Abstract

This paper examines abnormal stock returns around the publication of 1,167 reports issued by 26 brokerage firms on 37 small caps admitted to listing on the Italian stock market from 2003 to 2011. The focus is on small caps going public because, for such firms, information asymmetries may be severe and, therefore, analyst reports should be particularly valuable to the market. Several hypotheses are tested. First of all, the market impact is computed for the whole sample and for each recommendation category (buy, hold and sell). The exercise is repeated by controlling for the presence of contemporaneous news, by taking into account only the initiations of coverage and by selecting only the changes in recommendations. The results obtained through a standard event study methodology show that analyst reports on small caps are informative for the market, although the price impact differs across the samples considered. For the whole sample of reports, the cumulative abnormal returns estimated over a three-day event window around the publication date are statistically significant for buy and hold recommendations only (+0.98% and – 0.83% respectively); however after eliminating contaminating events only buys are significant (+1.13%). In addition the buy recommendations included in the initiation of coverage sample turn out to convey information to the market (+1,50%); the highest price impact, however, is estimated for the upgrades included in the revisions of recommendation sample (+2.19%). The second result of the paper is that information leakage is not widespread, given that abnormal returns are almost never significant before the report date. Therefore, on average, the timing of the market reaction to financial research dissemination does not signal tipping or selective disclosure.

The authors are grateful for comments to Giovanni Portioli, Giovanni Siciliano and the participants to the "XXI International Conference on Money, Banking and Finance", Rome, 10-11 December, 2012. We also thank Eugenia Della Libera for her excellent research assistantship. Any mistakes are exclusively ours; the views expressed herein are those of the authors and do not necessarily reflect the views of Consob.