The effect of audit quality on financial performance

(1)

Thesis

The effect of Audit Quality on Financial Performance

Name: Salihah Ahmed, 10750258

Program: BSc Economics and Business Specialization: Economics and Finance Supervisor: Mark Dijkstra

(2)

Statement of Originality

This document is written by student Salihah Ahmed who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

Abstract

This paper examines whether audit quality has an effect on the financial performance of firms. Audit quality, which stems from earnings quality, is measured by the estimated discretionary accruals using the Modified Jones model (1991). Furthermore, firm performance is measured by the stock returns. The sample consist of 19,513 firm-year observations representing 4,835 publicly listed firms. The data that is used is solely from US firms during the years 2005-2015. It is expected that audit quality will have a positive relationship with the financial performance. Given the negative association between discretionary accruals and audit quality, it is hypothesized that the discretionary accruals will have a negative relationship on financial performance. However, this hypothesis is not supported by the results of this study. The findings suggest that there is a positive association between discretionary accruals and the stock returns. This implies that the stock market attaches a positive value to the discretionary accruals. Furthermore, consistent with prior literature it is expected and also shown that clients of Big 4 auditors report lower discretionary accruals than clients of non-Big 4 auditors.

(3)

1. Introduction

In May 2017, The Financial Times reported that the world’s Big 4 audit firms are still failing in their profession.1 As auditors should deter errors and fraud on a company’s financial statements it is questioned by Sikka (2009) why these audit firms are still involved in scandals. In addition, The Financial Times criticized why regulators haven’t done enough to improve the audit quality after the recent financial crisis and the Enron fraud (FT, 2017 and Sikka, 2009). This indicates that audit quality hasn’t improved yet and therefore in this study it will be examined how firms can improve their audit quality.

The main research question of this study is the following: What is the effect of audit quality on financial performance? The stock returns will be used as a benchmark for financial performance and audit quality will be measured with discretionary accruals. This study will be practically relevant because not only firms but also the external users of audit reports, such as stakeholders, would like to see increasing stock returns. If the results of this study show that high audit quality will increase the returns, firms can be incentivized by simply improving their audit quality and thereby the financial statements to see increasing returns. Audit quality is difficult to define because there is no global accepted recognition of this term (The International Auditing and Assurance Standards Board, 2014). In addition, the quality of an audit is not publicly available information and therefore cannot be examined by users of financial statements (Khurana and Raman, 2004). Therefore, researchers are trying to define audit quality and investigate the determinants and effects of it. According to DeAngelo (1981) audit quality could be described as the joint probability of detecting and reporting material financial statement errors. However, one can criticize that the probability of finding mistakes increases with the number of mistakes made. According to Defond and Zhang (2014) audit quality could better be described as the assurance given by auditors that the financial statements correctly reflect the firms underlying economics. Furthermore, the research of Defond & Zhang (2014) states that the description of DeAngelo (1981) has a binary outcome, where auditors either succeed or fail in detecting errors. It is argued that auditors have more responsibilities than simply detecting errors. An example of such a responsibility is that an auditor should provide an audit opinion which gives the assurance that a firm’s financial report faithfully represents its performance. Therefore, according to

1

(5)

5 Defond and Zhang (2014) audit quality could better be described as the assurance given by auditors that the financial statements correctly reflect the firms underlying economics.

The research of Krishnan (2003) finds a link between the firm’s underlying economics and discretionary accruals. It is found that audit quality conditions the pricing of discretionary accruals. The discretionary accruals, which reflect private information of managers, can improve the ability of earnings in a way that the representation of the firm’s economic performance is improved. These discretionary accruals can arise because of the opportunistic reporting of managers and therefore reflect the private information of managers. Furthermore, Defond and Zhang (2014) also states that discretionary accruals are a common used proxy to measure audit quality. This study will differentiate itself from the study of Krishnan (2003) as this study will be based on a recent sample and as a consequence the effect of the recent financial crisis on audit quality will be examined.

For this study panel data of US public listed firms is used for the years 2005-2015.This study conducts two regressions. The first regression estimates the discretionary accruals using the modified Jones Model (1991). The estimated discretionary accruals are used as a proxy to measure audit quality. This is needed to run the second regression and answer the main question of this research. The second regression measures the stock returns using audit quality, auditor size and the recent financial crisis as independent variables. The sample of this study contains 19,513 firm/per year observations. The results of this study show that discretionary are positively associated with stock returns. Furthermore, the findings of this study indicate that firms audited by Big 4 auditors report on average lower discretionary accruals than firms audited by non-Big 4 auditors.

This study commences with a literature review which gives an overview of the existing literature about audit quality. Section three will describe the research methodology and it will therefore specify the methods which will be used. In section four, a specification of the data will be given. The results are presented and discussed in section 5 and subsequently in section six, a conclusion of this study will be given.

(6)

6

2. Literature review

2.1 Audit quality

According to Francis (2004) there is no single definition of audit quality because of its complexity. It is namely difficult to measure audit quality in an objective way as auditing is a service and the amount of service provided is unobservable. She therefore states that audit quality can be defined as a theoretical continuum in which audit quality, which is measured by the audit failures, is ranged from low to high audit quality. Audit failures happens in the lower end of this continuum and are therefore considered as low audit quality. According to Francis (2004) an audit failure occurs when auditors are unable to enforce the General Accepted Accounting Principles (GAAP) or are unable to deliver a qualified audit report. A qualified audit report should not contain any manipulations or errors. The GAAP are a set of official rules, standards and procedures that publicly traded companies are required to use in preparing their financial statements.The Financial Accounting Standards Board (FASB) are the source of the current United States (US) GAAP. GAAP principles are supposed to give investors confidence about the accuracy of a company’s financial statement and enhance their ability to accurately compare financial statements of companies, especially when they are in the same industry. However, despite GAAPs’ intentions, companies have discretion to use varying methods to value their assets and earning. Therefore, the role of an auditor is to enforce the GAAP and control whether it is followed by firms.

