Corporate Fraud, Self-Fulfilling Optimism, and Industry-Wide Over-Investment

Abstract

In this paper, we develop a theory of industry-wide over capacity and optimism among rational investors based on the manipulation of investors' beliefs by insiders. We analyze a dynamic model of corporate fraud, where manipulation and consequent investment distortions occur in equilibrium. Industry-wide over capacity and optimism emerge when manipulation by one firm promotes manipulation by other firms in the industry. This dynamic externality occurs because managers are more likely to over-state performance when market expectations on productivity are high; for example, following favorable past managerial disclosures by other firms in the industry. This externality can lead to industry-wide manipulation and a run-up in market expectations that converge to optimistic beliefs and high levels of investment. Alternatively, if there is no such run-up, investors' expectations converge to firms' true productivity, i.e., there is complete learning. Our model therefore exhibits a stochastic learning equilibrium: under certain conditions, this equilibrium produces complete learning on the firms' productivity in the limit; and, under other conditions, there is herding around optimistic beliefs about this productivity. By contrast, in most of the herding literature, informational cascades eventually occur with certainty and herding is symmetric. They do not therefore explain why some industries are more susceptible to investor optimism or overcapacity than others. In particular, we clarify that a run-up in market expectations and investment is more likely when the cost of capital is low, there is more uncertainty regarding industry productivity, the agency conflict with management is severe, and firms' growth potential is high. We also show that the possibility of manipulation by insiders induces excessive persistence in investment and find that downward movements in markets' expectations can be abrupt relative to upward movements.

Keywords: Asymmetric information, Managerial incentives, Learning, Investment

JEL Classification: G32, D23