We suggest that transparent bias in management disclosures may result from managers processing information in a heuristic, as distinct from Bayesian, fashion when they face imperfect or head-to-head competition. We predict that transparent bias in disclosures is positively related to the extent of head-to-head competition. In addition, when disclosure is discretionary, we show that managers who exhibit viable, heuristic behavior are less likely to disclose than managers who exhibit Bayesian behavior. Finally, when disclosure is discretionary, we show that the increase in the proportion of uninformed managers who exhibit viable, heuristic behavior encourages more disclosure by an informed manager.

We characterize the steady-state equilibrium in which informed traders who exhibit heuristic (i.e., representativeness, as opposed to Bayesian) and Bayesian behaviors achieve the same expected utility. Then, we show how the endogenous, steady-state proportion of heuristic traders is affected by the quality of public information and other exogenous features of our model. Finally, we discuss how the presence of heuristic traders potentially alters the link between improved public disclosure and market liquidity, the variance in the change in price, and market efficiency.

In addition to being a function of traditional fundamentals such as cash-flow persistence and the discount rate, the equilibrium association between a security price and a value-relevant statistic can simply be a function of what rational investors believe the association will be. We refer to this phenomenon as beliefs-driven price association (BPA). By explicitly considering the phenomenon of BPA, we show that the price response to information releases can vary over time even if the risk-free interest rate and investor preferences are static and the earnings/cash flow generating process is stable. This observation suggests, for example, that price-to-earnings associations and price volatility can vary over time even if a stable pattern of economic fundamentals suggests otherwise. The possibility of BPA suggests that measures of the cost of capital, information content, and growth prospects inferred from observed market prices will be confounded. While we do not predict when periods of BPA will arise, we provide empirically testable predictions about how prices should behave during periods of BPA. In particular, we predict that, during sufficiently long periods of high (positive or negative) BPA, price volatility, price levels, and expected returns will be higher than would be implied by a fundamental valuation framework. Finally, while BPA in the pricing of one security does not cause BPA in the pricing of other securities, the price levels of those other securities will be affected if the securities with BPA are sufficiently large relative to the market as a whole.

This paper provides evidence that uncontested director elections provide informative polls of investor perceptions regarding board performance. We find that higher (lower) vote approval is associated with lower (higher) stock price reactions to subsequent announcements of management turnovers. In addition, firms with low vote approval are more likely to experience CEO turnover, greater board turnover, lower CEO compensation, fewer and better-received acquisitions, and more and better-received divestitures in the future. These findings hold after controlling for other variables reflecting or determining investor perceptions, suggesting that elections not only inform as a summary statistic, but incrementally inform as well.