Defond and Zhang (2014) compared the different proxies that are used by researchers to measure audit quality. One of the proxies to measure it is by computing material misstatements; therefore, one could compute the amount of restatements that are made in previously disclosed financial statements. If the amount of restatements is high, it could be seen as relatively poor audit quality. Another proxy to measure audit quality is to look at the market reaction of audit-related events. This proxy should measure the perceptions of users of financial statements, such as investors, lenders, government regulatory agencies, tax authorities and audit committees, about the quality of the financial statements. Furthermore, audit fees could also indicate the level audit quality. High audit fees could indicate higher audit quality. However, Defond and Zhang (2014) questioned this because a higher audit fee could also be explained by other factors, such as monopoly pricing or a risk premium. This risk premium can be asked for litigation risk. Litigation risk refers towards the risk that a legal action will be taken against the company and the accountant may be prosecuted, which

(7)

7 can lead to financial penalties. Litigation risk is higher for companies that are financially distressed, have higher return volatility and riskier accounts (Defond and Zhang, 2014). According to Zaman, Hudaib and Hanifa (2011), financial leverage has a positive relationship with auditing costs. Highly levered firms may need more monitoring by auditors as a protection for financial and market risk. In addition, DeAngelo (1981) states that high transaction costs for firms to switch from an auditor makes it possible for the incumbent auditor to raise the fees, without losing their client. Therefore, it could be argued that it isn’t accurate to state that higher audit fees are an indication of higher audit quality. According to Defond and Zhang (2014) the commonly used proxy to measure audit quality are discretionary accruals. According to them, with this measure manipulation and violation against GAAP principles could be captured and it may signal more undetected misstatements.

2.2 Accruals

Accruals are transactions for which cash hasn’t been received or paid yet, but the effect of these transactions must be recorded in the accounts to match the revenues and expenses (Stolowy and Lebas, 2013). A firm is required to prepare its financial statements under the accrual basis of accounting according to GAAP. Under this method transactions should be recognized and recorded in the financial statements of the period to which they relate. This is the opposite of cash accounting, where transactions are recorded and reported to the period where cash is received or paid. For example: a service took place in June but the firm receives cash in September. According to accrual accounting this transaction should be recorded in the month it took place, even though the firm receives cash in September. So, the revenues will be booked in June and as a consequence ‘accounts receivables’ are created. The revenue recognition principle and the matching principle are two accrual basis accounting principles (Dechow, 1994). According to the revenue recognition principle, revenues should be recognized at the point of sales and expenses should be recognized as they incur, even though cash receipts or payments occur in another period. The matching principle however states that revenues and expenses should be matched. A fair presentation of the results from a firm’s operations during a period requires that all expenses incurred in generating that period’s revenues should be deducted from the revenues. Therefore, cash disbursements which are directly related to revenues should be expensed in the period in which the firm recognizes the revenues. Dechow (1994) documented that with these principles, accruals can mitigate the timing and mismatching problems of cash flows. The

(8)

8 timing and mismatching problems of cashflows arises because the underlying value creation, such as the achievements of the firm often differs from the timing of the related cashflows. Therefore, these principles could enhance the accuracy of financial statements.

The ability of a firm to generate more cash inflows than cash outflows indicates how successful a firm is. Therefore, net realized cashflows could be used as a performance measure. The net realized cash flows are the net cash receipts, so cash inflows minus cash outflows. But as mentioned above, because of the timing problems of cash flows, earnings are considered as a better performance measure (Dechow, 1994). Earnings consist of cash flows from operations which are adjusted by accruals (Dechow, 1994). Therefore, Dechow and Dichev (2002) stated that the role of accruals is to adjust the recognition of cash flows over time, so that firm performance can be measured more accurately better with the adjusted numbers. Furthermore, the FASB stated in the Statement of Financial Accounting Concepts (1978) that earnings of a firm which are measured by accrual accounting will give a better representation of the firm’s performance than net realized cash flows would do.

Despite that accruals enhance the accuracy of financial statements, they are based on assumptions and estimates. If for instance cash for a transaction is paid or received after the corresponding expenses and revenues are recognized in the earnings, managers have to make an estimation of the cash that would be received or paid in the future (Dechow and Dichev, 2002). As a consequence, managers may estimate accruals deliberately wrong in order to present higher and persistent earnings to investors. This process is called earnings management.

2.3 Earnings management

According to Healy and Wahlen (1999) earnings management occurs when managers manipulate transactions or use their own judgment in financial reports in such a way that stakeholders are misinformed about the performance of the company.

Due to information asymmetries between investors and firm’s managers, investors will demand a summary measure of firm performance. With this measure investors will evaluate the managers and investigate the firm’s ability to generate cash and create value. These information asymmetries arise because managers have more information about the firm and its investment opportunities than investors. Information asymmetries may lead to problems when agency costs arise. If for instance a firm is considering two projects to invest in and the

(9)

9 managers have more information about these projects and their net present values. It could be the case that managers will choose the project which will give the managers a higher payoff and which will not necessarily be favorable to investors. The costs that will arise because of these conflicts of interest are the agency costs (Berk and Demarzo, 2013).

According to Dechow (1994) investors, such as stockholders, use earnings as a summary measure of firm performance. As mentioned above, earnings consist of cash flows from operations adjusted by accruals. Managers may manipulate earnings, by estimating the accruals deliberately wrong, in order to present higher and persistent earnings or in order to maximize their own and/or the firm’s wealth. Incentives for managers to maximize their own wealth, by manipulating the earnings, arise when managers are compensated based on the firm’s earnings performance (Xie, Davidson and DaDalt, 2003). Examples of such situations are when managers receive bonusses, prestige and promotions if they achieve a pre-established target of earnings. If for instance a firm fails to achieve this target, managers may modify short-term expenses, such as advertising or training cost in order to achieve this target of earnings and as a result the bonusses and promotions (Stolowy Lebas and Ding, 2013). Beyond this compensation problem for managers, it could be that managers manipulate earnings to maximize the firm’s wealth. A situation in which this is the case is for instance when contracts between the firm and other parties, such as debtholders and suppliers, are based on reported earnings (Becker, Defond, Jiambalvo and Subramanyam, 1998). Lenders would like to know if the firm can pay back the amount of money borrowed and the interest upon it. The earnings of a firm indicate the financial position of a firm and may indicate the ability of the firm to hold to a debt contract. According to Dechow, Sloan and Sweeney (1996) high and stable earnings will encourage debtholders to lend their money to the firm. So, in order to attract external financing managers may manipulate the firm’s earnings. Another reason why managers will manipulate earnings is to avoid hostile takeovers. Easterwoord (2011) states that targets of hostile takeover attempts, will let their earnings increase in order to prevent that shareholders will support the takeover.

