NOTE: I am no longer developping sec_edgar. For a much cleaner (more comprehensive test suite, better modeled) replacement, based on XBRL reports (rather than hand-parsing raw text submissions), check out https://github.com/jimlindstrom/FinModeling.

Overview

sec_edgar is a gem that can download financial statements from the SEC/Edgar database, parse them, perform valuations, and write the results to CSV. It performs well for many NASDAQ tech companies, but has not been widely tested beyond that sector.

See Also

For an example of using this gem to perform valuations, see: https://github.com/jimlindstrom/stock_screener

Example

The following example retrieves all of Google's annual reports (10k's). It then reformulates them to highlight the business's operating fundamentals. Based on the reformulated statements, it performs a two-stage valuation, based on N years of forecast data and then a long-term growth rate.

 require 'sec_edgar'
 
 # set up the valuation with forecast data
 ticker = 'GOOG'
 beta = 1.15
 forecast_data = 
   [ { :revenue_growth => 0.25, :sales_pm => 0.20, :fi_over_nfa => 0.01, :ato => 1.50 },
     { :revenue_growth => 0.20, :sales_pm => 0.20, :fi_over_nfa => 0.01, :ato => 1.20 },
     { :revenue_growth => 0.15, :sales_pm => 0.20, :fi_over_nfa => 0.01, :ato => 1.10 },
     { :revenue_growth => 0.15, :sales_pm => 0.20, :fi_over_nfa => 0.01, :ato => 1.00 },
     { :revenue_growth => 0.15, :sales_pm => 0.17, :fi_over_nfa => 0.01, :ato => 1.00 },
     { :revenue_growth => 0.10, :sales_pm => 0.16, :fi_over_nfa => 0.01, :ato => 0.90 },
     { :revenue_growth => 0.04, :sales_pm => 0.15, :fi_over_nfa => 0.01, :ato => 0.90 } ]
 
 
 # get the financial summary (from all historical 10k's)
 summary = SecEdgar::Helpers.get_all_10ks(ticker)
 
 # calculate cost of capital
 tax_rate = 0.35
 rho_e = 1.0 + SecEdgar::Helpers.equity_cost_of_capital__capm(SecEdgar::Helpers::RISK_FREE_RATE, 
                                                              SecEdgar::Helpers::EXPECTED_RETURN_FOR_EQUITIES, 
                                                              beta)
 rho_d = 1.04 # decent average to work with
 rho_f = SecEdgar::Helpers.weighted_avg_cost_of_capital(ticker, summary, rho_e, rho_d, tax_rate)
 
 # perform valuation
 shares_outstanding = SecEdgar::Helpers.get_shares_outstanding(ticker) 
 v_cse_share_0 = summary.sf2_valuation(forecast_data, rho_f, shares_outstanding / 1000.0)
 
 puts "shares outstanding: #{shares_outstanding}"
 puts "             rho_f: #{rho_f}"
 puts "   value per share: #{v_cse_share_0}"
 summary.to_csv("summary.csv")