Hidden in plain sight: Correspondent banking in the 1930s

We present novel quantitative evidence on the number and location of correspondent banking relationships in the 1930s, a neglected area of international banking. Our data, collected from Thomas Skinners’ Bankers’ Almanac, captures over 2000 correspondent banking connections primarily based on London and New York and a smaller cohort of multinational banks. We draw on the new institutional economics and international business literature to explain the relative ubiquity of correspondent banking and the relative scarcity of multinational banks. Our argument that bilateral trade flows drive correspondent banking is tested empirically using an instrumental Poisson pseudo-maximum likelihood estimation.