Market Timing and Share Repurchase : A Regression Discontinuity Approach

We use plausibly exogenous shocks to stock price to study share repurchases. We exploit variations in stock price around the cutoff of Russell indexes. Russell 1000 consists of the largest 1,000 stocks, and Russell 2000 the next 2,000. This discontinuity leads to price decreases for stocks deleted from Russell 2000, and these price effects are not driven by firm fundamentals. We find that the likelihood of repurchases is lower by 25% for stocks deleted from Russell 2000. Our results support the market timing hypothesis of share repurchases