A classroom experiment on the benefit of auditing in financial markets.
In the experiment, students trade securities in a sequence of markets in which financial information pertaining to security values is made public. The quality of the information depends on whether it was audited, and varies from market to market. Data on transaction prices, security holdings, and investor profits from a typical sequence of markets are presented, as are a series of discussion topics designed to illustrate how to use the experiment and related data to address a variety of auditing-related issues.
In this paper, I discuss a classroom experiment that can be used to illustrate how auditing can impact investor behavior and financial market performance by providing assurance on the quality of financial information. In particular, the experiment demonstrates that auditing can influence security prices and investment portfolio compositions in a way that benefits investors and lowers the cost of capital for firms.
This exercise contributes to auditing education for the following reasons. First, the exercise was designed specifically to permit students to observe how auditing influences investor behavior in a set of financial markets in which securities are traded based on the contents of a financial report. The experimental environment is consistent with the typical presentation of auditing in textbooks and with the visible role of auditing in real-world financial markets. Accordingly, lessons from the experimental markets should be easily transferable to discussions based on textbook material as well as current events. Second, to facilitate discussion, I provide explicit guidance on how to utilize the experiment to demonstrate the economic value of auditing and to serve as a basis for discussing topics such as auditor independence, earnings management, reasonable versus absolute assurance, and investor confidence. Finally, the experiment requires minimal advance set up by the instructor, and can be conducted in less than one class period. Accordingly, this level of simplicity combined with the connection to financial reporting and financial markets makes this exercise an attractive choice for those who are interested in using experiments in an auditing class.
The next section of this paper provides background on research on the value of auditing and describes other classroom experiments on accounting topics. This is followed by a section that describes the experiment, and offers predictions as to what ought to happen. Then a set of discussion topics is provided along with results from a typical experimental session; the purpose of which is to illustrate how to utilize data generated from the experiment in classroom discussion. Finally, student feedback and concluding comments are offered. Detailed instructions that can be used to implement the experiment are included in Appendices A and B.
Traditionally, and in most auditing texts, auditing is viewed as a tool useful for reducing the occurrence of unintentional errors in reported financial information, and mitigating managerial bias in the recording and reporting of financial information (e.g., Robertson and Louwers 2002, 6; Rittenberg and Schwieger 2003, 3). The implication is that auditing financial statements is worth-while in part because it improves the quality of information available for investor decision making.
A considerable amount of academic research demonstrates how auditing creates economic benefits. In the broadest sense, this work has focused on two areas. The first area deals with contracting between owners (principals) and managers (agents). This work is based on the premise that owners, who are removed from the business, typically cannot directly observe managers' actions. This creates an opportunity for managers to take actions that are in their own personal interest, but which are inconsistent with the owners' objectives. Accordingly, performance contracts, which can rely on financial information, are written to align the incentives of owners and managers. Auditing plays a role by providing assurance that the financial information is reliable, and hence useful for contracting. Examples of theoretical and experimental work in this area include Antle (1982), Baiman et al. (1987), and Kachelmeier (1991).
A second area of research pertaining to audit quality deals with the broader role of auditing and transactions in organized markets. In this line of work, it has been demonstrated that auditing can improve the efficient allocation of resources between market participants, which in turn has a positive economic impact (e.g., Jensen and Meckling 1976; Blazenko and Scott 1986; Dopuch et al. 1989; Wallin 1992).
Simple experiments are an excellent means of integrating into the classroom the insights generated from these areas of work. The experiments can be basic enough for students to grasp readily, yet rich enough to allow students to develop subtle insights into the connection between auditing and economic concepts such as welfare and market efficiency (DeYoung 1993). (1)
One example of how experimental markets have been successfully implemented in accounting courses is Frischmann (1996). In this experiment, students participate in a market, which illustrates the role of implicit taxes in tax clientele formation, a topic addressed in graduate-level tax courses.
A second example of how experimental markets have been successfully implemented in accounting courses is an auditing experiment by Boylan (2000) (hereafter Boylan). In that experiment, students were assigned roles of sellers and buyers. Sellers could choose to sell either high-quality products (referred to as assets) or low-quality products, with high-quality products costing more to sell. In what were labeled "auditing markets," sellers' quality choices were visible to buyers, and buyers could make bids contingent on quality levels. In what were labeled "no-auditing markets," quality choices were unobservable to buyers. Classroom discussion focused on illustrating how economic performance was superior in the markets where the quality choice was observable to buyers (the auditing markets).
While that experiment illustrates an innovative way to teach students about the value of auditing, at least three features limit its usefulness in the classroom. The first limitation is the amount of time and effort required to set up and administer that experiment. In particular, two full 75-minute class periods are the norm for running the experiment. Discussion time is additional, and that experiment requires substantial set up by the instructor.
The second limitation is that the experiment is based on a research study on consumer choice in product markets (Lynch et al. 1986). As a result of the product market emphasis, the experiment is not particularly well-suited to addressing issues relating to financial statements and the demand for financial reporting (Boylan 2000, 49). Since most auditing texts focus rather heavily on financial statement auditing, and often motivate the subject in the context of financial markets rather than product markets (e.g., Rittenberg and Schwieger 2003, 12-16), the link between the lessons from the experiment in Boylan and the corresponding course material is not particularly strong.
The third limitation is that the experiment in Boylan provides only modest guidance pertaining to classroom discussion. Moreover, little of the guidance is specific to auditing or financial reporting.
