This paper shows that the results of variance-bound tests depend on how cash distributions to shareholders are measured. As in prior studies, the authors find apparent evidence of excess volatility when a narrow definition of cash flow (dividends only) is applied. However, they are unable to reject the hypothesis of market efficiency when the cash flow measure also includes share repurchases and takeover distributions in addition to ordinary cash dividends.

ABSTRACT This paper shows that the results of variance‐bound tests depend on how cash distributions to shareholders are measured. As in prior studies, we find apparent evidence of excess volatility when a narrow definition of cash flow (dividends only) is applied. However, we are unable to reject the hypothesis of market efficiency when the cash flow measure also includes share repurchases and takeover distributions in addition to ordinary cash dividends.

This paper shows that the results of variance-bound tests depend on how cash distributions to shareholders are measured. As in prior studies, we find apparent evidence of excess volatility when a narrow definition of cash flow (dividends only) is applied. However, we are unable to reject the hypothesis of market efficiency when the cash flow measure also includes share repurchases and takeover distributions in addition to ordinary cash dividends.