New York-A rumor is going around that the recent academic paper documenting significant “backdating” of analyst recommendations in Thomson Financial’s database has been retracted.
The paper, published in February by professors at NYU, UVA and London School of Business, discussed “ex-post” changes made to analyst recommendations for the period from 1993 through 2002, finding that almost 20% of Thomson’s I/B/E/S database had been altered. The changes were almost exclusively for the better, improving historical track records by 15-42%.
We received queries from prominent buy-side clients who had heard that the academic study had been retracted. We contacted one of the authors of the study, Alexander Ljungqvist, of NYU, who responded that the retraction was “news to me.” The source of the rumor is unknown.
Comment by Rob Tholemeier:
Backdating? No need. Why bother when all analysts have to do is downgrade a stock after it blows up (i.e., misses earnings expectations, reports an SEC investigation, etc) or upgrades after it jumps (i.e. announces a takeover bid with a premium) using the previous closing price vs. the next opening price for the shares. This method is as common as honey bees. It pads performance 10-25% per instance.