Cuts on Residual Analyst Coverage
Next we turn to the cuts based on residual analyst coverage. Here, and in everything else that follows, we exclude all stocks that are below the 20th percentile NYSE/AMEX breakpoint. Again, this is because the vast majority of these small stocks simply never have any analysts, so there is no variation to work with. Within this truncated universe, we create three subsamples based on residual analyst coverage, with the residuals coming from month-by-month cross-sectional regressions of log(l 4-ANALYSTS) on log(SIZE) and a NASDAQ dummy, just as in Model 1 of Table 2.
In implementing this technique, we choose to measure residual coverage six .months before we start our preformation ranking period. We use slightly “stale” data on analyst coverage in order to address a possible endogeneity concern. McNichols and O’Brien (1996) find that analysts are more likely to begin covering firms when they are optimistic about their near-term prospects. When one combines this finding with Womack’s (1996) evidence that there is stock price drift for up to six months in response to analyst recommendations, it raises the possibility that recent innovations in analyst coverage may be informative about future returns. Although we have no reason to expect that this form of endogeneity would bias any of our key tests one way or another, we adopt the stale data approach as a simple precaution. Intuitively, any patterns that we now find will be driven by the permanent component of coverage, and not by recent (and possibly return-predicting) innovations in coverage.
Table 4 presents the results of this approach. Before getting to the returns for the three subsamples, it is important to check that they have the desired characteristics with respect to size and coverage. Ideally, the subsamples will contain stocks of the same size, yet will display a healthy spread in coverage. As can be seen from the table, the variation in coverage is certainly there. The low-coverage subsample, which we denote by SUB1, has median coverage of 0.1 (mean of 1.5) and the high-coverage subsample SUB3 has median coverage of 7.6 (mean of 9.7). We do a little less well in terms of size matching. SUB1 has a somewhat larger mean size than SUB3 ($962 million vs. $455 million) and at the same time a smaller median size ($103 million vs. $180 million). Evidently, due to non-linearities in the analyst-size relationship, the simple linear regression technique is giving us residuals that do not have exactly the same size distribution across the three subsamples/tdtd