The following studies examined how capital market participants are influenced by manipulated earnings. According to Xie et al. (2003), financial information is crucial for capital markets as it is used to set security prices and investors use this information to design an investment strategy. A consequence of incorrect information is that it could lead to mispricing of securities. According to the efficient market hypothesis, securities will be priced fairly, given all the information available to investors (Mohanram, 2014). Therefore,

(10)

10 incorrect information will lead to mispricing of securities and this will reduce the probability that shareholders make correct informed decisions (Xie et al., 2003). Therefore, Xie et al. (2003) state that earnings management could be seen as agency costs because earnings management misrepresents the real performance of the firm.

Information asymmetries and agency costs between managers and investors such as stockholders should be reduced by auditing because you allow auditors, which are considered as outsiders, to verify the validity of financial statements. It is expected from auditors to discourage earnings management as they should enforce GAAPs’ principle of sincerity, which is to accurately represent a company’s financial situation (Defond and Zhang, 2014). Likewise, Watts and Zimmerman (1983) state that auditing is used by firms to reduce agency costs with stockholders. According to them, an auditor who is independent of the manager should reduce managers’ incentives to manipulate earnings. An independent auditor can be seen as a control mechanism for management that has to determine if the financial statements are presented reliably and fairly, without getting influenced by managers. So, an independent auditor should reduce the opportunistic behavior of managers to manipulate the earnings and therefore an independent auditor will give investors assurance that the financial statements are free from errors and manipulations (Defond and Zhang, 2014).

According to Dechow, Ge and Schrand (2010), earnings quality depends on the fairness of the representation of a firm’s financial performance. However, if earnings are manipulated the firm’s performance is not fairly represented and therefore the quality of earnings is low. In addition, earnings management also indicates a lower level of audit quality because it shows that the auditor has not been able to represent a fair view of the firm performance which therefore leads to lower audit quality. Therefore, earnings quality and audit quality are positively associated with each other (Dechow et al.,2010). An increase in the audit quality ill lead to an increase the earnings quality.

2.4 Prior empirical results

Healy (1985) defined accruals as the difference between the reported earnings and cash flows from operations. The total accruals of a firm can be divided into non-discretionary and discretionary accruals. Healy (1985) states that non-discretionary accruals, also called the normal accruals, are a firm’s expected level of accruals when there is no manipulation of earnings.

(11)

11 These normal accruals are accounting adjustments to the firm’s cash flows mandated by the FASB (Healy,1985). According to Healy (1985) for each firm, there is an expected level of accruals given the underlying business of the firm. These expected accruals are estimated by calculating the average of the total accruals of the previous period. These expected accruals, also called the non-discretionary accruals, should equal the total accruals of the current year. However, it could be the case that there is a difference between the total accruals and the expected accruals. This deviation may be due to the opportunistic adjustments to cash flows made by managers in order manipulate the earnings. The accruals that arises because of this opportunistic behavior are called discretionary accruals (Healy,1985). The downside of this approach by Healy (1985) is that it assumes that non-discretionary accruals remain constant and that there is no earnings management in the previous period. However, according to Jones (1991) the non-discretionary accruals of a firm can change over time due to an impact of economic circumstances or a change in the underlying business of the firm.

Jones (1991) estimates the normal level of accruals by taking a linear regression approach. He assumed that a change in total accruals, which is the difference between earnings and cashflows from operating activities, depends on a change in the economic drivers of accruals. These economic drivers are revenue growth and property, plant and equipment (PPE) in his model. The discretionary accruals are the residuals from regressing accruals on their economic drivers. The intuition behind this is that these residuals represent the discretion of managers, in manipulating the earnings, and the estimation errors (Dechow et al, 2010). Richardson, Sloan, Soliman & Tuna (2006) examined if there is an association between earnings management and firms with a high level of accruals. Their sample consists of 76,165 firm-year observations for the period 1962-2001. Earnings management is measured by the enforcement actions that were undertaken by the Security and Exchange Commission (SEC). The SEC is a control mechanism that undertakes action when firms don’t follow the GAAP. In their sample 169 firm-year observations are subjected to SEC enforcement actions for earnings manipulations. They concluded that firms have high abnormal accruals if earnings are manipulated and a reversal of these accruals in the periods after this opportunistic behaviour.

Beckers et al. (1998) have documented that high-quality auditors, which are independent of the manager, are more likely to detect and dissuade questionable accounting practices in order to qualify the audit report. They conclude that high quality auditors could deter

(12)

12 earnings management because if report errors and irregularities are detected and disclosed, this could damage the reputation of the management and the firm and therefore reduce the value of the firm. If for example a firm has overstated its earnings and this will be disclosed, then the reputation of the firm will be damaged as investors will no longer trust the firm and be able to assess the true value of the firm. A consequence of this is that the share prices will decline in the long run and as a result the value of the firm will be reduced (Dechow et al., 1996).

Krishnan (2003) also states that auditors are considered as high-quality auditors if they are more likely to detect and deter earnings manipulation and report material errors and irregularities. He states that it is the expertise and resources high-quality auditors have, that makes them more likely to constrain opportunistic and manipulated reporting of accruals by managers. Therefore, discretionary accruals, which indicates earnings management, are associated with lower audit quality as it indicates that auditors have not been able to constrain the opportunistic behavior of managers.

2.5 Auditor size

According to DeAngelo (1981), larger audit firms, measured by the number of clients, have less incentives to misbehave and this leads to higher perceived audit quality. For a large audit firm, measured by the amounts of firms to audit, each firm account for a relative small proportion of the revenues. This means that losing a client does not have a significant effect on their total revenue. Therefore, these large audit firms have less incentives to misbehave and get involved in the fraud of managers of the firms which are being audited. Small audit firms are more likely to get involved in frauds of their clients because they are afraid to lose their client. The difference in audit quality between small and large audit firms could also be explained by the research of Dopuch and Simunic (1980). They state that large audit firms have greater reputations to protect and will therefore deliver high audit quality. If they deliver low audit quality or get involved in another event of loss they will lose more reputation than small audit firms. This potential loss gives the auditors of large audit firms greater incentives to be independent.