In this paper, I present a classroom experiment that is designed to illustrate the economic impact of auditing, but which overcomes the three limitations previously identified. In particular, the following experiment is simpler and requires substantially less time and effort to set up and run than the one in Boylan. In addition, the experiment described in this paper offers a pedagogical improvement over Boylan because this experiment is based on a research study that was intended to address issues germane to investors and security markets (Plott and Sunder 1982), and thus is better suited to address issues specific to financial reporting and financial statement audits. In particular, participants in this experiment observe publicly released financial reports about a firm and then trade that firm's securities based on the information. Auditing, in this context, provides assurance on the quality of the information contained in financial reports, as opposed to providing assurance on the physical attributes of a product as in Boylan. This structure allows students to draw inferences about the economic impact of financial statement audits and audit quality in financial markets. Finally, this paper improves on Boylan by offering better and more explicit resources for classroom discussion. This paper offers eight specific discussion topics pertaining to the markets and to auditing, along with explicit guidance for managing class discussion for each topic. The topics are well developed, address current issues facing the accounting and auditing profession, and are relevant to audits of financial statements and the role of auditing in financial markets.
The objective of this experiment is to address how auditing can influence investor behavior in financial markets, and the attendant economic impact of such behavior. Two features make it desirable for use in the classroom. First, auditing is placed in a market context in which a firm releases financial information that can be used for trading. This structure mirrors the financial reporting process that occurs regularly in financial markets in the U.S. and around the world. Accordingly, the structure of the markets should be familiar to students, thus facilitating discussion on topics relevant to financial statement auditing, as well as making the experiment intuitively appealing. Second, students participate in "live" markets, in which they receive and process financial information, form strategies, and buy and sell securities for profit. Accordingly, they experience first-hand how auditing influences their strategies and decisions, as well as how it influences overall market dynamics.
Description of Markets
The markets used in this experiment are based on the work of Plott and Sunder (1982), who investigate the role of information in security markets (see also Davis and Holt 1993, 413-418). The experiment consisted of a four market sessions, each with three "investor types" and two groups of participants assigned to each investor type, for a total of six groups. At the beginning of each of the four market sessions, each group was endowed with a portfolio consisting of $10,000 in experimental currency and three shares of a generic security.
The value of those shares was determined by a draw from a deck of playing cards. If a black (red) card was drawn, the shares were worth the higher (lower) of two amounts depicted in Table 1, with the specific values depending on investor type. The outcome of the draw remained private (unobservable to the groups) until the conclusion of the experiment. A given type of investor group was aware of its own security values from Table 1, but was not aware of the values to other groups. However, it was common knowledge that multiple investor types existed, each with unique security values.
At the beginning of each of the four market sessions, a report that provided information about the outcome of the private draw (and hence security value) was released to the market. Qualitative characteristics of this report were manipulated between the sessions, and are described in detail below. Subsequent to the release of the information about share value, groups could trade shares with one another for up to seven minutes. Trades were arranged via an oral double-auction, in which groups could transact for one share at a time (details of the auction process are provided in the instructions included in Appendix A).
Information Structure and Manipulation
As indicated above, the experiment consisted of four separate sessions, and a report containing financial information was released in each session. The four sessions were divided into two pairs, with each pair referred to as "Market A" and "Market B," respectively. In each market pair there was one session in which the outcome of the private draw was a black card, and one in which the outcome was a red card. The order (red/black or black/red) was determined at random, and not revealed to participants until after the experiment. The two markets (A and B) differed based on the characteristics of the financial information contained in the public report. The correspondence between the public report and the private draw in both markets is depicted in the instructions provided to participants, which are included in Appendix A, and summarized below.
In Market A, the publicly available financial information always indicates that the securities are worth the high values from Table 1, regardless of the outcome of the private draw. As such, the information is biased in the sense that if one takes it at face value, one will either accurately estimate share values or overestimate share values, but never underestimate share values. Second, because there is no variation in the report (it always signals "black"), it is uninformative; it should not be used to update one's beliefs about the outcome of the private draw. Although no institutional or contextual detail is available to students at this stage of the experiment, the public report is meant to possess characteristics that are consistent with a set of unaudited financial statements that are prepared with an optimistic bias, which potentially contain misstatements, and which provide little predictive value for users. In short, Market A represents a worst-case scenario for auditing and the financial reporting process. This was done to establish an endpoint and to facilitate comparison with Market B. (2)
In contrast, in Market B, the publicly available financial information always corresponds to the outcome of the private draw. When the private draw generates high security values, the publicly available financial information reveals this. When the private draw generates low security values, the publicly available financial information reveals this as well. Accordingly, the financial information released to the investors in Market B is unbiased in the sense that if one takes it at face value, one should neither systematically overestimate nor underestimate share values. Second, because the public information always corresponds to the private draw, the report is perfectly informative; it faithfully reveals the outcome of the private draw. Accordingly, in Market B, the information structure is meant to provide a best-case representation of the auditing and financial reporting process. In other words, it represents a setting in which auditing is effective at mitigating management bias and preventing and detecting material misstatements in the financial statements, making them transparent and highly useful to investors. (3) However, the reader should not be distracted by the fact that auditing is perfect in this market. This is a convention to make the experiment easy to understand, and to facilitate comparison with Market A.
Participants and Procedures
The experiment described in this paper was conducted at a small, private university with a national student base, and has been replicated on five separate occasions. The total number of student participants in all five replications is approximately 85. Participants were primarily senior-level undergraduate accounting majors. Class sizes ranged from as few as nine to as many as 28 students. (4) The experiment, including preliminaries, lasts about 45 minutes. (5)
Upon arrival in class, students were divided into six groups, (6) and informed that they would participate in an experiment consisting of four market sessions, each lasting seven minutes. Each group received a set of written instructions (included in Appendix A), the contents of which were reviewed aloud by the instructor. Groups were informed that they would be endowed with a portfolio consisting of 10,000 experimental dollars and three shares of a generic security at the beginning of each market session. Groups were also provided with information about their own share values, and were informed that different groups might have different share values. In addition, groups were informed that they would have the opportunity to make profits by trading those securities after observing a financial report that provided information on security value, and that their ending balances in experimental dollars would be converted to chances of winning a $25 prize at a prespecified rate. (7) Groups were informed that they could make verbal offers to sell (asks) or buy (bids) one share at a time. In addition, they were informed that they could accept any outstanding bid or ask, which would consummate a trade. All activity was recorded in T-accounts on the chalkboard, for all to see. Details about the financial report and how it differed for Markets A and B were provided in the instructions. Finally, groups were informed that the Market A information structure would be in place for the first pair of sessions, followed by the Market B structure for the second pair of sessions. (8) At the conclusion of each session, each group recorded its trading profits on a summary sheet.
