Essays on Financial Economics

Abstract

This dissertation examines whether investors’ cognitive limitations can influence market prices and, if so, how this affects corporate decisions. In the first chapter, I investigate whether and how investors’ cognitive limitations influence market prices. I study this issue in the context of an accounting rule that enabled firms to avoid recognition of a compensation expense against reported earnings. I begin by noting that, because of the high cost associated to processing information, it is reasonable for an individual investor to make buy or sell decisions based on a few salient signals, such as front-page earnings. Following this premise, I argue, overstated earnings that lead investors to form over-optimistic views of a firm should result in the overvaluation of the firm’s stock under the presence of conditions that limit arbitrage.In this chapter, I provide evidence in support that market prices did not incorporate this compensation expense, which did not lead to a reduction in reported earnings but needed to be disclosed in a footnote to the financial statements. Consistent with the argument that this result is driven by investors’ cognitive limitations, I find that firms’ use of employee option grants –the form of compensation in question– and abnormal return performance are associated to the composition of firms’ investor base. Specifically, I find that these two are positively related to the presence of individual investors. Furthermore, I find that these are positively correlated to the presence of institutional investors with a short-term investment horizon, or transient investors, according to the classification developed in Bushee (1998). These results challenge the widespread notion that prices incorporate all publicly available information not only in the absence of well-funded rational investors, or arbitrageurs, but also in their presence.The results in Chapter 1, however, are subject to potential criticisms. Measuring market efficiency is a treacherous endeavor, being the “bad model problem” the chief challenge. As noted in Fama (1970), market efficiency can only be tested jointly with an equilibrium asset-pricing model characterizing normal returns. Therefore, the presence of abnormal return performance can be interpreted either as evidence against the hypothesis that market prices incorporate all publicly available information, the inappropriateness of the equilibrium pricing model under consideration, or a combination of these two. To mitigate concerns related to this issue, I propose three complementary methods to assess the informational efficiency of market prices: a calendar-time portfolio strategy, a regression analysis of abnormal returns, and an event-study around earnings announcements. Although the results of these analyses are all consistent with the behavioral view, it is not possible to fully reject the alternative that the “bad model problem” can at least partially explain the evidence here documented.After showing that due to the favorable accounting treatment of option grants firms had a stock-price motivation to use this form of compensation, in Chapter 2, I examine whether this affects firms’ behavior. Specifically, I explore whether it shaped CEO option-based compensation during the period under study and provide evidence that sustains that this is the case. These findings help explain the sudden shift away from executive option grants observed in 2002, and by doing so they contribute to fill in a void in the literature, which has so far not been able to determine the causes of this phenomenon (Frydman and Jenter, 2010). Existing studies have examined how the change in the accounting for options affected CEO compensation. Differently, in this chapter, I provide evidence that the most significant changes in firms’ CEO option-based compensation took place in 2002 –following the string of accounting and corporate scandals that rocked corporate America– well in advance of the mentioned change in the accounting rule. This finding is particularly important for studies that examine the effect that rules and laws have on corporate policies. As this dissertation demonstrates, in these studies, it is of first order importance to consider the events that lead to the implementation of any rule or law because the actions of lawmakers and rule-setters are rarely exogenous or unanticipated.