Using a sample of clients over the period 1989-1992, Becker et al. (1998) shows that high quality auditors, which were the Big 6, report lower discretionary accruals compared with

(13)

13 low quality auditors.2 The mean of the discretionary accruals reported by the clients of non-Big 6 auditors is on average 2.1 percent of total assets higher than the mean of the discretionary accruals reported by the clients of Big 6 firms. The sample of this research consisted of 10,397 observations of clients audited by the Big 6 auditors and 2,179 observations of clients audited by non-big 6 auditors. In addition, Francis et al. (1999) finds that even though clients of Big 6 auditors have higher total accruals, they have lower discretionary accruals.

However, despite the findings above that larger audit firms leads to higher audit quality, according to Louis (2005) Big 4 and non-Big 4 auditors have to follow the same accounting principles and regulatory standards. Therefore, there shouldn’t be any difference in the quality of both types of firms. In addition, Louis (2005) state that non-Big 4 auditors have close and long-time connections with their clients and have more knowledge of local markets. One explanation for this could be that on average the non-Big 4 don’t have a large client base as the Big 4. This can give the auditors of the non-Big 4 more time and opportunities to meet with their clients to enhance their relationship. Therefore, they may be more able to detect irregularities. According to Johnstone and Bedard (2004) risky clients, clients with earnings manipulation risk, will choose a small audit firm as they expect more inquiry from Big 4 auditors. Similarly, they state that the Big 4 will select clients with lower risk. This can be explained by the reason that Big 4 audit firms don’t want to be involved with firms having earnings manipulation as they don’t want to lose their reputation. Lastly, Lawrence, Meza and Zhang (2011) examined the effect of Big 4 and non-Big 4 auditors on audit quality with the following proxies: discretionary accruals, the ex-ante cost of equity capital and analyst forecast accuracy. They concluded that the effect of the Big 4 auditors are not significantly different from the non-Big 4.

2

The Big 6 refer to Arthur Andersen, Coopers & Lybrand, Deloitte & Touche, Ernst & Young KPMG and PriceWaterhouse. However, due to mergers and a fall of one of the largest audit firm namely, Arthur Andersen, the Big 6 are now the Big 4.

(14)

14

3. Research Methodology

3.1 Regression models

The main question of this research is: What is the effect of audit quality on financial performance? Audit quality will be measured by discretionary accruals and stock returns will be used as a measure of financial performance. In order to answer this research question, the following regression model will be used:

RET i,t = α + β1DAC i,t + β2Big4 i,t + β3DAC i,t * Big4 i,t + β5Crisis i,t + β6Crisis i,t * DAC i,t

+ β7LN(size) i,t + β8Leverage i,t + β9OCF i,t + ε i,t

Where the (-control) variables stand for:

RET i,t = Stock holding period return for firm i in year t;

DAC i,t = Discretionary accruals for firm i and year t;

Big4 i,t = 1 if firm i has a Big-4 auditor in year t and otherwise 0;

Crisis i,t= 1 if year is 2007 and 2008 and otherwise 0;

-Ln(Size) i,t = The natural logarithm of total assets;

-Leverage i,t = Total debt t-1/total assets t-1 ;

-OCF i,t= Cash flow from operations/total assets t-1.

In order to run the regression as mentioned above, the discretionary accruals have to be calculated first. In line with Lawrence et al. (2011), I will calculate the discretionary accruals by using the modified-Jones model (1991). With this model, the accruals are estimated for every industry group per year. Industry groups are defined by the two-digit SIC codes (Veenman, 2013). This is done, because it is expected that a regression for every industry group will estimate the discretionary accruals more precisely than a regression with all the firms together. Estimating the discretionary accruals with all the firms and industries together would not take in to account the differences of the underlying business activities/economics of the firms. However, with the two-digit grouping the differences in the business activities are taken into account. The model that is used is the same as in Kothari et al. (2005) where all the variables are scaled by lagged total assets. This model is the following:

(15)

15 TACC i,t = β0 + β1(1/Assets i,t) + β2(∆Sales i,t - ∆REC i,t) + β3PPE i,t + ε i,t

Where the variables stand for: TACC i,t = Total accruals

= (Income before extraordinary items t – net cashflow from operating activities t)/

total assets t-1

ASSETS i,t = Total assets in year t-1;

∆SALES i,t = (Sales in year t – Sales in year t-1)/ Total assets in year t-1 ;

∆REC i,t = (Accounts receivable in year t – Accounts receivable in year t-1)/ Total assets in

year t-1 ;

PPE i,t = (Net property, plant and equipment in year t) / total assets in year t-1;

ε i,t = Estimated Discretionary accrual. 3.2 Independent variables

The first independent variable is audit quality which is measured with the discretionary accruals by using the modified Jones (1991) accruals estimations model. This model is still the most used method to measure audit quality because of the high statistical power of this model in detecting earnings management (Lawrence et al., 2011). However, this model has been criticized by academics because of its weaknesses, such as it is subjected to large measurement errors and potential bias (Lawrence et al., 2011). As stated before, discretionary accruals are estimated error terms. These errors are the difference between total accruals and non-discretionary accruals and therefore represent the unexplained or discretionary part of the total accruals. However, it could be argued that in a given year a firm has high total accruals because of the underlying business activities. So, the level of the estimated errors, which are the discretionary accruals, can be a consequence of the higher total accruals. Therefore, these error terms are correlated with firm performance and this can lead to potential bias.

As already stated, this model is still widely used to calculate the discretionary accruals that are used as proxy for audit quality. Therefore, this study will also use this model to calculate the discretionary accruals.It is expected that high audit quality will have a positive effect on stock returns. Audit quality is measured by discretionary accruals and according to Dechow et al. (2010) audit quality and discretionary accruals are negatively associated with each other. This means that high discretionary accruals indicate lower audit quality because auditors have not succeeded in constraining opportunistic reporting of managers. Therefore, it is predicted the discretionary accruals, will have a negative effect on the returns.