The structure of share values depicted in Table 1 allows one to make several predictions about behavior in the markets. For instance, when a black card is drawn, shares are worth the most to Type I investors. Accordingly, in Market B, when the public report indicates "black," all participants should correctly infer that security values are high. Since shares are worth $300 to Type I investors in this case, and only $250 and $180 to Type II and III investors, respectively, it follows that Type I investors should purchase shares from the other two types of investors. If the purchase price is between $250 and $300, then all investors who trade will increase the values of their investment portfolios. Similar logic suggests that in Market B, when the public report indicates "red," Type III investors should purchase all of the securities for between $150 and $170, once again allowing all investors to profit, relative to their initial positions.
In contrast to Market B, the public report in Market A, which always indicates that the private draw was a black card, is biased and uninformative to investors, who know that there is an equal chance that the private draw produced a black or red card. Accordingly, rational investors should not rely on the contents of the financial report, and should trade based on the best information they have, which is the expected share values. In this case, since Type II participants have higher expected values ($200 from Table 1) than their Type I or Type III counterparts ($150 and $175, respectively), the outcome that generates the greatest amount of expected wealth, ex ante, would be for Type II investors to purchase all of the securities.
The preceding discussion indicates that participants collectively should generate more wealth via trading shares in Market B than in Market A. More specifically, in Market A, if Type II participants collectively acquire the securities, then a long-term average share value of $200 each will be obtained (0.5*$250 + 0.5*150). In Market B, Type I investors are predicted to accumulate shares when the public report indicates black, and Type III investors to accumulate shares when the public report indicates red. This would lead to a long-term average share value of $235 (0.5*$300 + 0.5*$170). As a result, average share prices would be 17.5 percent higher in Market B than in Market A. Finally, this should be good news to firms because the higher prices implicitly lower the cost of capital. (9)
RESULTS AND DISCUSSION TOPICS
The objective of this section is to present a set of results consistent with what one can expect to observe if one replicates the experiment, and to illustrate how one can use data generated from a replication of the experiment as a basis for discussion. (10) Prior to discussion, students should be provided with the results from their experiment (either by handout or by mini-lecture) along with the information in Table 1. The following results are based on an experiment run recently, and should be easily replicated.
Figure 1 displays the transaction price for each trade in each of four successive market sessions (A A B B). (11) The figure shows that transaction prices hovered around $190 per share in each session of Market A. In contrast, in Market B, when the public report indicated that the private draw was black, transaction prices quickly elevated to about $275 per share. Similarly, security prices fell to about $160 per share when the public report indicated that the private draw was red. Collectively, the data presented in Figure 1 indicate that the contents of public report in Market B influenced the prices at which investors were willing to transact. In addition, the data indicate that investors were willing to pay more for securities in Market B than in Market A after observing a public report that indicated that security values were high.
[FIGURE 1 OMITTED]
Figure 2 summarizes the data on the number of shares held in the ending investment portfolios for each type of investor. As predicted, it indicates that in Market B, most of the securities were held by Type I (Type III) investors when the public report indicated a black (red) card was drawn. In contrast, Type II investors held far more securities in Market A than they did in Market B. This also is consistent with the outcome predicted earlier in the paper. (12) The data presented in Figure 2 collectively indicate that the composition of investment portfolios was influenced by the contents of the public report in Market B. In addition, the data show that a public report indicating security values were high induced Type I investors to buy in Market B but not in Market A.
Figure 3 summarizes the data on average ending wealth in excess of the initial $10,000 cash endowment for each type of investor for each of three cases: Market A, Market B, and no-trade. The no-trade benchmark is useful because it shows how each investor type would have fared, on average, if they had simply held their initial endowments. Thus, it serves as a baseline for measuring the impact of trading with limited information (Market A), and the impact of the improved information quality available in Market B. The most telling feature of Figure 3 is that each type of investor did the best in Market B. Accordingly, not only did investment decisions change between Markets A and B (Figures 1 and 2), but decisions changed in a way that allowed all types of investors to be better off in Market B relative to Market A. This occurred because the information structure in Market B facilitated trades that got the securities in the hands of the investors with the highest valuations, ex post (Type I when the draw was black, Type III when the draw was red). Accordingly, more wealth was created in Market B, which implies there are more profits available to all investors.
[FIGURE 2 OMITTED]
The preceding results addressed individual and market performance without the added context of auditing or financial reporting. After presenting the students with this data, discussion can turn to auditing-related issues.
1. Explain how the private draw of the playing card and the related disclosure of the public report in this experiment relate to the financial reporting process in real-world financial markets.
In the experiment, the private draw (red or black) represents the actual payoff to be received at the end of trading and corresponds to the true or intrinsic value of a firm's shares. The difficulty in observing the outcome of the private draw corresponds to the difficulty in ascertaining a firm's intrinsic value. Intrinsic values are difficult to observe because of the fact that firms are going concerns, and that their value is in part dependent on future earnings (or dividends or cash flows), which in turn depend on numerous complex factors such as strategy and tactics, and environmental factors such as the economy and regulation, etc. Moreover, accounting conventions that prohibit or delay the recognition of assets and related revenues, or otherwise bias the measurement of their values, also create difficulties for investors who wish to use financial statement information as a basis for security valuation.
