This paper brings together the existing commercial bank balance sheets for New England for the period 1785–1837. Approximately 79 percent of all balance sheets issued have been found, and data from these are presented in aggregated form. Data compiled from available statements were then used to estimate the balance sheets of missing banks and the missing items on individual balance sheets. The variables used in estimating included authorized capital stock, a time trend, age of bank, city population, paid-in capital, deposits, loans and discounts, bank notes in circulation, and reserves. The actual and estimated balance sheets are then combined and presented.

Langdon Cheves's role as president of the Second Bank of the United States (1819–1822) warrants reassessment. First, he did not save the bank from threatened failure, since it was, in truth, already sound. Second, he was irresponsible in accumulating huge excess reserves during the depression year of 1820; reserves estimated here as sufficient to support an increase of up to 17 percent in the volume of notes and deposits in the U.S. money supply.

We examine (1) why, when a national capital market developed in the United States, regional interest rate differentials existed and seemed to narrow toward 1900, and (2) the changing patterns of banking behavior in the post-Civil War period. We show that a national capital market was established early (1870s), and regional rate differentials were a response to variations and changes in the interest sensitivity of business loan demand. Bank structure was characterized by monopoly, but it declined because of increasing sophistication of business financing decisions, which became more sensitive to open market rates (in particular the stock market) and to loan rates themselves.

A radical shift in the industrial composition of investment and final demand played an important role in delaying a complete recovery from the trough of 1932. Under the conditions of technological discovery prevailing in the 1930s, this shift in the composition of demand made full employment virtually impossible to achieve as recovery from the crash got under way. A complete recovery required a mass of interrelated new techniques and human and physical capital which, in the timid financial environment prevailing after 1929 (and given the uncertainties of the New Deal), could not be organized on the necessary scale by private investment markets. These findings pose a new agenda for research on the political economy of the New Deal and on the economic history of the post-World War II era.

Over the 1919–1929 period, fluctuations in the value of stock trading on the New York Stock Exchange exercised statistically significant and economically important impacts on the demand to hold cash balances. The marked post–1925 rise in the volume and value of stock trading led to a measurable increase in the transactions demand to hold cash balances, an increase in demand not recognized or seriously discussed by individuals inside or outside of the system. Had it been recognized, it is unlikely that the Fed would have persisted in its antispeculative policies in 1928–1929, policies associated with rises in interest rates and the beginnings of a downturn in real activity in the second quarter of 1929.