Whilst in some financial systems in the early twentieth century commercial and investment banking activities were carried out by functionally separate firms, in others both kinds of operation were conducted under one roof by "universal banks". Explaining the evolutionary paths that lead to these divergent banking structures has remained a hot topic of multidisciplinary debate for many years. So has their respective exposure to financial crises. On the one hand, universal banks -which hold both

more »

... ong-and short-term assetsare able to reduce information asymmetries and internalise risk. But on the other hand, their mixed asset structure arguably decreases versatility during an economic downturn and may create a "dual market for lemons" in which information asymmetries cause financially sound clients and banks to exit the market, leaving only the riskier crisis-prone ones behind. This paper analyses these debates using the case study of the Netherlands in the early 1920s. The literature argues that it is during this decade that the Netherlands experienced her one and only traditional banking crisis from 1600 to the present day, and after which her short-lived experiment with a system of universal banking came to an end. By calculating an equitydeposit ratio panel for the Big Five Dutch banks, this paper attempts to measure to what degree the sector evolved to become universal and subsequently returned to functional separation. It then conducts a matched pair comparison of two similar-sized banks operating in the Netherlands in the 1920s: the Amsterdamsche Bank and the Rotterdamsche Bankvereeniging. Whilst the first escaped the crisis relatively unscathed, the second required assistance from the Nederlandsche Bank, the Dutch central bank. A new and detailed narrative of one episode of the crisis using as yet unused primary sources is developed for this comparison.