The financial report released to the market in the experiment was meant to provide information about the value of the firm's shares. In Market A, this information was not useful because the public report always indicated that the private draw was black. Thus, one could not tell whether the private draw really was black or whether it was red. In contrast, the financial information released in Market B correlated perfectly to the private draw; fully revealing the final payoff. This process is comparable to the financial reporting process because financial statements, much like the report in the experiment, are released to the public regularly and used by investors as a basis for trading decisions.
[FIGURE 3 OMITTED]
2. In which market does the public report appear to have been audited? Explain.
Audits are designed to detect material misstatements resulting from either unintentional errors or management bias. In Market B, the public report appears to have been audited because it gives an unbiased picture of the firm's prospects. Second, the public report in Market B reveals the underlying economic status of the firm with less of a chance of error than is the case in Market A.
In contrast, in Market A the public report always indicates that the private draw was black, which implies that security values are high. Accordingly, in Market A, the firm's prospects are always presented in the best possible light. This is consistent with the notion of management bias that is often cited in texts as one source of demand for auditing (e.g., Messier 2003, 5-7).
3. Use Markets A and B to make an argument for why investors benefit from audited financial information in financial markets.
Joint consideration of Markets A and B demonstrates the economic value of auditing. Market B is consistent with a setting in which an audit has provided investors with assurance that publicly available financial information is unbiased and informative about underlying security values. In Market A no such assurance exists. Data from the experiment demonstrates that in the absence of assurance that financial information is unbiased and reliable, investors did not rely on that information. In turn, this affected what they were willing to pay for securities as well as the composition of their investment portfolios. Figures 1 and 2 indicate that the level of assurance that investors had about the reliability of the information in the public report influenced decisions in the marketplace; it influenced transaction prices as well as security holdings. Figure 3 indicates that assurance about the reliability of the information affected decisions in a way that provided systematic benefits to investors. In particular, it promoted the allocation of securities to individuals with the highest valuations whether the underlying security value was high (black) or low (red).
4. Use Markets A and B to make an argument for why firms whose securities are publicly traded benefit from providing audited financial information.
By comparing the average price at which securities trade in each market, one can verify that the firm whose securities are traded in the experiment benefits from having those securities traded in Market B. In Market A, the weighted average price over both sessions was $189. In Market B, however, the weighted average of security prices was $203, or 7.4 percent higher than in Market A. (13) This is beneficial for the firm in Market B, because the higher security prices imply that its cost of equity capital is lower (e.g., it can raise more money in exchange for fewer shares in Market B). Thus it appears that a win-win situation exists in Market B; not only are investors better off, so is the firm. The implication from the experiment is that audited financial information supported systematically higher security valuations than unaudited financial information, thus benefiting the firm by lowering its cost of raising equity capital.
5. Can auditing provide absolute assurance that financial statements are free of material misstatements, as is the case in Market B? Explain.
This question is intended to generate discussion on the limitations of audits, and the potential impact of those limitations. In particular, several practical constraints preclude auditors from providing absolute assurance that financial statements are free of material misstatement. Time constraints preclude exhaustive investigations of financial reports. Human error can result in well-designed audits failing to detect material misstatements, as can collusion by management. The risk that the auditor fails to identify material misstatements (and modify its audit opinion accordingly) due to these and other limitations is known as audit risk, and is managed by the auditor. The resulting standard of assurance provided by audits is "reasonable" assurance rather than absolute assurance.
Students can explore the economic implications of limited assurance by comparing the results of Markets A and B. In Market B, auditing provided absolute assurance about the financial information pertaining to security values. In Market A, no assurance was provided. As was noted in connection with discussion Questions 3 and 4, the level of assurance provided in Market B relative to Market A yielded both higher profits for investors and a lower cost of capital for firms. Accordingly, the experiment suggests that as the level of assurance about financial information released in the marketplace diminishes, investors become less reliant on that information. This, in turn, can result in lower investor profits and higher capital costs.
Some students might argue that since audits provide reasonable rather than absolute assurance, investors and firms will always benefit from increasing the level of assurance provided by audits. This line of reasoning, however, fails to consider the costs of increasing assurance levels. For instance, if the cost of providing the assurance levels in Market B relative to Market A exceeds the additional wealth that is created, it does not make economic sense to consume the resources necessary to provide the audit and related assurance.
Finally, this question can be used to discuss how the Sarbanes-Oxley Act of 2002 and the related creation of the Public Company Accounting Oversight Board are expected to alter the costs and benefits of providing financial statement audits. These regulatory changes are intended to promote increased investor confidence in the information contained in financial reports. One example of the potential for increased costs related to these changes is the new requirement to provide assurance on the effectiveness of an entity's internal control structure, which could lead to additional audit work.
6. Use Markets A and B to discuss why investors should be concerned about whether auditors are independent.
This question can be used as a basis for discussing the concepts of independence in fact and independence in appearance. Independence in fact suggests that auditors will not place the interests of management above the interests of other parties such as investors. This suggests that an auditor who is independent in fact is unlikely to accept financial reports that are known to contain a biased representation of the firm's performance and financial position beyond the scope of what is permitted by generally accepted accounting principles. Accordingly, the auditor provides reasonable assurance of detecting material misstatements resulting from management bias. This is consistent with the information structure in Market B, in which the public report does not systematically overstate value. The overstatement bias and lack of informativeness present in Market A is absent in Market B; hence making the information provided in the financial report in Market B more reliable and useful to investors, which leads to the economic benefits outlined in discussion Questions 3 and 4.