(16)

16 In addition, audit firm size will also be used as an independent variable in the regression. As discussed in the literature review, there are contradicting views regarding the association between audit firm size and audit quality. DeAngelo (1981), Becker et al. (1998) and Francis et al. (1999), have shown that a large audit firm size is positively associated with audit quality. Besides, they have shown that the Big 4 auditors provide higher audit quality than the non-Big 4 as their clients report lower discretionary accruals. However, others such as Lawrence et al. (2011) have shown evidence that there is no significant difference in the audit quality of those auditors. This indicates that auditor size doesn’t influence the audit quality. In line with prior research, Lawrence at al. (2011), Becker et al. (1998), Francis et al (1999), audit firm size will be measured through Big 4 versus non-Big 4. The Big 4 are EY, Deloitte, KPMG and PwC. This will be done by including a dummy variable in which a firm audited by Big 4 auditors will be classified as one, whilst non-Big 4 auditors will be classified as zero. Krishnan (2003) shows that that a small auditor size has a negative effect on the stock returns. Therefore, it is predicted that a large audit firm size, measured by the Big 4, will have a positive effect on the returns.

The third variable is the variable that measures the interaction effect of discretionary accruals and audit firm size. With adding this interaction effect, it is expected that there is a difference in the effect of discretionary accruals in combination with a larger audit firm. Becker et al. (1998) and Francis et al. (1999) state that auditors from large audit firms, report lower discretionary accruals because they are more likely to detect and constrain earnings manipulation. In line with this, it can be argued that the discretionary accruals reported by clients of the Big 4 doesn’t give an indication of the audit quality because it is already expected that Big 4 auditors will constrain earnings manipulation and thereby the discretionary accruals. As stated before discretionary accruals are estimated error terms. Therefore, these discretionary accruals of clients audited by the Big 4 could be a consequence of high total accruals or potential bias. So, it is predicted that the discretionary accruals will have an insignificant effect on the returns when the audit firm is a Big 4. For the whole sample, it is expected that discretionary accruals will have a negative effect of on the returns. Therefore, to hold to the prediction that discretionary accruals will have an insignificant effect on the returns when the audit firm is a Big 4, it is predicted that the interaction term will have a positive effect on the returns.

However, it can also be argued that the interaction term may have a negative effect on the returns. The reason behind this is the following: it is expected from auditors of the Big 4 to

(17)

17 constrain earnings manipulation and thereby discretionary accruals (Dechow et al. 2010). This may imply that clients of the Big 4 should have low discretionary accruals or no discretionary accruals at all. Furthermore, if it is disclosed that clients of the Big 4 have discretionary accruals, it would indicate that the Big 4 have not been able to constrain earnings manipulation. So, because the Big 4 haven’t succeeded in constraining earnings manipulation and the reason that for the whole sample a negative effect of discretionary accruals on returns is expected, it is predicted that discretionary accruals in combination with the Big 4 will have a stronger negative effect on the returns.

The fourth independent variable, crisis, is also a dummy variable, which will be classified as one for the years 2007 and 2008, and 0 otherwise. Sikka (2009), states that in January 2008 Lehman Brothers received an unqualified audit opinion on its annual financial statements of 2007. By the summer of 2008 it was still facing severe financial problems and as consequence it went bankrupted. This indicates that Lehman Brothers already faced problems in 2007. These problems came to arise in 2007 when banks were panicked and stopped lending to each other. In addition, according to Mckibbin and Stoeckel (2009) the financial crisis started in 2007 with the bursting of the housing market bubble. This resulted in a collapse of the mortgage- and fnancial markets which subsequently caused the collapse of Lehman Brothers in 2008. As a consequence, the risk premium in the mortgage- and financial markets increased sharply. For this reason, the period 2007 and 2008 will be considered as the financial crisis. It is expected that the financial crisis will have a negative effect on the returns as firms were facing problems.

The fifth variable is the variable that measures the interaction effect of discretionary accruals and the crisis. With adding this interaction effect, it is expected that there is a difference in the effect on returns of discretionary accruals during the crisis and discretionary accruals outside the crisis. During the crisis, it is expected that earnings will decrease as firms are facing problems. To cover this up managers may estimate the accruals deliberately wrong in order to present higher earnings to investors, which leads to more discretionary accruals (Dechow and Dichev, 2002). For the whole period in the sample, it is expected that discretionary accruals are negatively associated with the returns. Therefore, it is predicted that during the crisis the effect of discretionary accruals on returns is negatively stronger. In line with Lawrence at al. (2011) and Krishnan (2003), the control variable Leverage is included in the model. This is to control for the impact of financial risk on returns. It is

(18)

18 predicted that leverage will have a negative effect on the returns. In addition, the variable Ln(size), which is the natural logarithm of total assets, is included to control for the client size. According to the political cost hypothesis of the positive accounting theory, managers of larger firms are more likely to manipulate earnings in such a way that earnings are understated to decrease political costs (Watts and Zimmerman,1986). This affects earnings quality negatively and this indicates lower audit quality. Therefore, it is predicted that this variable will have a negative effect on the returns. Lastly, the variable OCF, which is cash flows from operations divided by the total assets at the beginning of the year, is included to control for the profitability of the firm. Higher operating cash flows will indicate that the firm is more profitable and this will positively affect the returns. Therefore, it is predicted that OCF will have a positive effect on the returns.

3.3 Dependent variable

The dependent variable, financial performance, will be measured trough the monthly stock returns of publicly listed firms. The monthly stock returns are the total holding period returns, so including the possible dividends paid out. Because, the independent variables, on which I am going to regress the stock returns on, are calculated on an annual base, I must annualize these returns as well. This is done by compounding the returns of each month per year.

(19)

19

4. Data

In the paper, data of publicly listed firms, in the USA will be used. The sample period will be a ten-year period from 2005 – 2015. Compustat is used to get access to the data of financial statements of publicly listed firms in the USA. Furthermore, the stock returns are downloaded from CRSP. Consistent with prior research, financial institutions and utility companies are excluded. In addition, the firms who do not have all the required information that is needed to compute the proxy and regression are eliminated.

This study started with 52,605 observations but after deleting the firms with not enough information and merging the files with the returns and financial statement data, there were 19,513 firm-year observations left. The sample consists of 15,326 Big 4 observations and 4,187 non-Big 4 observations

In the following table, it can be seen that the first independent variable, discretionary accruals, has a mean of 0.0011. These discretionary accruals can be separated by clients of the Big 4 and non- Big 4. The mean of this variable is 0.0002 for the firms audited by the Big 4 whereas the mean of the discretionary accruals of firms audited by the non-Big 4 is 0.0044. In addition, the mean of the absolute value of the discretionary accruals (ADA) is also lower for the clients of the Big 4. This result is consistent with the results of Becker et al. (1998) and Francis (1999) who showed that the discretionary accruals of firms audited by the Big 6 are lower and therefore their audit quality is higher.

