Look who is talking now

This paper examines analyst recommendations on internet firms that went public during 1997-2000. The motivation for the paper is that analyst recommendations during the internet bubble and its subsequent burst were the cause of the Global Settlement from April 2003. In addition, by using a single branch of industry cross-sectional industry effects do not play a role in the analyses. Key for this paper is whether the affiliation of the analyst (lead manager, co-lead manager or unaffiliated brokerage firm) plays a role in the recommendation. Also, we have examined the relation between the analyst recommendation and the prestige of the investment bank. Furthermore, we have run regressions to analyze the relation between the analyst recommendations and both firm-specific and equity market characteristics. Finally, we examine to which extent the analyst recommendations for the same firm around a specific date are qualitatively similar. Our findings provide evidence that the type of investment bank does matter. In general, regardless the periods of time examined the recommendations of the unaffiliated brokerage firms have significantly lower ratings compared to those of each of the other two types of investment banks. The outcome from the regressions shows that analyst recommendations are a decreasing function of the number of shares offered as a percentage of total shares after IPO: the higher this percentage, the better the recommendation. Also, the higher the percentage changes of the NASDAQ index, the better the analyst recommendations. Using a matched pairs approach we find that the average recommendation for the same firm around the same date of a co-lead manager is significantly lower than that of the lead manager or unaffiliated investment bank. The outcome holds under different model specifications.