What is the role of journalists in the stock selection process? Are they experts, signal amplifiers or noise amplifiers? Is their collective view short-term or long-term? In two recent papers, Alexander Kerl and Andreas Walter examine the nature and value of the stock filtering role of journalists writing for German personal finance magazines (such as Effecten-Spiegel and Börse Online).
In their December 2006 paper entitled “The Usual Suspects: The Effects of Attention on Journalists’ Stock Recommendations”, they investigate whether journalists are susceptible to attention bias in the same ways as individual investors. Using a sample of specific stock recommendations from all target publications during 1996-2003 (2,412 buy and 1,051 sell), they conclude that:
- Much as do individual investors, journalists focus on recent attention-grabbing events (news, unusual returns, unusually high trading volume) in their buy and sell recommendations.
- When journalists recommend stocks with exceptionally high (low) recent returns, the average short-term performance of their buys (sells) is 2.54% (-4.92%), indicating a short-term momentum strategy.
- Journalists recommend their favorites, with a preference for low-priced stocks, over and over again.
In summary, journalists amplify the attention bias of individual investors, especially for sell recommendations.
In their January 2007 paper entitled “Long-run Performance Evaluation of Journalists’ Stock Recommendations”, they evaluate the long-term performance of buy and sell recommendations issued by journalists. The authors maintain that “the educational and professional background of [German financial] journalists is similar to the one of security analysts.” Using a sample of specific stock recommendations from all target publications during 1996-2003 (2,637 buy and 1,168 sell), they conclude that:
- Buy recommendations overall have a significantly positive average excess return of +4.83% compared to a broad market index after 24 months. However, this excess return derives substantially from size/value/momentum/subperiod characteristics of recommendations and is insignificant relative to a characteristic-adjusted benchmark.
- Sell recommendations overall have a significantly negative average excess return compared both to a broad market index (-7.43% after 24 months) and to a characteristic-adjusted benchmark (-12.6% after 24 months).
- Buy (sell) recommendations for value stocks have a significant characteristic-adjusted average excess return of +8.65% (-21.4%) after 24 months. However, recommendations for growth stocks are less predictive.
- As noted above, journalists apparently subscribe to a momentum investment strategy, with an average 3-month prior market-adjusted return of +1.17% (-6.91%) for buy (sell) recommendations. Buy (sell) recommendations from the group of past winners (losers) have an average characteristic-adjusted excess return of +8.07% (-18.29%) after 24 months.
- Journalists provide valuable long-term buy recommendations during 1995-1997, but not since. In contrast, sell recommendations have value across the entire sample period.
In summary, (German financial) journalists show predictive ability in their (1) sell recommendations and (2) buy recommendations for value stocks and positive momentum stocks.
Differences in financial cultures (e.g., education/professionalism of financial journalists and presence/absence of analogs to Regulation FD) may limit transferability of conclusions to stock recommendations from journalists in other countries.
The authors do not address the possibility of special relationships between journalists and large investors, which might entail attempts by large investors to influence retail investors and/or frontrun publication of recommendations. Nor do they examine evolution of company fundamentals to assess whether financial performance or ongoing media attention per se is decisive in future stock performance.
