Ebook Auditing, Governance And Reporting: An Experimental Investigation
The need to avoid future audit and governance failures such as Enron, Tyco and World-Com has renewed the interest of, regulators, executives, auditors, investors and academicians in understanding the role of auditing in improving transparent reporting and reducing expropriations by managers. Managers can expropriate investor wealth in at least two ways. First, by understating the true performance of the firm, managers can expropriate the difference between the “achievable” and “achieved” performance. For example, in a study spanning 31 countries, Leuz et al. (2003) argue that if the governance and/auditing is weak, managers of the firm engage in the consumption of private control benefits such as related party transactions, empire building and other “hidden” transactions that effectively transfer wealth from investors to managers.
The second way in which managers transfer wealth from investors to themselves is by earnings management, a notion that is supported by extensive empirical literature (Healy and Wahlen, 1999; Bruns Jr. and Merchant, 1990; Burgstahler and Dichev, 1997; Burgstahler and Eames, 2003; Christie and Zimmerman, 1994; Dechow et al., 1996; DuCharme et al., 2004; Richardson, 2000). Numerous empirical studies have examined the effectiveness of audit and internal governance mechanisms in restraining such expropriation and opportunistic earnings management behavior of managers. These studies have ranged, from the studies on auditor independence (Frankel et al., 2002; Ashbaugh et al., 2003; DeFond et al., 2002; Larcker and Richardson, 2004), to the effect of the internal corporate governance mechanism such as board independence and audit committee structure on managerial reporting behavior (Klein, 2002a; Klein, 2002b; Becker et al., 1998). Srinidhi and Sen (2007b) show that weakening of the market-based corporate control mechanisms in the form of adopting poison pills increases earnings management and decreases the value-relevance of earnings. Further, there is evidence that managers increase their compensation after poison pill adoption (Srinidhi and Sen, 2007a; Bebchuk et al., 2002). The overall findings of these studies support the contention that audit quality and strong governance mechanisms restrain earnings management and other expropriating behaviors of managers but managers try to weaken these mechanisms by creating economic bonds with the auditors and by adopting poison pills (Gompers et al., 2003; Srinidhi and Sen, 2007b).
We use a controlled laboratory setting to experimentally examine the role of auditing and market based governance in restraining managerial expropriation and inaccurate financial reporting. Our study is motivated by two primary considerations. First, empirical studies are limited in their ability to manipulate audit and governance variables in a controlled manner and in isolating the contexts under which the effect of auditing can be investigated. A laboratory-based controlled experiment involving human participants overcomes these limitations by creating settings in which the experimenter can vary the incentives and the choice set of the participants. The experimenter can also introduce specific variations in the treatment variables, document the decisions made by the participants and evaluate the results under different regimes1. Second, few empirical studies in the literature have addressed the issue of private benefits of control and other non-financial perquisite consumption by managers in situations of weak audit and governance. One reason for this lack of empirical analysis is that the data on such private consumption is decidedly hidden from view and is neither accessible to the investor nor to the researcher. In such a situation, only an experimental investigation or an analytical formulation can yield insights that can help regulators and policy makers in devising policies to constrain such expropriation. By its very nature, analytical formulations need to make a number of assumptions that cannot be validated in practice. Even in the presence of an analytical analysis, an experimental investigation will offer a richer set of insights into the phenomenon.
Prior literature identifies three roles of auditing signaling by managers of private information to investors and other external parties; deep-pocket hypothesis which holds that auditors provide a means of recovering investment losses from bankrupt firms and reduce agency costs by providing assurance that accounting numbers are fairly and accurately presented (Chow et al. 1988). Of these, audit quality is mainly associated with the third role, namely the reduction of agency costs between investors and managers in the presence of information asymmetry. Clearly, auditing cost itself is an agency cost in this relationship. Agency costs do not benefit anyone and therefore, contracts are devised to minimize the total agency costs. To the extent that incurring audit costs reduces more of the other agency costs, it becomes part of the optimal contract. Our view of auditing derives from this framework and therefore, its main benefit is in reducing the information risk faced by investors in the presence of information asymmetry between investors and managers. The information risk in financial statements mainly stem from the accruals which are added to the underlying realized cash flows to determine the earnings. Managers have little discretion in reporting the realized cash flows per se. The accruals are determined both by the inherent characteristics of the firm as well as the discretion employed by managers in choosing accounting methods and estimates about the future.
Auditing reduces information risk in two ways. First, the auditors independently assess the reasonableness of the estimates and the appropriateness of accounting methods used by managers to compute the accruals. Based on the private information that they gather during the course of their audit(s), they seek economic justification for deviations of managerial estimates from their own independent estimates. Further they examine the design of and compliance with the internal control procedures to ensure that the numbers being produces truly represent the estimates of managers. To the extent that they cannot find adequate justification, they conduct detailed substantive tests to verify the reported amounts and negotiate with the managers to change the reported amounts to more justifiable amounts. The second way in which the auditor influences information risk is by having a credible threat of costly qualification. Managers, in order to avoid the possibility of such qualification, become more disciplined in their reporting and thereby decrease the information risk to the investors.