Another interesting point is the difference in the statistics of returns between clients audited by the Big 4 and clients audited by the non-big 4. The returns of Big 4 clients are on average higher and less volatile than the returns of non-Big 4 clients.

(20)

20 Table 1: Descriptive Statistics

Variable Mean Standard Deviation Minimum Median Maximum

Pooled Sample DAC 0.0011 0.0755 0.4821 0.0088 0.3980 ADA 0.0508 0.0558 0 0.0343 0.4821 LEV 0.2281 0.2462 0 0.1761 1.4029 OCF 0.1271 0.0840 0.0053 0.1101 0.4969 Ln(Size) 6.7708 1.7964 1.7066 6.7573 11.1389 RET 0.1263 0.5894 -0.9823 0.0493 25.0800 Big 4 Clients DAC 0.0002 0.0719 0.4821 0.0081 0.3699 ADA 0.0480 0.0535 0 0.0326 0.4821 LEV 0.2426 0.2468 0 0.1972 1.4029 OCF 0.1271 0.0812 0.0053 0.1110 0.4969 Ln(Size) 7.2615 1.5628 1.8155 7.2030 11.1389 RET 0.1272 0.5849 -0.9823 0.0567 25.0800 Non-Big 4 Clients DAC 0.0044 0.0874 0.4409 0.0110 0.3980 ADA 0.0611 0.0636 0 0.0422 0.4409 LEV 0.1751 0.2364 0 0.0793 1.4029 OCF 0.1270 0.0938 0.0053 0.1067 0.4969 Ln(Size) 4.9756 1.4112 1.7076 4.9187 10.1321 RET 0.1230 0.6057 -0.9732 0.0175 5.6734

(21)

21

5. Results

In table 3 of appendix A, the variance inflation factors(VIFs) are presented. The highest VIF is 3.49. According to Krishnan (2003); Chatterjee and Price (1997), a VIF higher than 10 indicates serious multicollinearity problems. This implies that multicollinearity is not a serious problem in this study. In table 4, it can be seen that all variables, except Crisis, Leverage and Ln(size), are positively correlated with the stock returns. In addition, several regressions have been done to determine the effect of each independent variable on the dependent variable, stock returns. These results can be seen in table 2.

Discretionary accruals (DA) have a positive coefficient and therefore a positive effect on stock returns. This effect is significant in all the models. However, this result is not the same as predicted. It was predicted that discretionary accruals will have a negative effect on the returns because of the negative association between discretionary accruals and audit quality. This contradicting result can be explained by the studies of Krishnan (2003) and

Subramanyam (1996) who shows that discretionary accruals are positively associated with the stock returns. According to Subramanyam (1996) managers may smooth the reported earnings to counteract the effect of temporally changes in earnings. He states that not all smoothing is opportunistic. If earnings are smoothed to present a stable trend in earnings, it may enhance the value relevance of earnings as it improves the persistence of earnings (Subramanyam, 1996). This is however inconsistent with the notion made in this study, that accrual manipulation lowers earnings quality and thereby reduce the value relevance of earnings.

Furthermore, according to Krishnan (2003) and Subramanyam (1996), discretionary accruals are priced by the stock market. A possible explanation for this is that discretionary accruals may contain private inside information of managers about future profitability. According to Subramanyam (1996), prior literature state that dividends may signal private information about the firm’s future performance as well. Therefore, he examined and showed that there is positive association between discretionary accruals and dividend changes. As a consequence, he concludes that both earnings, which include accruals, and dividends signal information about future profitability. In addition, Louis and Robinson (2005) also state that firms will use discretionary accruals to signal private information about future profitability. These studies and the contradicting result may imply that using discretionary accruals as a proxy to measure audit quality may not be suitable in this setting.

(22)

22 The variable, Big 4, has a positive coefficient in all the models. Hence, it can be stated that auditor size is positively associated with the returns. However, the interesting fact is that this effect is only significant in the model including all variables and without fixed year effects. The variable, Big4xDA, represents the interaction effect of discretionary accruals in

combination with the Big 4 on returns. A binary outcome was predicted for this variable, which can be seen in the methods. From the results, it can be seen that this variable has a negative coefficient. This means that discretionary accruals in combination with a Big 4 auditor has a negative effect on the returns. A possible explanation for this is that investors value a greater weight to discretionary accruals of Big 4 clients relative to discretionary accruals of non-Big 4 clients (Krishan, 2003). As discussed earlier, it expected that Big 4 auditors will be more likely to constrain opportunistic reporting of managers. Therefore, if it is disclosed that clients of Big 4 have discretionary accruals, investors may perceive this as a sign that the auditor could not constrain managers in earnings manipulation and therefore this will affect the returns negatively.

Furthermore, the variable crisis negatively affects the returns negatively. This coefficient is significant in all the models. The negative effect of the crisis on returns is in line with the prediction that the crisis affects firm performance negatively. It was predicted that the crisis will lead to a drawback of the earnings and as a consequence lowers the returns.

The last explanatory variable, represent the interaction term of discretionary accruals during the crisis. This variable positively affects the returns. This positive association is marginally significant in the model including all variables without fixed year effects. This result is not in line with the expectation that discretionary accruals, during the crisis, will have a stronger negative effect on the returns. The results suggest that discretionary accruals, during the crisis, have a stronger positive effect on the returns. This may also be explained by the studies ofKrishnan (2003) and Subramanyam (1996) who explains the positive relationship between discretionary accruals and returns. They state that discretionary accruals are priced by the stock market because they contain valuable information for investors about future

profitability.

Lastly, Leverage, Ln(size) and OCF were included as control variables. The signs of the coefficients are in line with what was predicted. The remarkable result is that leverage is not significant in the regression including all the variables without fixed year effects.

