eThesis - School of Business electronic theses

» hse_ethesis_14584.pdf   size:2 MB (1669962)

BACKGROUND AND OBJECTIVES OF THE STUDY

The key objective of this study is to make sense of the information environment surrounding listed U.S. companies that have exactly one sell-side stock analyst covering them. Despite the large amount of stock analyst literature, this "One man consensus" setting has previously been left without academic attention. For small and illiquid listed companies that receive little attention from the investors, coverage from even one analyst might become decisively important, either in a good or bad way depending on their recommendation. Hence, this thesis studies the market reactions to the buy-hold-sell recommendation revisions issued by stock analysts who form the analyst consensus on their own.

DATA AND METHODOLOGY

The data set of this study covers all single analyst consensus recommendation revisions in the U.S. stock exchanges NYSE, NASDAQ and AMEX between 1994 and the end of 2014. Market reactions for these revisions are analyzed in an event study context for short-term event windows, as well as in an Ordinary Least Squares regression with and without fixed effects. A more novel influential recommendation method is also applied, based on both cumulative abnormal returns and abnormal turnover figures. A control sample of multiple-analyst consensus revisions for the same companies is introduced for comparative analysis. Both the One man consensus sample and the control sample are separately examined for upgrades and downgrades, and implications are thoroughly discussed.

EMPIRICAL FINDINGS

Results from the event study and the regression analysis show that on average, markets react to single analyst upgrades and downgrades in a statistical significant manner. However, the influential recommendation tests reveal that the proportion of visibly strong reactions remains small. One man consensus upgrades tend to induce stronger short-term reactions than the control sample upgrades, whereas negative reactions to downgrades are more muted than in the control sample. The tests control for firm quarterly earnings announcement events that might otherwise provoke incorrect conclusions about stock analysts' influence. Provided evidence suggests that in most cases, analyst coverage even from only one person improves the information environment for listed companies.