Independence in appearance relates to investors' perceptions about the auditor. For instance, if investors presume that an auditor has a financial interest in its client due to lucrative consulting contracts, etc., investors might conclude that the auditor has a vested interest in permitting biased financial reports to be issued. If investors believe financial statement information to be tainted due to a combination of management incentives and a perceived lack of auditor independence, they will likely decrease reliance on the information. Market A illustrates what might happen if investors reduce their reliance on financial statements because of a fear that auditors have permitted management to issue statements containing biased and unreliable information. In particular, comparing profits and the cost of capital in each market demonstrates the sub-par performance in Market A. The lesson is that if there is the appearance of an independence problem, investors might defend themselves by decreasing reliance on financial statements. In the experiment, this type of investor behavior had a negative impact on security prices, which, in turn, stunted economic growth and raised the cost of capital for firms.
7. Discuss the economic implications of earnings management in the context of Markets A and B, and explain the connection between earnings management and audit quality.
This requirement can be used to help students understand the potential economic consequences of abusive earnings management policies. (14) Details of abusive accounting practices by companies such as Enron, WorldCom, Sunbeam, Cendant, and Xerox have become common knowledge in recent years. These companies have used practices such as improper revenue recognition, improper restructuring charges, etc., to manipulate earnings. Moreover, auditors of these companies have either failed to detect these problems on a timely basis, or have (at least initially) concurred with their treatment. These audit failures potentially have a number of root causes, such as the inherent limitations in audit technology that result in audit risk (see Question 5), lack of independence (see Question 6), and others.
Earnings management can be viewed in the context of Markets A and B as follows. Market B is consistent with a setting in which auditing is of very high quality; preventing and detecting material misstatements associated with abusive accounting practices and helping make the financial reporting process as transparent as possible. Market A is consistent with a setting in which investors lack confidence either in the quality of the financial information provided or in the level of assurance provided about that information. As noted, this lack of confidence might stem from the use of abusive accounting practices by management. Or, it might stem from audits that are either ineffective or perceived to be ineffective when it comes to preventing and detecting malfeasant reporting practices.
Comparing Markets A and B provides perspective on the potential economic consequences that result if auditors knowingly or unknowingly permit abusive earnings management practices to taint the contents of financial reports. The lesson from this is that if investors believe that the integrity of the financial reporting process has been impaired due to earnings management and ineffective audits, then their rational response will be to decrease reliance on the financial information provided. In the experiment, this leads to lower investor profits and higher capital costs for firms.
8. What mechanisms exist to help promote high-quality auditing?
The premise for this question is that the benefits that accrue to investors and firms in Market B (as compared to Market A) depend on the fact that the financial reporting process provides reliable information that is useful to investors. Moreover, prior discussion has suggested that audits can be an effective tool in preventing and detecting material misstatements in financial statements resulting from unintentional errors or biased reporting by management. This question departs from the experiment, and serves as a though exercise to help students understand the role of certain mechanisms designed to help ensure that audits are in fact effective.
First, the auditing profession itself has borne a significant share of the responsibility for promoting and monitoring audit quality. Until recently, the Auditing Standards Board of the profession's trade organization, the American Institute of Certified Public Accountants (AICPA) has been in charge of establishing authoritative guidance addressing items such as auditor qualifications, audit planning, and audit execution that is to be followed by auditors. However, the Sarbanes-Oxley Act of 2002, which was created in the aftermath of numerous highly publicized accounting and auditing scandals, has a resulted in a fundamental shift away from self-regulation. In particular, standardsetting authority now rests outside of the profession with the newly created Public Company Accounting Oversight Board, an independent body created to oversee audits of public companies.
Second, legal liability provides investors with a channel to monitor and discipline auditors who have failed to conduct audits with the proper degree of professionalism that is expected of them; thus promoting quality audits and the aforementioned economic benefits (e.g., Dopuch and King 1992; Wallin 1992).
Finally, auditors who develop reputations for providing high-quality audits that reduce the cost of capital for firms (as they did in Market B) should be able to charge higher fees than counterparts who do not enjoy similar reputations. Accordingly, the audit marketplace can serve as a mechanism that encourages auditors to develop reputations for providing high-quality audits so that they can reap the related economic benefits (e.g., Simunic 1980; Simunic and Stein 1987).
This section reports student feedback on the experiment. Formal feedback was solicited via a post-exercise questionnaire completed by 17 students from the most recent class in which the experiment was administered. Student responses are measured on a five-point scale, with 5 = strongly agree, 4 = agree, 3 = neutral, 2 = disagree, 1 = strongly disagree. When asked whether the experiment was effective in improving student understanding of the relation between the quality of financial information available to investors and the creation of value, the mean response was 4.24/5.0. When asked whether the experiment was an effective platform for discussing the role that auditing can play in influencing the quality of reported financial information, the mean response was 4.29/5.0. When asked whether the experiment was an effective platform for "addressing the economic consequences of topics such as audit quality, auditor independence, and investor confidence," the mean response was 4.18/5.0. Finally, when asked whether the experiment should continue to be used in future sections of the auditing course, the mean response was 4.71/5.0. The picture that emerges from the responses to these questions is that the students believe that the experiment is a useful learning tool and a valuable part of the auditing course.
This paper builds on previous work that illustrates how classroom experiments are effectively used in accounting and auditing courses. The experiment described herein is particularly well suited to the auditing classroom because of its ease of implementation, its brevity (it can be run in about 45 minutes), and because it explicitly incorporates financial reporting as a means of communication between a firm and its shareholders in a market setting.