(23)

23 Table 2: Regression output of Stock returns

RETURN RETURN RETURN RETURN RETURN RETURN RETURN RETURN RETURN

DA 0.3694*** 0.3701*** 0.6482*** 0.6748*** 0.5755*** 0.5612*** 0.8522*** 0.8526*** 0.8266*** (0.066) (0.066) (0.1386) (0.1342) (0.144) (0.144) (0.1418) (0.1415) (0.1304) Big4 0.0057 0.007 0.0026 0.0027 0.0066 0.0062 0.0361*** 0.0154 (0.0104) (0.0105) (0.0101) (0.0101) (0.0103) (0.0101) (0.0123) (0.0108) Big4xDA -0.3905** -0.4236*** -0.4253*** -0.4143*** -0.3872*** -0.3530** -0.3641*** (0.1571) (0.1515) (0.1512) (0.1511) (0.1453) (0.1451) (0.1338) Crisis -0.3555*** -0.3559*** -0.3568*** -0.3627*** -0.3663*** -0.5822*** (0.009) (0.009) (0.009) (0.0088) (0.0091) (0.0122) CrisisxDA 0.3648*** 0.3641*** 0.237* 0.2287* 0.2321** (0.1309) (0.1309) (0.1249) (0.1249) (0.1143) LEVERAGE -0.0591*** -0.0399** -0.0166 0.034** (0.0179) (0.0174) (0.0180) (0.0164) OCF 1.1354*** 1.1302*** 1.0222*** (0.089) (0.0885) (0.0164) Ln(Size) -0.0138*** -0.0045* (0.0028) (0.0024) Constant 0.1259*** 0.1214*** 0.1202*** 0.1923*** 0.1922*** 0.2029*** 0.0554*** 0.1214*** 0.065*** (0.0042) (0.0094) (0.0093) (0.0093) (0.0093) (0.0097) (0.0136) (0.01670 (0.0171)

Year Fixed Effects No No No No No No No No Yes

N 19513 19513 19513 19513 19513 19513 19513 19513 19513

R2 0.0022 0.0023 0.0028 0.0595 0.0599 0.0605 0.0853 0.0865 0.2405

Adjusted R2 0.0022 0.0022 0.0026 0.0593 0.0596 0.0602 0.085 0.0861 0.2364

(24)

24

6. Conclusion

The effect of audit quality on financial performance has been examined in this empirical study. Audit quality cannot be measured objectively because of its complexity. Therefore, it has been measured with the estimated discretionary accruals using the modified Jones model (1991). Furthermore, financial performance has been measured with stock returns. The sample to conduct this research contains observations from 4,835 publicly listed USA firms during the years 2005-2015. By using this sample, the effect of the recent financial crisis on stock returns has been examined.

As discussed in the theoretical part of the literature review, managers have a discretion in manipulating the reported earnings to represent the firm performance better. Accruals, which are a component of earnings, are often based on assumptions and estimates. This allows managers to behave opportunistically, when reporting earnings. The accruals that arise because of this opportunistic behaviour are called the discretionary accruals. Beckers et al. (1998) state that high quality auditors should constrain this opportunistic behaviour of managers. Therefore, discretionary accruals are negatively associated with audit quality (Dechow et al., 2010). Prior empirical evidence, which can be seen in the literature review, indicates that clients of Big 4 auditors report lower discretionary accruals than firms audited by non-Big 4 auditors.

Given the results of prior literature, it was predicted that discretionary accruals will have a negative effect on the stock returns. Furthermore, it was expected that clients of Big 4 auditors will report lower discretionary accruals than clients audited by non-Big 4 auditors. The results of this study reveal that Big 4 auditors indeed report lower discretionary accruals compared with clients of non-Big 4 auditors. Contrary to the predictions of this study, the results reveal that discretionary accruals have a positive significant effect on the returns. However, this result is in line with the research of Subramanyam (1996) who shows that discretionary accruals are positively associated with the returns. According to Subramanyam (1996) discretionary accruals are priced by the stock market because they contain valuable information for investors about future profitability.

Given these results, it can be concluded that there is not enough evidence to state that audit quality is positively associated with stock returns. This may imply that using discretionary accruals as a proxy to measure audit quality is not the most suitable proxy is in this setting.

(25)

25 Furthermore, the measurement error and potential bias associated with discretionary accruals is a concern. Therefore, this study gives enough evidence for further future research in which audit quality may be measured with another proxy.

(26)

26

7. References

Becker, C. L., DeFond, M. L., Jiambalvo, J., & Subramanyam, K. R. (1998). The effect of audit quality on earnings management. Contemporary accounting research, 15(1), 1-24.

Berk, J., & DeMarzo, P. (2013). Corporate finance. Edinburgh Gate: Pearson Education Limited.

Chatterjee, S., & Mächler, M. (1997). Robust regression: a weighted least squares approach. Communications in Statistics-Theory and Methods, 26(6), 1381-1394.

DeAngelo, L. (1981). Auditor size and audit quality. Journal Of Accounting And Economics, 3(3), 183-199.

Dechow, P. M. (1994). Accounting earnings and cash flows as measures of firm performance: The role of accounting accruals. Journal of accounting and economics, 18(1), 3-42.

Dechow, P., Ge, W., & Schrand, C. (2010). Understanding earnings quality: A review of the proxies, their determinants and their consequences. Journal of Accounting and

Economics 50,344-401.

Dechow, P.M. and I.D. Dichev, 2002, The Quality of Accruals and Earnings: The Role of Accrual Estimation Errors. The Accounting Review 77, 35-59.

DeFond, M., & Zhang, J. (2014). A Review of Archival Auditing Research. Journal of Accounting and Economics, 58(2–3), 275-326.

Dopuch, N., & Simunic, D. (1980). The nature of competition in the auditing profession: a descriptive and normative view. Regulation and the accounting profession, 34(2), 283-289.

Easterwood, C. M. (2011). Takeovers and incentives for earnings management: an empirical analysis. Journal of Applied Business Research (JABR), 14(1), 29-48.

FASB ,1978, The Statement of financial Accounting Concepts, No.1, par.44

Francis, J. R., E. L. Maydew, and H. C. Sparks. 1999. The Role of Big 6 Auditors in The Credible Reporting of Accruals. Auditing: A Journal of Practice and Theory 18(2), 17-34.

(27)

27 Francis, J. R. (2004). What do we know about audit quality?. The British accounting review, 36(4), 345-368.

Healy, P. M. (1985). The effect of bonus schemes on accounting decisions. Journal of accounting and economics, 7(1-3), 85-107.

