Similarly, for each firm sample, an abnormal return (ARit) is generated around the event dates of -1, 0, +1 (day 0 representing the day that the firm's financials were available per DJNRS
Empty explanations represent days for which either no articles were found in the DJNRS or that the articles that were found did not state a specific explanation.
Table 1 Distribution of calendar event days I search the DJNRS for articles concerning the peace process in the Middle East in the Wall Street Journal, the New York Times, the Washington Post and the Los Angeles Times during 1993-1997.
Data are obtained from the DJNRS
database, the CRSP and Compustat files, and the US Department of Commerce.
Although we are unaware of any documented biases in our data sources that could confound our analysis, sampling of periods available for disclosure and actual forecast disclosures is limited by the completeness of sources such as Compustat and DJNRS.
7) Our DJNRS search keywords are: "expects earnings," "expects net," "expects income," "expects losses," "expects profits," "expects results," and three similar lists with first words "forecasts," "predicts," and "sees.
Table 1 Study Sample Summary Sample 1 Sample 2 (1987-1998) (1999-2010) Forecasts identified by DJNRS
12,802 23,408 Firms removed due to insufficient (898) (1,201) Compustat data Firms removed due to insufficient (87) (108) CRSP data Final overall sample 11,817 22,099 Table 2 Test of Hypothesis 2 Table entry is Average Management Forecast Error Deflated by Firm's Stock Price 180 Days Prior to Forecast Model: [summation] [[fe.
is also reviewed to insure that confounding factors, such as change of corporate ownership or form, or management change, are minimized by excluding any firms which contain these events.
Thus, the DJNRS
reports the events as they occur, so that price reactions attributable to any of the reported events occur either the day of, or the day DJNRS