Agency Problems and the Choice of Auditors: Evidence from the MENA Region

2.1 Dividend Payout Ratio

An important consequence of inadequate governance and enforcement mechanisms is that managers can use the excess cash available with them on unprofitable opportunities. Grossman and Hart (1980) and Jensen (1986) argue that high dividend payouts reduce cash available with managers, and thus diminish any opportunity for managers to squander excess cash. This, therefore, reduces some of the agency conflicts that may be present in a firm. In another related study, La Porta et al. (2000) formalize the above arguments in a theory known as the substitute model of dividends. This model maintains that firms disgorge cash as dividends to establish a reputation for acting in the best interests of minority shareholders. One of the reasons why firms care about their reputation is their persistent need to raise capital in future. Continuous need to raise capital induces firms to act in a way that is beneficial for outside investors. Paying high dividend is one mechanism via which investors can ensure that not enough cash is left with managers to expropriate. High dividend payments, therefore, result in lower agency problems. We argue that since high dividend paying firms already have low agency problem, therefore there is less incentive for them to hire one of the big-four auditors. While low dividend paying firms, having high agency problems, have more incentives to hire one of the big-four auditors to signal investors that they have lower agency conflicts.

2.2 Ownership Concentration

Prior literature considers ownership concentration as an important characteristic of emerging market firms. Farooq and Kacemi (2011) document that insiders own, on average, almost 50% of shares in firms listed at MENA stock exchanges. Similar observation is also made by Farooq and Seffar (2012) regarding firms in the MENA region. Concentration of ownership in the hands of few allows managers and controlling shareholders to evade effective disclosure of information (Leuz, Nanda, and Wysocki 2003). Poor information disclosure exacerbates information asymmetries between insiders and outsiders and results in agency problems. Prior literature suggests that high ownership concentration creates an entrenchment problem that allows self-dealings by controlling shareholders to go unchallenged by boards of directors. This entrenchment problem is noticed by investors and they discount the stock prices of firms with high ownership concentration. Therefore, we argue that firms have to introduce certain mechanisms that can limit the ability of controlling shareholder to expropriate minority shareholders. One such mechanism is appointing one of the big-four auditors as an external auditor of a firm.

2.3 Operational Complexity

Operational complexity of a firm also intensifies agency problems. Prior literature maintains that operational complexity allows broader scope to management for control to assure effective operations (Abdel-Khalik 1993; Hay and Davis 2004). This strand of literature also indicates that an increasing amount of complexity gives rise to moral hazard problems between managers/controlling shareholders and minority shareholders. Knechel, Niemi, and Sundgren (2008) argued that operational complexity, eventually, leads to the manipulation of information by managers/controlling shareholders, thereby increasing agency problems. We hypothesize that high agency problems associated with operationally complex firms lead to hiring one of the big-four auditors. The big-four auditors, being more objective, are one of the mechanisms via which firms can reduce agency problems embedded in complexity of their operations.

2.4 Transactional Complexity

Another mechanism that can increase agency problems in a firm is transactional complexity (Abdel-Khalik 1993). Transactional complexity relates to the amount of accruals and inventory. Ge and McVay (2005) observe that the two largest sources of reported material weaknesses in internal control of any firm relate to accruals (receivables) and inventory. Firms with high transactional complexity usually have less informative financial statement and thus have higher agency problems (Knechel, Niemi, and Sundgren 2008). We argue that higher agency problems result in a higher level of information asymmetry, which may allow managers and controlling shareholders to take advantage of minority shareholders. Therefore, transactional complexity in a firm is perceived as a negative signal by investors. Given the information asymmetries embedded in firms with high transactional complexities, we hypothesize that these firms are more likely to hire one of the big-four auditors than firms with less transactional complexity.

3.1 Choice of Auditors

This paper defines the choice of auditors by firm’s decision to appoint one of the big-four auditors as its auditor. Our definition of big-four auditors consists of KPMG, Ernst & Young, PriceWaterhouseCoopers and Deloitte & Touche. The data for auditors was collected from Infinancials. Table 1 documents the descriptive statistics for the choice of auditors. Panel A documents the number of firms audited by one of the big-four auditors and the number of firms audited by one of the non-big-four auditors for each year, while Panel B and Panel C document similar statistics for each industry and each country, respectively. Table 1, Panel A, shows that most of the firms choose to get audited by one of the big-four auditors throughout our sample period. For example, our results show that 299 firms chose one of the big-four auditors in 2008 in comparison to 154 firms who chose one of the non-big-four auditors. The results also indicate that big-four auditors dominate the audit market in the MENA region. Our results in Table 1, Panel B, show that all industrial sectors, except health care, consumer goods and oil and gas, prefer one of the big-four auditors as their external auditors. Our results in Table 1, Panel C, confirm our previous findings that big-four auditors dominate the audit markets in the MENA region. The only exception is Qatar, where non-big-four auditors are the preferred choice of auditors.

