In unreported results, using the subsample in which CEO tenure starts after CFO tenure starts, the coefficient estimate of CEO-CFO LSM is positive but not statistically significant with CFO compensation as the dependent variable. Using the subsample in which CEO tenure starts before CFO tenure, the coefficient estimate of CEO-CFO LSM is positive but not statistically significant either. Using the subsample in which CEO tenure starts after CFO tenure, we conduct a logistic regression to examine whether CEO-CFO LSM is positively associated with the likelihood of a CFO being appointed as a board member. The coefficient estimate of CEO-CFO LSM is positive and statistically significant. We do not use event history analyses to conduct subgroup analyses because the subgroup can artificially create time gaps for the same CFO, which can bias estimation. With the number of M&As and value of M&As as the dependent variables, the coefficient estimates of CEO-CFO LSM are positive and statistically significant using the sample that CFO tenure starts before CEO tenure. These findings further mitigate the concern that CEO selection of CFO drives our findings.