New Bank and Transportation Stock Indexes from 1793 to 1871, with Comparisons Across Region and Sector, and Against Prior Indexes

Abstract

Good data on US stock market returns before the advent of the Cowles’ (1939) dataset in 1871 have been scarce. Small samples and an inability to observe dividends render current estimates suspect. I report total return for a much larger sample of stocks before 1871 than heretofore seen. I observe prices and dividends for all the large banks trading in the markets of Boston, Philadelphia, Baltimore and Virginia, where past research had been confined mostly to banks headquartered in New York. I also observe prices and dividends for all the large turnpikes, canals, and railroads that traded in this period. In both cases I find significant survivorship bias in prior compilations. Banks that went bust or went to the wall in the Panics of 1819 and 1837 have been ignored. Early transportation firms that soared on speculation and then collapsed without ever paying a dividend have also been omitted from the record. Likewise, the dividend cuts and omissions characteristic of hard times have been overlooked. Net of correcting these survivorship biases, in the period before the Civil War I find significantly lower stock market returns than reported in Siegel (2014). I likewise find bonds out-performing stocks during this period. The paper concludes with a discussion of why stocks should have performed poorly, and bonds well, under the distinctive macroeconomic conditions that prevailed before the Civil War. [This paper has been updated by later work. See revision notes that follow this abstract.]