3.2 Agency Variables

This paper uses dividend payout ratio, ownership concentration, operational complexity and transactional complexity to capture various aspects of agency problems. In this paper, we define dividend payout ratio as the proportion of earnings paid as dividends. Dividend payouts are supposed to alleviate agency conflicts through the reduction of free cash flow available to managers (LaPorta et al. 2000). Ownership concentration, another variable used as a proxy for agency problems, is defined as the proportion of shares held by insiders. Concentrating ownership in the hands of few exacerbates the information disclosure and increase information asymmetry. Leuz, Nanda, and Wysocki (2003) document that whenever ownership is concentrated in the hands of insiders, quality of disclosed information goes down. This paper uses operational complexity and transactional complexity to capture those aspects of agency problems that are related to transparency. Prior literature defines operational complexity as the ratio of salary expense to total operating expenses, while transactional complexity is defined as the ratio of inventory and receivables to total assets (Knechel, Niemi, and Sundgren 2008). High complexity is supposed to decrease the transparency of firms, thereby making it hard for investors to decipher information reported by firms. The data for the calculation of above-mentioned variables was collected from Worldscope and Thomson Financials. Table 2 documents the descriptive statistics for agency variables used in this study. The results in Table 2, Panel A, show a gradual increase in dividend payout ratios and decrease in ownership concentration for our sample firms during our sample period. It may indicate lowering of agency problems in the MENA region over the period of time. However, when comparing the findings of Table 2 against the developed financial markets, we argue that firms in the MENA region pose significant agency problems. Our results show that an average firm in the MENA region is owned by insiders – an indication of information asymmetry for the MENA region firms. Similarly, dividend payout ratio of less than 35% indicates that insiders have greater cash flow to expropriate in the MENA region. Table 2, Panel A, also shows that operational complexity and transactional complexity remained more or less the same during our sample period. The results in Table 2, Panel B, indicate that, except for firms in telecommunication and health-care sector, all others have relatively low dividend payout ratios. We report almost 47% dividend payout ratio for telecommunication sector and 44% dividend payout ratio for health-care sector. It indicates relatively low agency problems for firms belonging to these sectors. The results in Table 2, Panel B, also show that firms belonging to health-care sector are the most widely held firms in the MENA region. We report 15% ownership by insiders for firms belonging to health-care sector. The results in Table 2, Panel B, also indicate the highest operational complexity for firms belonging to financial sector, while the same table indicates the least operational complexity for firms belonging to utility and consumer goods sector. The results are intuitive because financial sector deals with more complex operations, while utility sector usually has the least operational complexity. Our results also show that unusually low transactional complexity for telecommunication and utility sector. We show in Table 2, Panel C, that North African firms have the highest ownership concentration and the highest transactional complexity. We report average ownership concentration of almost 63% and 70% for Egypt and Morocco, respectively, and average transactional complexity of almost 0.31 and 0.48, respectively, for the same countries. Furthermore, our results also show high payout ratios for Moroccan firms, i.e. almost 43%.

4.1 Payout Ratio and Choice of Auditors

Prior literature considers payout ratio as a mechanism via which firms can reduce agency problems. Grossman and Hart (1980), for example, document that high dividend payouts alleviate agency conflicts through the reduction of free cash flow available to managers. In order to test our hypothesis, we estimate a logistic regression with a dummy variable representing whether a firm is audited by one of the big-four auditors or not (AUDITOR) as a dependent variable and payout ratio (PoR) as an independent variable. We define PoR as the proportion of earnings paid as dividends, while AUDITOR is defined as a dummy variable that takes the value of 1 if firm is audited by one of the big-four auditors and 0 otherwise. Our basic equation takes the following form. The country dummies, industry dummies and year dummies are represented by CDUM, IDUM and YDUM, respectively:

However, there may be concerns that some of the firm-specific characteristics might be driving the decision of a firm to engage a particular type of auditor. For example, prior literature suggests that agency cost increase with the percentage of debt in the capital structure (Kim and Sorensen 1986). Firth and Smith (1992) argue that increased agency costs lead firms to hire reputable auditor. Therefore, we add total debt to total asset ratio (LEV) in our regression equation. In addition, we also control for the size of firm. Citron and Manalis (2001) document that bigger firms usually appoint one of the big-four auditors as their external auditors. Log of firm’s market capitalization (SIZE) was, therefore, added in eq. [1] to control for the effect of size on the choice of auditor. Furthermore, we add growth in earnings (GROWTH) in our regression equation as a proxy for growth opportunities. High growth firms should have excessive need for external financing and thus should have incentives to hire one of the big-four auditors (Das and Sengupta 2001). We also added earnings per share (EPS) as a proxy for firm’s profitability. Abbott and Parker (2000) suggest that more profitable firms can engage reputable auditors because they have means to pay the high fee charged by one of the big-four auditors. Our modified regression equation takes the following form:

