Have the anomalies following share buybacks disappeared?

Figures & data

Table 1. Summary of alternative methodologies

Methods	Test statistics
Buy-and-Hold Abnormal Return (BHAR) method	$t_{BHAR}=\frac{\overline{BHA{R}_H}}{\mathit{\sigma }(BHA{R}_H)/\sqrt{n}}$
Standardized Calendar Time (SCT) approach	$t=\frac{\overline{CTAR}}{S\text{tandard}Error}$
Mean Monthly Calendar Time Abnormal Return (MMCTAR)	$t=\frac{MeanMonthlyCTAR}{S\text{tandard Error}}$

Note: This table summarizes the test statistics of alternative methodologies taken into account in our empirical analyses. The standard error is the conventional standard error computed using the CTARs defined in Equation (1). We consider only size-BM-matched control firm approach to measuring the long-run abnormal stock returns after the event.

Table 2. Analysis of buy-and-hold abnormal returns

Holding period	Sample	Benchmark	Mean BHAR	p-Value
Panel A: Equal weight
1 Year	0.28	0.20	0.08^*	0.002
2 Years	0.57	0.63	−0.06^*	0.000
3 Years	0.32	0.33	−0.01	0.082
Panel A: Value weight
1 Year	0.81	0.75	0.06^*	0.000
2 Years	0.31	0.36	−0.05^*	0.008
3 Years	0.74	0.72	0.02	0.521

Note: BHARs following share repurchases are calculated for one-, two- and three-year holding periods. Both equally-weighted as well as value-weighted cases are analyzed.

^* Level of significance at 5%.

Table 3. Calendar time approach analysis

Holding period	Equally-weighted portfolios	Value-weighted portfolios
Panel A: SCT approach
1 Year	0.071 (0.000)^*	0.053 (0.021)^*
2 Years	0.024 (0.54)	0.017 (0.23)
3 Years	0.029 (0.09)	0.021 (0.14)
Panel B: MMCTAR method
1 Year	0.114 (0.000)^*	0.082 (0.000)^*
2 Years	0.036 (0.33)	0.024 (0.78)
3 Years	0.028 (0.16)	0.020 (0.21)

Note: Abnormal returns following share buybacks are calculated for one-, two-, and three-year holding periods using the calendar time portfolio approaches. Panel A shows the results for standardized calendar time method, while Panel B indicates the mean monthly calendar time abnormal returns. Both equally weighted calendar portfolios and value-weighted calendar portfolios are analyzed. The values in the brackets indicate the p-values.

^* Level of significance at 5%.