The product market opportunity loss of mandated disclosure

Abstract

In the interests of protecting decision makers in the financial markets, the Securities and Exchange Commission requires publicly traded companies to publicly disclose certain accounting information. Such disclosure requirements however create a potential opportunity loss. They may destroy firms’ opportunities for implementing an alternative information acquisition and exchange regime – one that would optimize the firms’ product market profits. Nevertheless, we show that despite the previous imposition of an opportunity loss, firms may still favor future increased disclosure requirements. Finally, when information cost declines, although firms’ welfare may decrease, their desire for increased disclosure requirements always strengthens.

Introduction

Companies publicly disclose accounting information such as a balance sheet and an income statement every quarter. Such disclosures are mandated and regulated by the Securities and Exchange Commission (SEC), and used by (potential) shareholders and bankers to make investment decisions.¹ These accounting disclosures are also observable by competitor firms and used in making decisions such as setting production quantities and prices.² Consequently these disclosures may have significant effects on a firm’s profits. The view taken here is that disclosure regulation by the SEC/FASB may preclude firms in an industry from creating a jointly optimal information sharing program, or from committing (explicitly or implicitly) not to exchange information. As such, mandating disclosure standards imposes an opportunity loss on firms’ expected profits earned in the product market.

The following research questions are addressed. In the absence of mandated disclosures, what level of information acquisition and disclosure would firms optimally jointly select? How do these optimal levels of acquisition and disclosure depend on industry parameters? What is the opportunity loss suffered when disclosure is instead mandated? How do firms view subsequent changes in disclosure levels? How does diminishing information cost affect firms’ welfare and firms’ attitudes toward additional disclosure requirements? Finally, can firms recover this opportunity loss?

A stochastic oligopoly model is used to gauge the impact on firm expected profits of changing information regimes and to address these questions. We assume many different industries, parameterized by differing demand and cost coefficients, but each composed of identical firms. Each firm purchases costly private information about an unknown industry-specific market demand parameter. An industry’s information regime consists of all firms each acquiring m conditionally independent private observations about the unknown market parameter (via the installation of an internal management accounting system) and subsequently each disclosing f randomly selected signals from the set of m.³ For example, revenues earned, order backlog data and revenue projections and forecasts can all be considered noisy observations of future market demand. The greater the proportion of these observations which are shared or disclosed the more accurate are firms’ post-disclosure information sets (with respect to the unknown market parameter), and the less private information there remains in the market following disclosure, i.e. the greater the correlation between firms’ post-disclosure information sets.⁴

The analysis shows that the optimal level of mandated disclosure is either full disclosure of all acquired information or no disclosure. Surprisingly, partial disclosure (disclosure of some of the acquired information) is never optimal in this setting. As expected, the optimal level of information acquired is a decreasing function of the cost of information. However this optimal level ranges discontinuously from no information to the acquisition of perfect information, with the discontinuity occurring as the optimal disclosure level switches from none to full. Thus the optimal mutual disclosure regime may be industry specific: for some industries the optimal disclosure regime may be the acquisition and full disclosure of a few observations. For other industries the optimal disclosure regime may be the acquisition of many observations with no subsequent disclosure. These results represent an unconstrained optimum in the problem of cooperative information regime choice on the part of firms in a particular industry.

It is in this environment that we consider the impact of disclosure mandated by the SEC/FASB. Such mandated disclosures compromise the ability of firms to implement the optimal information regime, and insodoing create an opportunity loss. In this way this research attempts to evaluate disclosure choices within the context of other information and other information regimes. We show that this opportunity loss consists of two components: the first due to the mandated disclosure level not being the optimal full disclosure level, and the second due to full disclosure potentially not being optimal.

Further, we show that his notion of opportunity loss is helpful in formulating the FASB/SEC’s problem by providing a well-defined way of comparing competitive effects of a given disclosure level across industries. For industries in high information cost environments, ceteris paribus, the impact of increasing disclosure levels is always to exacerbate firms’ opportunity loss, while for those in low information cost environments, there is a desire for increased disclosure levels. For industries with intermediate levels of the information cost parameter, firms prefer more stringent mandatory disclosure levels if the existing levels are low and less stringent if the existing levels are high. Noteworthy, is that even in industries for which no disclosure is actually optimal, we see that firms might still prefer increased disclosure requirements.

