Abstract
Prior literature finds that earnings management is negatively correlated with institutional ownership. The question is whether institutional investors drive down earnings management of the firms they invest in, or they choose firms with lower earnings management. In this paper, we use the instrument variable design of the Russell 1000 and 2000 indices reconstruction to obtain an exogenous variation in institutional ownership. We find that institutional investors do not drive down earnings management. Instead, institutions choose firms with lower earnings management when they make investment decisions. To further support the preference hypothesis, we add measures of institution preference in the panel regression and find that the negative relation between institutional ownership and earnings management disappears.
Highlights
Managers tend to report earnings towards some desired level of profit
Column 1 shows the results based on the total institutional ownership and column 2 separate institutional investors into dedicated (DED), quasi-indexer (QIX), and transient investors (TRA)
We can find there is a significant negative relationship between institutional ownership and earnings management, which indicate that the Russell sample is in line with results reported in the prior literature
Summary
Managers tend to report earnings towards some desired level of profit. the cost of earnings management is quite high. There are two possible underlying mechanisms to explain this negative relationship between institutional ownership and earnings management: (1) institutions drive down earnings management through their monitoring role; (2) institutions prefer firms with lower earnings management when they make investment decisions Both hypotheses can lead to a negative relation between institutional ownership and earnings management, but prior literature fails to disentangle which channel it is, since it is difficult to show that institutions drive earnings management because institutions may simultaneously choose stocks based on corporate earnings management. Many studies use the actual assignment instead of market capitalization rank as the instrument to capture the change in institutional ownership, which may lead to selection bias [16] To mitigate this concern, rather than using the actual assignment, we use firms switching from one index to the other as the instruments for institutional holdings [12].
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have