The capital market consequences of analyst coverage initiation and termination

This paper consists of three essays. The main theme is to examine the capital market impact of analyst coverage initiation and termination. First, I investigate whether analyst coverage initiation and termination are associated with institutions' investing decisions. After controlling for simultaneity between initiation/termination and changes in institutional ownership, I find that termination is associated with a decline in institutional ownership. The decline is mainly attributable to quasi-indexer institutions, and is more pronounced when firms are covered, in the pre-termination period, by multiple analysts or by analysts from highly reputable brokerage houses. In contrast, I find no evidence that coverage initiation leads to changes in institutional ownership except for a sub-sample of quasi-indexer institutions (which are covered by analysts from highly reputable brokerage houses in the post-initiation period), which increase their holdings (Essay 1).;I further predict that the impact of initiation/termination of analyst coverage on institutional holdings stems from the information benefits associated with analyst coverage. As predicted, I find that the initiation (termination) of coverage is associated with decreased (increased) information asymmetry (Essay 2).;I also investigate changes in firms' disclosure behavior around analyst coverage initiation and termination, and find increases in disclosure activity following both events. The increase in disclosures for the initiation sample stems primarily from increases in earnings and sales forecasts; the increase for the termination sample is attributable mainly to increases in direct communication and miscellaneous items. For a larger sample of firms, I find that one specific type of voluntary disclosure---management earnings forecasts---increase/decrease after initiation/termination. These increases and decreases are pervasive across all types of earnings forecasts, and are more pronounced when multiple analysts follow the firm in the post-initiation period (for the initiation sample) or in the pre-termination period (for the termination sample) (Essay 3).