Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Finance (Business Administration)

First Advisor

Ji-Chai Lin


This dissertation examines the information content and market microstructure effects of analysts' initial coverage. The first essay centers on the valuation effect of analysts' initial coverages and discusses the roles of trading volume and firm size in the revaluation process. First, a large average abnormal return (7.46%) is observed on the day the initial coverages are released via the Dow Jones News Wire. The average abnormal return of initial coverages is significantly larger than the average abnormal returns of analysts' upgrade recommendations, implying marginal information content for firms with initial coverage is greater than firms which are already in the analysts' recommendation lists. Second, small firms with initial coverage experience a larger abnormal return (11.13%) than large firms (4.13%). Third, the abnormal return is found to be positively related to normal trading volume. This is consistent with Bhushan (1989) that the price system of firms with higher expected trading volume contains more noise, but inconsistent with Barber and Loeffler (1993) that an order imbalance due to noise traders' response causes a larger price reaction for firms with less volume and less liquidity. In second essay, market microstructure effects of analysts' initial coverages is examined. Since analysts' recommendations are usually disseminated to clients first and then released to the public, this provides a unique setting to empirically examine the behavior of informed trading. In particular, three issues are investigated. First, how quickly do stock prices incorporate private information? Using a sample of 87 initial coverages, most of abnormal returns on the release day occur at the opening trade and at trades within ten minutes after the opening trade. Second, comparing the efficiency of the call market and dealership market in reflecting private information, private information can be more efficiently revealed in the call market (NYSE/AMEX) than in the dealership market (NASDAQ). Third, when trade size choice of informed traders is investigated, the proportion of trades in medium size, ranging from 500 to 9,900 shares, increases significantly in the private information period. The results are consistent with Barclay and Warner's (1993) stealth trading hypothesis in that traders prefer to use medium size trades when they have private information.