用推测变分模型分析银行在各市场的战略行为及跨市场互动，对研究银行竞争与市场结构的学者有参考价值。
The distinction between retail and corporate banking markets is of much importance in real life banking organizations.The two markets differ with respect to concentration, the importance of informational asymmetries, and the extent of customer mobility.Within a standard conjectural variation model we empirically characterize the strategic behaviour within each of these markets, and also focus on cross market interactions to see whether initial moves in one market affects the equilibrium in the other market.' See, e.g., the papers by Benston, Hanweck and Humphrey (1982), Murray and White (1983), Clark (1984), Gilligan and Smirlock (1984), Gilligan, Smirlock andMarshall (1984), Kim (1986), Hunter and Timme (1986), Mester (1987), Berger, Hanweck and Humphrey (1987) and many others.2 Shaffer (1989Shaffer ( , 1993) ) tests for competition in the US and the Canadian banking industries, but does not consider the multi-output nature of banking.Gelfand and Spiller (1987) do consider multiple outputs, making a distinction between peso and dollar markets in the Uru guyan banking sector.' Feinberg and Sherman (1982).' Hunter. Timme, and Yang (1990) are the only ones, we are aware of, to have used a similar distinction between different loan market segments, albeit for a different purpose, by specifying retail and wholesale outputs in their study of cost subadditivity in large U.S. banks.5 Also, larger banks tend to employ a higher proportion of purchased funds.' Or engage in cross subsidization practices.8 We treat the notion of interrelated costs as jointness in production.This phenomenon may or may not result in scope economies.In our present context jointness is sufficient for the strategic interaction among multioutput oligopolies.I The reason is that small changes in bank i's equilibrium strategy in market I will cause small changes in its competitor j's marginal profit schedule and thus induce small changes in j's market 2 strategy.These small changes in j's strategy have first-order effects on i's profits.* 0;J indicates the expected response in market i to an initial change in market j.Subscript 1 represents the retail and subscript 2 the corporate banking markets respectively.