In the experiment, students trade securities in a sequence of market periods in which the quality of the financial information available for decision making depends on whether it was audited. By design, the experiment permits students to make direct cause-and-effect inferences about how auditing influences investor behavior, investor profits, the cost of capital to firms, as well as overall economic performance in a financial market. Accordingly, the experiment can be used to motivate the demand for auditing, or it can be used as a basis for discussing the importance of auditing to the functioning of financial markets. Finally, it provides a natural platform for discussing issues that are of current interest, such the economic impact of investor perceptions about audit quality and auditor independence.
In this experiment we are going to conduct a series of markets in which you will have the opportunity to buy and sell shares of a fictitious company. The experiment will consist of multiple rounds of buying and selling under varying conditions. You will find several "Information & Record Sheets" as well as a "Profit Sheet" along with these instructions. These sheets will help you keep track of your earnings as well as determine the value of any decisions you make. The information on these sheets is private. Please do not share it with anyone.
The currency used in this experiment is called "experimental dollars." At the end of the experiment your experimental dollars will be converted to lottery chances or a cash prize at a pre-specified rate. The higher your ending balance in experimental dollars, the greater your chances of winning the cash prize.
Your profits come from two sources: (1) from selling shares to other groups, and (2) from holding shares at the end of each market period (the value of these shares will be announced at the end of the experiment). During each market period you are free to purchase or sell as many shares as you wish, provided you follow the rules below. For each share you hold at the end of a market period, you will receive one of two amounts (the process will be explained later). Please refer to your Information & Record Sheet for these amounts. Recall that this information is private, and that different groups may have different values. Please do not reveal this information to anyone. The total value of shares held at the end of a market period is simply the number of shares you hold multiplied by the share value listed on your information and record sheet. For example, if you hold five shares and each share is worth 10 experimental dollars, your holdings are worth 50 experimental dollars (5 shares @ $10 each).
Purchases of shares reduce your cash balance by the price you paid. Sales of shares increase your cash balance by the amount that you receive. Thus you can either make or lose money by purchasing and reselling shares. At the end of each market period, your final positions will be liquidated (converted to experimental dollars). At the beginning of the first period and each subsequent period your initial allocation will be restored to $10,000, and three shares.
Attached are Information & Record sheets for each market period as well as a Profit Sheet. You should keep track of sales and purchases on these sheets. We will now work a simple example to illustrate how to keep track of your activity.
All transactions are for one share at a time. You may not sell more shares than you own (i.e., your portfolio cannot contain a negative number of shares). After each sale/purchase you must record the transaction price on your information sheet, and update the count of your holdings. At the end of the experiment, you will compute the total value of your portfolio for each market period and transfer this amount to your Profit Sheet.
Each market period will last seven minutes. During this time, anyone who wishes to sell a share may raise his/her hand, and when called on, specify an asking price. This will be recorded on the board. For example, if Group 1 wishes to sell a share for $5, a member should (after being called on) announce, "Group 1 ASKS $5." In contrast, if Group 4 wishes to purchase a share for $3, a member should (after being called on) announce, "Group 4 BIDS $3." Groups may continue to lower ASKS or raise BIDS until an agreement has been reached or until time expires.
Finally, if say group 5 wishes to accept the outstanding BID ($3 according to our example), a member should (after being called on) announce, "Group 5 BUYS." In contrast if, say group 7 wishes to accept the outstanding ASK ($5 according to our example), a member should (after being called on) announce, "Group 7 SELLS." In either case, a binding contract has been reached. The share should be transferred to the new owner and each group should record the appropriate transaction price.
After each trade, we will start the bidding process over again. You will be notified of the time remaining in each market period at various intervals.
Information about Share Price
In each market, you will be provided with information that you can use to help assess share prices. Share price is determined by a draw from a standard deck of playing cards. However, the result of the draw will not be revealed to the market until after trading is complete. Instead, a public report will be released to the market prior to trading. The relation between the private draw and public report will depend on which market is in place, Market A or Market B. The relation between private draws and public reports for each market is detailed below:
Market A Information Structure Private Draw Public Report Spade (black) Black Club (black) Black Heart (red) Black Diamond (red) Black Market B Information Structure Private Draw Public Report Spade (black) Black Club (black) Black Heart (red) Red Diamond (red) Red
Note that since there are an equal number of spades, clubs, hearts, and diamonds (13 each) in a standard deck, there is a 50/50 chance of drawing a red card or a black card. Accordingly, in the absence of any additional information, there is a 50/50 chance that shares in any given period will be worth the "red" amount or "black" amount as indicated on your Information & Record Sheet. The information provided by the public report may help (in some cases) clear up some of this uncertainty.
Finally, we will conduct two separate sessions of each market (A and B). The private draw has occurred prior to class. The card drawn has been sealed in an envelope, with the corresponding public report marked on the outside. Envelopes will be opened at the end of the experiment, after all four sessions have been completed. Thus, you will not observe the results of the private draw at any point in time prior to the end of the experiment. Finally, to make subsequent analysis more simple, the private draws are such that both a red and black card have been drawn for each Market (A and B). In other words, one envelope for Market A contains a red card, and the other contains a black card. Again however, until the experiment is over, the only information you have on the contents of the envelopes is the public report.
Now is the time for questions on any aspect of the markets that is unclear. Once trading
CHECKLIST FOR INSTRUCTORS
Prior to Class
1. Prepare packets for each group. Include the following in each packet:
a. One set of instructions
b. Four "Information & Record Sheets" (one for each period) stapled together
* fill-in information on share values using Table 1
* use same share values for each period
c. One "Profit Sheet"
2. Note: There should be an equal number of identical packets for each type of investor in Table 1. For instance, if using six groups, there should be two identical packets for each investor type in Table 1. If using nine groups, there should be three identical packets for each investor type, etc.