The results of the above set of regression are reported in Table 3. Our results show that as dividend payout ratio goes up, the probability of appointing one of the big-four auditors goes down significantly. Our result from eq. [2], the most comprehensive equation, shows that for every one unit increase in dividend payout ratio, we expect a 0.0081 decrease in the log odds of appointing one of the big-four auditors. Our results are consistent with La Porta et al. (2000) who consider high dividend payouts as a mechanism via which firms can establish a reputation for acting in the best interest of minority shareholders. High dividend payouts result in lower agency problems because of the reduced cash flows available with managers. We argue that low agency problems of high dividend paying firms reduce the need to appoint one of the big-four auditors. Consistent with our expectations, we show that EPS and leverage are positively related to firm’s choice of appointing one of the big-four auditors. However, contrary to previous literature, our results show negative relationship between growth opportunities and firm’s choice of appointing one of the big-four auditors. We show that for one unit increase in growth rate, there is a 0.0001 decrease in the log odds of appointing one of the big-four auditors. Our results show no significant relationship between size and firm’s choice of appointing one of the big-four auditors.

4.2 Ownership Structure and Choice of Auditors

Concentrated ownership has been one of the main characteristic of emerging markets. Claessens, Djankov, and Lang (2000) document that more than two-thirds of firms are controlled by a single shareholder in emerging markets. Concentration of ownership in the hands of a few results in exacerbating agency problems in these markets (La Porta, Lopez-de-Silanes, and Shleifer 1999). In order to test whether ownership concentration gives rise to the need of having a reputable auditor, we estimate a logistic regression with AUDITOR as a dependent variable and ownership concentration (OWNERSHIP) as an independent variable. We define OWNERSHIP as the proportion of shares held by insiders. Our regression equations take the following forms. As before, we also control for various firm-specific characteristics:

and

The results of the above set of regression are reported in Table 4. Our results show that as the ownership concentration goes up, the probability of appointing one of the big-four auditors also goes up significantly. Our result from eq. [4] shows that for every one unit increase in ownership concentration, there is a 0.0576 increase in the log odds of appointing one of the big-four auditors. Our results may be driven by the fact that high ownership concentration allows controlling shareholder to take decisions without taking into account wishes of minority shareholder. Therefore, high ownership concentration exacerbates agency problems in a firm and thus induces firms to hire one of the big-four auditors to reduce some of agency problems. Our results are consistent with Farooq and Kacemi (2011) who document positive relationship between ownership concentration and firm’s choice of appointing one of the big-four auditors in the MENA region. They also report that market considers appointing one of the big-four auditors as a value relevant signal and reward firms with one of the big-four auditors with better stock price performance.

4.3 Operational Complexity and Choice of Auditors

The more operationally complex a firm is, the broader is management’s scope of control to assure effective operations (Hay and Davis 2004; Abdel-Khalik 1993). Higher control by management increase agency costs, decrease transparency and enhance moral hazard problems. Therefore, firms with higher operational complexity have greater desire to appoint of the big-four auditors to lower these problems. In order to test our hypothesis, we estimate a logistic regression with AUDITOR as a dependent variable and operational complexity (OPERCOMP) as an independent variable. We define OPERCOMP as the ratio of salary expense to total operating expenses (Knechel, Niemi, and Sundgren 2008). Our regression equations take the following forms:

and

The results of the above set of regression are reported in Table 5. Our results show that as the operational complexity goes up, the probability of appointing one of the big-four auditors also goes up. Our result from eq. [6] shows that for every one unit increase in operational complexity, there is a 0.9518 increase in the log odds of appointing one of the big-four auditors. Our results are consistent with our earlier arguments as operational complexity goes up, it becomes hard for outside investors to understand and interpret information. Appointment of one of the big-four auditors provide reliability to information and help investors resolve some of the information asymmetry associated with operationally complexity firms.

4.4 Transactional Complexity and Choice of Auditors

As a last step, we test whether transactional complexity of a firm affects the choice of auditors or not. Prior literature argues that firms with more transactional complexity are less transparent and thus will have more agency problems. This strand of literature argues that complexity hinders investor’s ability to understand firm’s information and thus introduces adverse selection problems for investors (Knechel, Niemi, and Sundgren 2008). It, eventually, leads to having a need of reputable auditor. In order to test our hypothesis, we estimate a logistic regression with AUDITOR as a dependent variable and transactional complexity (TRANSCOMP) as an independent variable. We define TRANSCOMP as the ratio of inventory and receivables to total assets (Knechel, Niemi, and Sundgren 2008). Our regression equations take the following forms:

and

The results of the above set of regression are reported in Table 6. Consistent with previous results, we show that as the transactional complexity goes up, the probability of appointing one of the big-four auditors also goes up. Our result from eq. [8] shows that for every one unit increase in ownership concentration, there is a 3.4150 increase in the log odds of appointing one of the big-four auditors. We argue that high transactional complexity increases agency problems and it induces firms to appoint one of the big-four auditors. By appointing reputable auditors, a firm can signal outside investors that there are less agency conflicts embedded in its structure.