Healy, P.M. and J.M. Wahlen, 1999, A Review of the Earnings Management Literature and Its Implications for Standard Setting. Accounting Horizons 13, 365-383.

IAASB (2014). A Framework for Audit Quality: Key Elements that Create an Environment for Audit Quality. Retrieved on 1 June 2016 from: https://www.ifac.org/publications-resources/framework-audit-quality-key-elements-create-environment-audit-quality Jere R. Francis and Michael D. Yu (2009) Big 4 Office Size and Audit Quality. The

Accounting Review, 84(5), 1521-1552.

Johnstone, K. M., & Bedard, J. C. (2004). Audit firm portfolio management decisions. Journal of Accounting Research, 42(4), 659-690.

Jones, J. J. (1991). Earnings management during import relief investigations. Journal of accounting research, 193-228.

Khurana, I. K., & Raman, K. K. (2004). Litigation risk and the financial reporting credibility of Big 4 versus non-Big 4 audits: Evidence from Anglo-American countries. The Accounting Review, 79(2), 473-495.

Kothari, S. P., Leone, A. J., & Wasley, C. E. (2005). Performance matched discretionary accrual measures. Journal of accounting and economics, 39(1), 163-197.

Krishnan, G. V. (2003). Audit quality and the pricing of discretionary accruals. Auditing: A Journal of Practice & Theory, 22(1), 109-126.

Lawrence, A., Minutti-Meza, M., & Zhang, P. (2011). Can Big 4 versus Non-Big 4 Differences in Audit-Quality Proxies Be Attributed to Client Characteristics?.

The Accounting Review, 86(1), 259-286.

Louis, H. (2005). Acquirers’ abnormal returns and the non-Big 4 auditor clientele effect. Journal of Accounting and Economics, 40(1), 75-99.

(28)

28 Louis, H., & Robinson, D. (2005). Do managers credibly use accruals to signal private

information? Evidence from the pricing of discretionary accruals around stock splits. Journal of Accounting and Economics, 39(2), 361-380.

Maijoor, J., & Vanstraelen, A. (2006). Earnings management within Europe: the effects of member state audit environment, audit firm quality and international capital markets. Accounting and Business Research, 36(1), 33–52.

McKibbin, W. J., & Stoeckel, A. (2009). The potential impact of the global financial crisis on world trade.

Mohanram, P. S. (2014). Analysts' cash flow forecasts and the decline of the accruals anomaly. Contemporary Accounting Research, 31(4), 1143-1170.

Richardson, S. A., Sloan, R. G., Soliman, M. T., & Tuna, I. (2006). The implications of accounting distortions and growth for accruals and profitability. The Accounting Review, 81(3), 713-743.

Sikka, P. (2009). Financial crisis and the silence of the auditors. Accounting, Organizations and Society, 34(6), 868-873.

Simunic, D. A., & Stein, M. T. (1996). Impact of litigation risk on audit pricing: A review of the economics and the evidence. Auditing, 15, 119.

Stolowy, H., Lebas, M. & Ding, Y (2013). Financial accounting and reporting: a global perspective. Cengage Learning EMEA.

Subramanyam, K.R., 1996, The Pricing of Discretionary Accruals. Journal of Accounting and Economics 22, 249-281.

Vander Blauwhede, H. (2003). Resultaatsturing en kapitaalmarkten: Een overzicht van de academische literatuur, Maandblad voor Accountancy en Bedrijfseconomie, 4 (5), 196-204.

Veenman, D. (2013). Introductory Guide to Using Stata in Empirical Financial Accounting Research. RSM, Erasmus University Rotterdam.

Watts, R. L., & Zimmerman, J. L. (1983). Agency problems, auditing, and the theory of the firm: Some evidence. The Journal of Law and Economics, 26(3), 613-633.

Watts, R. L., & Zimmerman, J. L. (1986). Positive Accounting Theory. Contemporary Topics in Accounting Series

(29)

29 Xie, B., Davidson, W. N., & DaDalt, P. J. (2003). Earnings management and corporate

governance: the role of the board and the audit committee. Journal of corporate finance, 9(3), 295-316.

Zaman, M., Hudaib, M., & Haniffa, R. (2011). Corporate Governance Quality, Audit Fees and Non-Audit Services Fees. Journal of Business Finance & Accounting, 38(1-2), 165-197

(30)

30

8. Appendix

8.1 Appendix A - Correlation Matrices

Table 3: Variance inflation factors to quantify multicollinearity

Variable VIF 1/VIF

DA 3.91 0.255716 Big4 1.39 0.721832 Big4xDA 3.49 0.286493 Crisis 1.01 0.993414 CrisisxDA 1.38 0.722719 Ln(Size) 1.51 0.6621 Leverage 1.1 0.90988 OCF 1.06 0.947659 Mean VIF 1.86

Table 4: Pearson Correlations

Variables RET DA Big4 Big4xDA Crisis CrisisxDA Leverage Ln(Size) OCF

RET 1 DA 0.0473 1 Big4 0.0029 -0.0227 1 Big4xDA 0.0279 0.8437 0.0014 1 Crisis -0.2378 0.0019 -0.0133 -0.003 1 CrisisxDA 0.0394 0.5248 -0.0157 0.4447 0.0141 1 Leverage -0.0209 -0.0375 0.1125 -0.0154 -0.0244 -0.0217 1 Ln(Size) -0.018 0.0519 0.5226 0.0834 -0.0707 0.014 0.2886 1 OCF 0.1394 -0.216 0.0004 -0.1882 0.026 -0.0747 -0.0421 -0.0444 1

(31)

31

8.2 Appendix B – Descriptive statistics of Modified Jones Model

Table 5: Descriptive statistics

Variable Mean

Standard

Deviation Minimum Median Maximum

DA 0.0011 0.0755 -0.4821 0.0088 0.3980

β0 -0.0526 0.0407 -0.7248 -0.0485 0.1329

β1 -0.3633 4.3542 -68.7045 -0.2231 48.0482

β2 -0.0028 0.1050 -0.6721 -0.0029 0.7869

β3 -0.0320 0.1243 -3.2291 -0.0321 3.4905

These variables are estimated for every industry group per year. Industry groups are defined by the two-digit SIC codes (Veenman, 2013).

The effect of audit quality on financial performance

Thesis