3. Make four copies of the "Trade Log."
4. Prepare one pair of envelopes for Market A by placing a playing card (one red and one black) in each envelope. Seal, and write "BLACK" on each envelope. Mark the envelopes to indicate they are for Market A. Do the same for Market B, but write "BLACK" on the envelope containing the black card, and "RED" on the envelope containing the red card.
5. Obtain a device capable of keeping track of time in seven minute intervals.
1. Draw several "T" accounts on the board, with "bid" and "ask" in the debit/credit columns, respectively.
2. Break students into groups and distribute one packet per group.
3. Go over instructions (paraphrasing is sufficient) and do an example of how to record trades on an "Information & Record Sheet." Use values that are different than those in the experiment (e.g., securities are worth $1,000, and you buy for $900, or sell for $1,100).
4. Ask for questions.
5. Indicate which market (A or B) is about to open, select an envelope, and show students the public report, reminding them of the correspondence between public reports and private draws for that particular market.
6. Commence trading, keeping track of activity on the board.
7. Once the market is closed, record each completed contract on the "Trade Log."
8. DO NOT open the envelope! Either wait until both sessions in a given market are finished, or wait until the end of the experiment.
9. Repeat steps 5-8 for the remaining three market sessions.
10. Open envelopes. Ask students to record their profits and names.
11. Collect packets at the end of the experiment.
TABLE 1 Intrinsic Share Values (a,b,c) Private Draw Outcome pr. = 0.5 pr. = 0.5 Investor Type Black Red Expected Value Type I 300 0 150 Type II 250 150 200 Type III 180 170 175 (a) This table depicts the value to each investor type of holding a share of the security under each of the two possible outcomes of the private draw. Each outcome is equally likely. Security values are generally higher (lower) when the private draw is black (red). For example, a black draw is consistent with a firm successfully adopting a business strategy, whereas a red draw is consistent with a poorly performing business model. (b) The term "intrinsic value" is commonly used in valuation settings. It is equivalent to the notion of the underlying economic value of a firm's shares. In contrast, accounting conventions result in a balance sheet that reports a firm's "book value." Restrictive accounting policies regarding the recognition of assets (e.g., certain intangible assets such as brand loyalty, a knowledgeable work force, superior business strategy, etc.) and revenues result in book values that are typically different from underlying economic (intrinsic) values. (c) Shares of a security can have different values to different investors due to tax profiles, investment horizons, risk preferences, etc. (Plott and Sunder 1982, 665; Davis and Holt 1993, 408). For example. Type III investors might have a long investment horizon, and might consider information about short-term results to be relatively unimportant for security values. In contrast, Type I investors might have a short-term horizon, in which current profitability is important. In addition, these investor types might be in different tax brackets and may have differential abilities to utilize tax loss carry-forwards, etc.
This paper has benefited from discussions with Elizabeth Oliver and Geoff Sprinkle, as well as comments from Thomas Hall, two anonymous reviewers, and participants at the 2002 American Accounting Association Southeast Regional Meeting.
(1) Classroom experiments can be a useful tool for developing students' critical thinking and analytical reasoning skills in general (Bonner 1999, 24-28). Moreover, there is evidence that using active learning exercises like experiments has a positive impact on student academic performance (Gremmen and Potters 1997; Berg et al. 1995).
(2) From an experimental design standpoint, making the public financial report strictly uninformative in Market A and perfectly informative in Market B is desirable because it is the strongest manipulation possible, and thus sharpens any contrasts between the two experimental conditions (Friedman and Sunder 1994, 31). Benefits to this type of manipulation are that it maximizes the chances of obtaining observable results and that the results should obtain in the fewest number of replications (market periods).
(3) In the experiment, the audited financial information in Market B is of superior quality to the unaudited information in Market A, where the term "quality" is defined as the degree of correspondence between the information contained in the financial report and the underlying security value. In reality, this result depends in part on the quality of the audit itself. Accordingly, several questions designed to elicit discussion on audit quality and its related economic consequences are included in the text.
(4) The experiment can accommodate larger classes by increasing the number of groups. For example, for classes of approximately 50 students, one could have 12 groups of investors (four of each type), each endowed with two shares of the security, and each consisting of about four students. In general, groups can include anywhere from one to four members and still function well.
(5) Discussion can follow in a single class or be spread over multiple class periods (e.g., ten-minute blocks), and tied to the subject matter scheduled for those periods. Approximately 30 minutes for discussion is sufficient to address numerous questions of interest.
(6) Groups, rather than individuals, were used to manage the auction process. In the oral-double auction described in Appendix A and used in the experiment, participants make offers to trade securities by raising hands. The auctioneer calls on bidders and solicits their bids. Six to 12 bidders (groups) makes it reasonable to recognize groups roughly in the order that they raise their hands. As the number of bidders increases, it becomes more difficult to recognize bidders in an orderly manner.
(7) One could reduce or eliminate the rewards if so desired. However, by doing so, one takes the risk of not sufficiently motivating students. This violates a fundamental tenet of experimental economics and can potentially impact the results from the markets, and the ability to draw inferences about the causes of the results (Smith 1976). However, in practice, the lack of rewards (either monetary or nonmonetary) will not necessarily doom the outcome of the experiments. For example, Frischmann (1996) finds that tax students, participating in a much more complex market, performed in a manner consistent with economic theory despite rewards consisting solely of snack-food items. Accordingly, whether to make external rewards a part of this exercise is a judgment call for the instructor.
(8) In other replications of the experiment, the order of the pairs was reversed, with Market B sessions preceding Market A sessions. See footnote 11 for additional discussion.
(9) The cost of capital is lower because the firm's shares are more valuable in the marketplace. Given that firms can use their shares as currency in business combinations, higher valuations imply that a firm can enter into a specific combination in exchange for fewer shares. Similarly, if a firm wishes to raise capital through the equity markets, then it follows that higher valuations will result in more capital being raised in exchange for a fixed number of shares.
