Fig..mw

Figure 1 depicts the influence of overconfidence on each firm’s actual profit (versus believed profit). It can be observed that although firms believe that they can earn more profit as the upper bound of the preference increases, actually, the profit of each firm demonstrates only an initial upward trend, but this is followed by a subsequent decline.

Figure 2 shows that the disclosure cost decreases the probability that a firm gets benefits from overconfidence. It is possible that a firm may earn negative profit when its overconfidence level is sufficiently high. Under such a case, each firm over-discloses its information relative to the ideal level , implying that it spends too much money on information disclosure. When the actual revenue cannot balance the cost, the firm eventually earns a negative profit. Figure 2 also shows that when c=0, the cutoff point r* is about 0.25. Yet, when c=0.1, r* is about 0.23, and when c=0.2, r* is about 0.14. As disclosure is more costly, firms are more inclined to be worse off in the presence of overconfidence.

As r increases, firms are more (less) likely to disclose information when r is small (large) enough (see Figure 3a). This is different from Proposition 3 in which firms’ disclosure probability does not depend on r. This difference stems from the influence of consumer heterogeneity on demand elasticity.When r is large enough, on the contrary, the price elasticity under disclosure is higher than that under non-disclosure, leading to the decrease of disclosure probability with r (see Figure 3b).

Overconfidence is always harmful to competing firms (see Figure 4). This is also different from Proposition 4 in which firms may be better off in the presence of overconfidence. This different result is intuitive because, as we have shown, the overestimation of the upper bound ofimplies less fierce competition while the overestimation of the lower bound implies more fierce competition. It is obvious that the decrease of price will hurt both firms.