|OT: "Rewriting History"|
Interesting paper by Alexander Ljungqvist, Christopher Malloy and Felicia Marston; discussed in this week's Barron's, Technology Trader column.
papers.ssrn.com (free download)
For the subset of affected stocks, however, the changes result in a decidedly more conservative distribution of recommendations, especially in the late 1990s – a period that continues to be of central interest to researchers.
The database in question is Thomson Financial's I/B/E/S database of research analyst stock recommendations. In some instances, Thomson permitted analysts to review their track records and request that corrections be made (subject to Thomson review). The impetus for other database alterations is known only to Thomson; they have blamed software glitches, but that seems inconsistent with the non-random nature of the changes. One type of alteration involves anonymization, i.e. removal of the analyst's name from a particular recommendation. The authors found, in reviewing all upgrades by the same analyst, that returns for anonymized upgrades were significantly lower than returns for upgrades that weren't anonymized; the bad calls were preferentially anonymized. This wouldn't have changed the overall record of a brokerage firm, but could have made their best-known analysts look better. A disproportionate number of revisions affected analysts subsequently named Institutional Investor all-stars or "top stock pickers" in the Wall Street Journal's annual ranking. Thomson calls the paper a "hatchet job," but in response they have restored most of the analyst names that were anonymized.