Information precision, signaling, and strategic knowledge sharing

Mehmet Ozbilgin, Purdue University

Abstract

In this dissertation, two research questions are studied. In the first study I theoretically analyze a privately informed firm's decision to adopt managerial initiatives that improve the precision of information the firm uses for internal decision making and the stock market reaction to that decision. I develop a signaling model that provides the result that in a separating equilibrium, a firm with a high intrinsic value would choose a less precise information system when the cost of precision is low and a more precise information system only when the cost of precision is sufficiently high. The former result predicts an unfavorable market reaction to these initiatives and reverses the classic signaling outcome implied by the latter by demonstrating that choosing a precise information system is not a sufficiently costly action for a valuable firm to distinguish itself when the cost of precision is low. Instead, such a firm reduces the precision of its information system and increases the probability of making a poor investment decision to impose costs high enough to discourage mimicking by less valuable firms. As a special case of this result, I solve the model under costless precision and obtain a separating equilibrium to explicitly show that an inverse relation between intrinsic value and precision can exist. In the second study I develop a model to explain why firms voluntarily share with their rivals some (but not necessarily all) of the very basic knowledge that drives their profitability. I model a competitive setting and find that, in equilibrium, a firm may decide to share a portion (but not all) of its knowledge with a rival with the expectation that the rival will turn out to be less competitive than it otherwise would have been. This occurs because knowledge sharing reduces the rival's incentive to acquire additional knowledge by bolstering its confidence to the point where it no longer perceives a net benefit from learning anything new. These results are consistent with phenomena such as partially informative financial statement footnotes, selected operational information shared through benchmarking, and a constrained openness with customers, suppliers, and analysts.

Degree

Ph.D.

Advisors

Penno, Purdue University.

Subject Area

Accounting|Economic theory

COinS