We also consider the impact of changes (specifically reductions) over time in the information cost parameter. We show that although the attitude toward increasing disclosure requirements strengthens for firms in all industries as information cost decreases, the actual welfare of firms in some industries will increase and others decrease.

Finally, we consider the possibility that firms within an industry may cooperatively be able to recover some of this opportunity loss. Full recovery is possible only for those industries in which full disclosure is optimal and the mandated disclosure level is below the optimal full disclosure level. In general, however, full recovery is not possible, since if the mandated disclosure level is greater than zero, then the optimal information regime choice is now a constrained version of the original problem. No recovery is possible for those industries in which full disclosure is optimal, but the mandated disclosure level already exceeds the optimal full disclosure level. Partial recovery however is feasible in some cases, either by acquiring and sharing (disclosing) more information than mandated, or by acquiring additional observations but keeping them private: in other words by implementing a partial disclosure regime. This contrasts with the earlier result that partial disclosure is never optimal.

The paper is organized as follows. A brief literature review completes this section. The product market game is presented in Section 2 and solutions derived for equilibrium output and generic expected profit levels as functions of the information regime. In Section 3 the jointly optimal information regime is derived in the absence of mandated disclosure. In Section 4 we examine properties of the opportunity loss created when the SEC/FASB mandates a particular level of disclosure. Concluding remarks are given in Section 5.

The industrial organization literature has examined incentives for information sharing.⁵ Typically, expected equilibrium profits are compared for a limited set of information sharing alternatives: sharing all or no private information. Sharing noisy market demand information when cost functions are linear and firms are Cournot competitors reduces expected firm profits. See Clarke (1983).⁶ However, Vives (1984) shows that expected profits are higher when sharing single signals in a Cournot duopoly when the goods produced are complements. Kirby (1988) shows that complete sharing is preferred to keeping information private when cost functions are sufficiently convex. Since firms are better off as the accuracy or precision of their post-disclosure information sets increases, and worse off as the correlation between firms’ post-disclosure information sets increases, one might conjecture that partial disclosure (which retains some private information for each firm) might be the preferred disclosure regime.

Regarding partial information sharing previous research provides conflicting results. In a simplified setting, Novshek and Sonnenschein (1982) allowed for partial information sharing but showed that equilibrium expected profits were undominated only for the extremes of no disclosure and full disclosure of all observations and furthermore that both extremes lead to identical levels of expected profits. By contrast, Kirby (1993) showed that partial information sharing in the form of multiple information sharing pools is optimal in some cases.⁷

The question of information acquisition was examined by Fried (1984) using a duopoly model which allowed for asymmetric, binary choices regarding costless production and disclosure of cost information by the duopolists. Such duopolists are motivated to disclose information, in turn requiring the costless acquisition of the information. Li et al. (1987) analyzed the case of endogenous research by Cournot oligopolists (without the possibility of subsequent information sharing) and showed inefficiencies relative to social welfare maximizing levels. The inefficiency was attributed to a lack of information pooling.⁸

Accounting researchers have used stochastic oligopoly models of information sharing to examine the ex ante product market effects of alternative mandated accounting disclosures. Hughes and Kao (1991) show that full disclosure (interpreted as selective capitalization of R&D expenses) produces greater expected profits than does partial disclosure (interpreted as immediate writeoff). Feltham et al. (1992) show that under conditions of Cournot competition and demand uncertainty, multisegment firms receiving private information are worse off under line of business reporting than under aggregate reporting, but that expected consumer surplus and expected social welfare are greater under line of business reporting.⁹

The current paper differs from those mentioned above in that it simultaneously considers the problems of selecting information acquisition and information disclosure levels. Moreover, it views the product market effects of mandatory accounting disclosures in the context of existing information arrangements, and to this end uses an opportunity loss notion of firm welfare.