In a similar vein, governance practices restrain private consumption as well as opportunistic reporting behavior by managers. Governance practices could originate from the regulatory and legal framework (such as the board composition, the requirement of audit committee, the procedures for nominating and voting for directors, the handling of share-holder proposals etc.) or from the market in the form of corporate control (Martin and McConnell, 1991; Grossman and Hart, 1988; Dahya and Powell, 1998). While a number of empirical studies on governance have focused on the board characteristics, relatively fewer studies have investigated the market-based governance that comes from the market for corporate control. The set of studies that have examined these perspective have investigated the effect of poison pills and other anti-takeover provisions on the wealth and the governance of the firm (DeAngelo and Rice, 1983; Pound, 1987; Harris, 1990; Comment and Schwert, 1995; Datta and Iskandar-Datta, 1996; Borokhovich et al., 1997; Sundara-murthy, 2000). While the evidence is mixed regarding the wealth effects, there seems to be a general agreement that management gets entrenched by adoption of anti-takeover provisions and is thereby able to increase its expropriation (Bebchuk et al., 2004; Bebchuk et al., 2002).
Auditing and corporate governance mechanisms form two important factors that restrain managerial expropriation and reporting. The interaction between the board effectiveness and auditing is shown to be complementary (Carcello et al., 2002) in the sense that a stronger board demands higher levels of audit. Overall, the gist of empirical studies is that expropriation and opportunistic reporting behavior of managers are both curbed by strong auditing, effective boards and in the presence of a strong market for corporate control.
Our experimental set-up captures the basic elements of the above relationships. We first develop a theoretical framework to establish the expected relationships in a context which is more structured but less varied than the real world. For our analytical model we assume the existence of managers of different “types,” where the type space represent differences in inter-temporal preferences of managers. Relatively “patient” mangers have higher discount factors, while relatively “myopic” managers have lower discount factors. This provides us with some novel insights which are then tested through our experiment. The main contention from our theoretical framework is that under certain conditions there exists a separating equilibrium such that differences in managerial reporting behavior is driven by their inherent differences in inter-temporal preferences as captured through their discount factors.
In our experiment, in a multi-period setting, we allow the managers to report in every period a return that could be different from what they observe. Under-reporting results in direct expropriation similar to the real life phenomena of perquisite consumption, private benefits of control and empire building. This reduces the residual amount available for investment in the subsequent periods. On the other hand, over-reporting, particularly in the early periods, could help the managers in attracting more investments from different investors and increase the “firm” size and their own expected future compensation. This is akin to managers who have stock interest in the firm engaging in income-increasing earnings management that increases the stock price and results in greater compensation. In this scenario, investors are allowed to choose a costly audit and different levels. An effective audit results in the “true” earnings being revealed but the effectiveness of audit is determined by the level chosen by the investor. This is theoretically equivalent to an imperfect audit that decreases the extent of earnings management but does not eliminate it. We also have a treatment (poison pill) that does not allow the investor to switch his/her investment to another manager. This treatment simulates the entrenchment of the manager. In this setting, we study whether the investment levels drop to zero in the absence of an audit; the effect of audit on expropriation; whether higher audit levels are associated with higher levels of investments; and the effect of poison pills on the level of audit, the expropriation level and the overall welfare level.
The experimental results support our hypothesis regarding the differences in expropriation levels being driven by the presence of different “types” of managers. In early periods of all treatments we find relatively higher expropriation and audit levels (where applicable). But this metamorphoses into low expropriation and low audit in later periods through manager selection, where investors reward “patient” managers with more investment and punish “myopic” ones with bankruptcy. We also find that the investment level does not drop to zero in the absence of an audit. This is supportive of the contention that reputation and the need to attract more capital disciplines the managers to some extent and this allows the market to function in the absence of an audit. However, we also show that the level of expropriation drops with audit level; that higher investment levels are associated with higher audit levels; and the overall pay-off is higher when audit levels are allowed to vary rather than fixing them. We also find that firms with poison pills attract higher levels of audit but less capital. Further, we find that even with higher levels of audit, there is more expropriation in the firms with poison pills than in other firms.
Our findings contribute to the literature in two important ways. First, our results show that auditing is desirable but not necessary for the market to function. Second, these results confirm the empirical findings which show a lower level of expropriation and earnings management when governance and audit levels are high. Empirical findings suffer from the endogeneity problem, i.e., that governance and auditing decisions are not necessarily driving less expropriation and better reporting but rather, are simply the outcomes of the same underlying variables such as managerial ethics etc. By exogenously controlling the parameters in the experiment, we show that the empirical findings are not driven by endogeneity alone. Third, we show that the market-driven governance is a substitute rather than complementary to auditing. In other words, if there is a weakening of the market for corporate control, more auditing is demanded to substitute for the disciplining effect.
The rest of the paper is organized as follows. In the next section, we develop the theoretical framework. The third section develops and presents the hypotheses. In the fourth section, we provide the experimental design. The results and their discussion is given in Section 5. The last section concludes the paper.
Download
PDF Ebook Auditing, Governance And Reporting: An Experimental Investigation
Posted in :
