Previous research documented that soon after companies cross-list; they achieve significant negative post-listing abnormal returns (the post-listing anomaly). The evidence presented in this study shows that companies cross-list based on either a market-timing consideration or a genuine performance consideration. The host market condition is significant in explaining both the sign and the significance of post-listing abnormal returns. On the one hand, the evidence reveals, companies that cross-list in a host market while that host market is "positive" and achieve a significant negative post-listing abnormal returns, those companies time the market, and the post-listing anomaly is explained in the context of market timing. On the other hand, if companies cross-list in a host market while that host market is "negative" and achieve a positive post-listing abnormal return whether it is significant or not, means those companies did not time the market, which also indicates that the post-listing anomaly does not exist.
Event studies are based on grouping companies by certain characteristics, such as the choice of benchmarks affects the method of forming portfolios of different companies, which in turn affects both the sign and the significance of the post-listing abnormal returns. The sample evidence shows for some of the different characteristic index benchmarks used, different estimation procedures changed the sign and the significance of the post-listing abnormal returns. For the entire different market index benchmarks used, however, different estimation procedures did not change the sign or the significance of the post-listing, abnormal returns. The LARCH (generalized autoregressive conditional heteroscedasticity) estimation procedure is a better fit when using daily returns to estimate abnormal returns, and the characteristic index is a better fit when forming portfolios of companies based on certain characteristics.
Discretionary accruals research reports if companies have a high degree of discretionary accruals, then those companies engage in earnings management. I built a dummy variable DTIMERS that takes the value of 1 if the companies time the market and 0 if they do not. I ran multiple regression models where Absolute Discretionary Accrual is the dependent variable, with DTIMERS as an independent variable along with other control variables. I used a wide variety of both parametric and non-parametric tests, and diagnostic regression analyses adjusting for heteroscedasticity and autocorrelation. The evidence shows the companies that time the market engage in earnings management and that may explain why those companies in the post-listing period achieve significant negative abnormal returns.
This study contributes to the literature by highlighting the relationship between the cross-listing decision, host market condition, post-listing abnormal return, and earnings management. Researchers of cross-listing must take into consideration all those factors, investors ought not buy shares of cross-listing companies without conducting due diligence, and financial analysts should not recommend buying a firm's stock that is cross-listed unless they have examined the timing of cross-listing and signs of earning.
This study leaves open the possibility for further research to study such questions as does cross-listing create value for non-market timers, and does the market generally overreact to cross-listing, regardless of whether or not the companies time the market.