Banking Structure and the National Capital Market, 1869–1914: A Reply

In Gene Smiley's comment on our study of nineteenth-century American financial markets several issues are discussed. ' We group them into three areas. First, he argues that Davis's net rates of return on earning assets are not satisfactory estimates of the return on investment in loans. Second, he argues that our theoretical model of bank behavior is too simplistic to justify our empirical specifications and our conclusion that the New York Stock Exchange had an important impact on financial integration. Third, he argues that the relationship between the yield on securities of the New York Stock Exchange and the regional quantity of loans, which we estimated as a loan-demand relationship, could be explained in terms of the loan-supply behavior of regional banks. However, it is our contention that none of Professor Smiley's suggestions is warranted and that the results of our analysis remain valid. It should be emphasized that Professor Smiley had access to the longer, original version of our article which contained a more extensive treatment of the theoretical model, numerous references to the relevant literature, and an extensive amount of empirical results that were reviewed but not reported in the condensed published version.2 Smiley argues that Davis's net rates of return on earning assets are not satisfactory estimates of interest rates and that gross returns (although not available prior to 1888) are more useful.3 Although more refined data are always desirable, there are ample justifications for using net returns in our analysis of the integration of nineteenthcentury capital markets. In a well-integrated market expected rates of return (adjusted for risk) are equated by movements in the flow of capital. But this is distinct from the notion that stated rates of interest are equated for two reasons. First, investors recognize that a promised (stated) rate is not necessarily the rate actually realized given business and financial risk.4 Thus, the expected yield will be less than the promised yield on all but risk-free debt. In other words, if all ex ante rates were realized ex post, then all securities would be risk free. Clearly, riskier loans require higher promised returns to compensate bankers for the higher probability of default. Second, operating costs (monitoring costs and collection costs, for example) must be absorbed by bankers in their lending operations. If operating costs were not important factors, bankers would have shifted their entire portfolios into consumer loans and