This Article presents an economic model of corporate fraud arising from shareholder incentives. First, the model shows that a firm’s current shareholders have a preference for higher reported values. Current shareholders are, in expectation, net sellers of the firm’s shares; a higher reported value of the firm increases current shareholder returns in expectation. Second, these preferences for inflationary misreporting translate into equilibrium misreporting behavior, which generates inefficiencies due to asymmetric information among secondary-market traders. Informed traders undertake inefficient research costs, noise traders demand a discount to trade, and selling shareholders face deadweight illiquidity costs. Third, in general, some ex post penalty for misreporting can eliminate misreporting incentives and result in a unique truth-telling (that is, separating) equilibrium. This improves social welfare. With joint-welfare maximization among the firm’s initial stakeholders and unlimited liability, it does not matter on whom the penalty is placed. Finally, the specific mechanism of firm-level (or “vicarious”) fines has desirable qualities from the perspective of administrative feasibility: the optimal fine is a simple function of observable market data. Compensation does not affect this formulation, yet compensation may be desirable in the event of incomplete deterrence because it reduces asymmetric information liquidity costs. The same liability formula applies for alternative targets of liability, such as the manager, and the approximate magnitude of the optimal fine remains the same; however, judgment-proof defendants and limited liability may militate toward firm-level fines.