This paper studies whether directors and (or) CEO overconfidence change the relationship between board size and the performance of a company, such that we can understand whether the mixed empirical results about the corporate governance effect of board size in the past literature are related to insider overconfidence. Using listed and OTC companies (excluding financial sector) from 1996 to 2015 as a sample, the findings are that when director board is large and all insiders are overconfident, firms have relatively poor performance; there is no insider overconfident in a small board creating the better performance of a company. Further, under a large board, the performance of firms without overconfident insiders is better than that of firms with overconfident directors and un-overconfident CEOs, but insignificantly different from that of firms with un-overconfident directors and overconfident CEOs. This result indicates that the negative effect of director overconfidence is more serious than that of CEO overconfidence. Moreover, when the board is small and only one type of insiders (directors and CEOs) is overconfident, the firm performance is superior to that when all insiders are overconfident. Finally, although the small number of directors is favorable for performance, all insiders or just directors are overconfident making performance worse than the large board with all insiders not overconfident or only CEO overconfident. Thus, insider overconfidence is detrimental to performance, especially directors. These results are more significant for operating performance and weaker effect for the market-side performance. On average, insider overconfidence really affects the relationship between board size and firm performance.
