Where can I get financial indicators of public companies in the format friendly for statistical analysis?

"Friendly" is mainly about:

  • Panel data in long format.
  • Variable names compatible with statistical packages.
  • Datapoints properly formatted.

Capital IQ Compustat is the example of such friendliness, but what are its open-data alternatives?

A couple of sources that have the necessary information are:

  • 1
    What do you mean by format friendly for statistical analysis? The NYU datasets you reference are a spreadsheet, but you could do a SaveAS->CSV to get them into a comma delimited text file. Jan 3, 2014 at 22:30
  • 1
    NASDAQ also provides a downloadable CSV file for public companies that contains the basics: name, address, last sale, market cap : nasdaq.com/screening/company-list.aspx Jan 3, 2014 at 22:31
  • @Andrew-OpenGeoCode The NYU datasets aren't in the long format and need additional work before they can be imported in Stata or R. That's okay unless there're better sources. Jan 4, 2014 at 12:12

1 Answer 1


The SEC publishes standardized machine readable data in XBRL format on public companies and other entities, such as mutual funds. Public company financial statements are among the data sets available. These data sets are available on the interactive data home page of the SEC, www.xbrl.sec.gov.

In addition, the SEC maintains a market structure and data analysis site. (See announcement.) This site, www.sec.gov/marketstructure, contains data such as metrics for individual securities. The individual security data sets provide metrics for more than 4,800 securities. Columns include: Ticker, Date, Security, McapRank, TurnRank, VolatilityRank, PriceRank, Cancels, Trades, LitTrades, OddLots, Hidden, TradesForHidden, OrderVol, TradeVol, LitVol, OddLotVol, HiddenVol, TradeVolForHidden.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.