STOCK RETURNS BEFORE AND AFTER CALLS OF CONVERTIBLE BONDS

Ofer and Natarajan (1987) report negative, statistically significant cumulative average abnormal returns over five years following convertible bond calls. We show that these results are obtained only if returns preceding the call dates are used for market model parameter estimation. Returns preceding calls tend to be positive and unusually large. This means that predicted post-call returns, based on pre-call parameter estimates, are biased upward. Consequently, the corresponding abnormal returns are biased downward. We also discuss a corrected test statistic. We conclude that the evidence does not indicate market inefficiency in the stock price reaction to convertible calls. © 1990, School of Business Administration, University of Washington. All rights reserved.