Figure A.1
A graph that shows a time series plot of the number of S&P 500 constituents that we can recover from Compustat, CRSP, and Optionmetrics. The highest number of constituents is recovered in Compustat, right from the beginning of the series in 1964 until 2018. The number of constituents recovered from CRSP is rapidly increasing and almost at the level of Compustat recovery in 1974, which is the beginning of our sample used for long training. The revovery rate from Optionmetrics is the lowest, but also increasing over time, from 1996, the first year of availability of Optionmetrics data. A second plot in the graph shows the corresponding time series of the market capitalization of the recovered constituents for each of the 3 data sources. The time market capitalization series are close, in particular after 1996, when data from all sources are available.

Identification of S&P 500 constituents. The figure illustrates the ability to detect historical S&P 500 constituents according to the implemented identification strategy. Panel (A) presents the coverage of HSPC achieved at different stages of the data processing. The line in light grey refers to the HSPC found in Compustat. The blue line shows for how many of these constituents it is possible to find stock price information in CRSP. The red line starting in 1996 illustrates for how many HSPC it is also possible to find information in OptionMetrics. Panel (B) depicts the aggregate market capitalization for each of these three groups of HSPC.

Close
This Feature Is Available To Subscribers Only

Sign In or Create an Account

Close

This PDF is available to Subscribers Only

View Article Abstract & Purchase Options

For full access to this pdf, sign in to an existing account, or purchase an annual subscription.

Close