The paper describing the data collection for the new database can be downloaded at SSRN, Working Paper # 3011486.
This post serves mostly to provide a link to the spreadsheet that contains the data (link at bottom of post). The new database is intended for scholarly use. I don’t know if it will make much sense to an ordinary investor. But here is a brief description of the data aimed at the non-scholar, along with the intended contribution.
The History of Stock Trading in the United States
If you are not a scholar in this area, you might assume that stock market trading began sometime before 1926, but was not well-recorded until that date. You might make that assumption because almost any historical account you would read in the Wall Street Journal or similar outlet would include a phrase like “Since 1926 …
But if market history has developed into somewhat of a hobby, you might find yourself wondering why, if the Dow Jones Industrial Average was launched in 1896, you keep seeing the year 1926 instead. If you asked around, you would be told that yes, stock trading does go that far back, but reliable dividend data can’t be obtained before 1926, so that it is not possible to calculate total return before that date. And total return should be an investor’s focus, right?
Except that if market history has become a hobby indeed, you’d be familiar with Robert Shiller’s book Irrational Exuberance, and its charts showing stock prices back to 1871. “Ahh,” you might think, “that year must be close to the true point of beginning of US stock trading.”
However, that’s not correct; stock trading in the US dates back at least to the 1790s. But for the period prior to 1871, there have been only a few scholarly attempts at constructing an index of total return (see the SSRN paper for a detailed discussion of prior efforts).
The New Database
The new database captures prices and dividends for all the largest railroad stocks from January 1831 through January 1871. That might sound like a partial, industry-specific take on the stock market of that era; but except for the earliest years, railroads accounted for over 90% of the stock market capitalization in the US. The stock market was a market for railroad stocks. Moreover, excellent information on railroad dividends and capitalization has long been available. The new database makes a contribution by bringing together price, capitalization, and dividend data for that era.
The new database is not the first to attempt total return calculations for this period; but I believe it represents an advance on prior attempts, as detailed in the SSRN paper, particularly its Appendix B.
Here is the spreadsheet in Excel. Please let me know if you find any errors; and I would be curious to learn of the results of any analyses you attempt.