Quasi-natural experiment: Institutional blockholder mergers.

This table reports the results of the difference-in-differences tests under a quasi-natural experiment based on mergers between institutional blockholders located in the U.S. during 1995–2012. Following [15], we apply a symmetric 7-year window around the event year. Panel A shows the results in the unmatched sample. Panel B shows the DID results after a propensity score matching, where the matching variables include firm size, Tobin’Q, book leverage, collateral, investment, profitability, analyst coverage, and institutional ownership, retained earnings, breadth of institutional ownership, industry dummies, and year dummies. The dependent variable, AvgCSR[t+1, t+2], is the 2-year average of firms’ CSP. Standard errors shown in the parentheses are adjusted for heteroskedasticity and clustered by firm. *, **, and *** indicate significance at the 10%, 5%, and 1% levels, respectively.