|dc.description.abstract||Security analysts play a central role in the functioning of financial markets through their privileged position as intermediaries between firms and investors. Analyst activity is important to reduce information uncertainty but it is not unbiased. On the one hand, the literature shows that these sophisticated agents promote market efficiency by facilitating the incorporation of new information into stock prices. On the other hand, there is evidence that analysts underreact to negative information and that they tend to be optimistic about firms they follow.
Recent studies show that the market does not assimilate immediately the disclosure of a first-time going-concern modified (GCM) audit report. This accounting event is part of a wide range of bad news events which investors are particularly inefficient at dealing with. My thesis explores how analysts deal with the GCM audit report and whether they facilitate the correct assimilation of such information into stock prices. In particular, I use a sample of 924 firms for which their auditors disclose a GCM audit report for the first-time between 01.01.1994 and 31.12.2005.
I find that security analysts anticipate the publication of a first-time GCM audit report. My results show that within the one-year period before the GCM disclosure, security analysts downgrade the average recommendation for GCM firms from “buy” to “hold” whereas similar non-GCM firms maintain an average “buy” rating. A number of robustness tests confirm that this finding is not sensitive to the criteria used to select the non-GCM control firm. Moreover, analysts are more likely to cease coverage of GCM firms prior to the GCM event than for matched control firms. In addition, I show that analysts react to the publication of a GCM audit report by ceasing coverage of GCM firms.
My results suggest that investors do not recognize an average “hold” recommendation for a stock of a firm immediately before the announcement of a GCM audit report as an unfavourable message even considering that it represents a downgrade from a previous “buy” rating. In particular, I find that the negative short-term market reaction to the publication of a GCM audit report is significantly higher for firms with pre-event analyst coverage compared to firms with no pre-event analyst coverage. This suggests that analyst activity may be misleading the market in terms of the saliency of pre-GCM unfavourable news by issuing “disconfirming opinions” to the market and thus increasing the “surprise” associated with the publication of a GCM audit report.
In addition, I show that analyst post-GCM coverage does not increase the efficiency with which the market assimilates the GCM audit report into stock prices. In particular, I fail to find significant differences between the post-GCM return performance of covered firms compared to firms with no analyst coverage. However, I show that the percentage of covered firms following the GCM disclosure is significantly higher for those with best post-GCM return performance than for those with worst post-GCM return performance. This suggests that post-GCM return performance explains the decision of analysts to cover GCM firms but analyst coverage does not influence significantly the post-GCM return performance of such firms.
Overall, my thesis contributes to the accounting and finance literature by showing that analyst activity is not providing investors with adequate value-relevant information for their investment decisions in the GCM bad news domain. Firstly, the reluctance of analysts to issue a clear unfavourable message about the stocks of GCM firms seems to explain why the “surprise” associated with the publication of a GCM audit report is greater for covered firms than for non-covered firms. Secondly, the tendency of analysts to cease coverage of GCM firms and the low level of analyst coverage following the GCM announcement may explain why analyst coverage does not reduce the magnitude of the post-GCM negative drift. As such, analyst contribution to the price-discovery process in this case is likely confined to firms with high levels of analyst coverage.||en