(10) I have run this experiment five times in class. In each experiment the results are comparable to what is reported in the text. Moreover, the results are consistent with those of Plott and Sunder (1982) upon which this experiment is based. Finally, the results are consistent with what is predicted by theory. Collectively, these observations suggest that results from future experimental sessions are highly likely to be consistent with past results.
(11) To verify that the pattern of data observed in Figure 1 is attributable to the differences in the markets, rather than the order in which the markets were run, additional sessions have been run in the reverse order (B B A A), and with various combinations of private draw realizations. The results reported in the paper are consistent with results from other sessions, and hence, appear to be driven by the differences between Market A and Market B, rather than by the order of the sessions.
(12) Students can be reminded at this point that at the time of the experiment, despite the fact that they were told nothing more than what the security was worth to them and that a cash prize was at stake, they made decisions that were consistent with the optimal behavior predicted in the discussion.
(13) Alternatively, one could take a simple average of the closing prices in each market session. For the sessions described in the text, the average closing price in Market A was $191, and the average closing price in Market B was $212.5, or an 11.3 percent increase. Closing prices are often used in analysis because they are thought to be a better indicator of the underlying competitive equilibrium price than averages, which tend to include noise from early trades (Davis and Holt 1993, 168).
(14) The term "abusive" is used to modify the phrase "earnings management" to recognize that some degree of earnings management is permitted by generally accepted accounting principles, and may in fact be beneficial to shareholders. Abusive policies, in contrast, are those that violate the letter or spirit of GAAP.
Antle, R. 1982. The auditor as an economic agent. Journal of Accounting Research 20 (Autumn): 503-527.
Baiman, S., J. H. Evans, and J. Noel. 1987. Optimal contracts with a utility-maximizing auditor. Journal of Accounting Research 25 (Autumn): 217-244.
Berg, J., J. Dickhaut, J. Hughes, K. McCabe, and J. Rayburn. 1995. Capital market experience for financial accounting students. Contemporary Accounting Research 11 (Spring): 941-958.
Blazenko, G., and W. Scott. 1986. A model of standard setting in auditing. Contemporary Accounting Research 3 (Fall): 68-92.
Bonner, S. E. 1999. Choosing teaching methods based on learning objectives: An integrative framework. Issues In Accounting Education 14 (1): 11-39.
Boylan, S. 2000. Using experimental asset markets to illustrate the value of auditing. Issues in Accounting Education 15 (1): 43-65.
Davis, D., and C. Holt. 1993. Experimental Economics. Princeton, NJ: Princeton University Press.
De Young, R. 1993. Market experiments: The laboratory versus the classroom. Journal of Economic Education 24 (Fall): 335-351.
Dopuch, N., R. King, and D. Wallin. 1989. The use of experimental markets in auditing research: Some initial findings. Auditing: A Journal of Practice & Theory 8 (Supplement): 98-127.
______, and ______. 1992. Negligence versus strict liability regimes in auditing: An experimental investigation. The Accounting Review 67 (1): 97-120.
Friedman, D., and S. Sunder. 1994. Experimental Methods: A Primer for Economists. Cambridge, U.K.: Cambridge University Press.
Frischmann, P. 1996. Real time classroom tax planning using experimental markets. Issues in Accounting Education 11 (2): 281-296.
Gremmen, H., and J. Potters. 1997. Assessing the efficacy of gaming in economic education. Journal of Economic Education 28 (Fall): 291-303.
Jensen, M., and W. Meckling. 1976. The theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics 3 (October): 305-360.
Kachelmeier, S. J. 1991. A laboratory market investigation of the demand for strategic auditing: Auditing: A Journal of Practice & Theory 10 (Supplement): 25-48.
Lynch, M., R. Miller, C. Plott, and R. Porter. 1986. Product quality, consumer information and "lemons" in experimental markets. In Empirical Approaches to Consumer Protection Economics, edited by P.M. Ippolito, and D. T. Scheffman, 251-306. Washington, D.C.: Federal Trade Commission, Bureau of Economics.
Messier, W. F., Jr. 2003. Auditing and Assurance Services: A Systematic Approach. 3rd edition. Boston, MA: Irwin/McGraw-Hill.
Plott, C., and S. Sunder. 1982. Efficiency of experimental security markets with insider information: An application of rational expectations models. The Journal of Political Economy 90 (4): 663-698.
Rittenberg, L., and B. Schwieger. 2003. Auditing: Concepts for a Changing Environment. 4th edition. Fort Worth, TX: Dryden.
Robertson, J., and T. Louwers. 2002. Auditing. 10th edition. Boston, MA: Irwin/McGraw-Hill.
Simunic, D. 1980. The pricing of audit services: Theory and evidence. Journal of Accounting Research 18 (Spring): 161-190.
______, and M. Stein. 1987. Product Differentiation in Auditing: Auditor Choice in The Market for Unseasoned New Issues. Vancouver, British Columbia, Canada: Canadian Certified General Accountants' Research Foundation.
Smith, V. L. 1976. Experimental economics: Induced value theory. American Economic Review Papers and Proceedings 66: 274-279.
Wallin, D. 1992. Legal recourse and the demand for auditing. The Accounting Review 67 (1): 121-147.
Scott J. Boylan is an Associate Professor at Washington & Lee University.
|Printer friendly Cite/link Email Feedback|
|Author:||Boylan, Scott J.|
|Publication:||Issues in Accounting Education|
|Date:||May 1, 2004|
|Previous Article:||Characteristics of master's in accounting degree programs.|
|Next Article:||Accounting majors' perceptions regarding the 150-hour rule.|