Note(s): Column (1) of Table 4 reports the regression results after matching on the 1:1 nearest neighbor matching principle. Column (2) reports the results of the placebo test. Specifically, the regression analysis is performed after randomly assigning CEO narcissism levels. Also, we repeat the above randomization process 500 times, where the coefficient of CEO narcissism (CEO_nar) is significantly negative at the 5% level 16 times and at the 1% level 7 times, ruling out the presence of a placebo effect. Other variables are defined in Table 1. All regressions are firm-level clustered robust standard errors. *, ** and *** indicate significance at the 10%, 5% and 1% levels, respectively