This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.
This chapter summarizes the findings of a great volume of prior published researches that refer to earnings management, the incentives and other factors that promote it. Also, the role of IFRS together with the advantages and drawbacks that are involved in them are analysed. Furthermore, an analysis of the consequences from the adoption of IFRS on earnings management, either voluntary or mandatory from other code-law countries is involved. Additionally, this section deals with references to the BIG 4 audit firms, their attributes and their differentiation compared to the rest audit firms. Finally, various earnings management estimation models are presented.
3.1 DETERMINANTS & INFLUENTIAL FACTORS OF EARNINGS MANAGEMENT
There have been plenty of assertions about the provenance of earnings management and the parameters that prompt the utilization of the corresponding means. The study of Healy & Wahlen (1998) is a landmark with respect to these assertions and indicates that there are three basic categories of incentives. The first is relative to capital market issues, such as the managers' will to meet the expectations of financial analysts. The second refers to contractual relations in both lending and management's reward terms. In the case of the former, earnings management patterns were used to eliminate the possibility of breaking a contract and subsequently suffering the respective negative implications. In the case of the latter, it was used to boost the accounting figures that, under specific circumstances, would ensure the growth of managers' personal wealth, just like bonuses.
Zeghal and Othman (2006) draw upon previous studies and they briefly refer to motives such as the remuneration of managers, the adherence to debt covenants and regulations, the aversion for reporting losses and the attainment of the analysts' forecasts. Apart from that, they also define a triptych of motives categorized as follows: i) Firm attributes such as the size, the industry in which it operates and the degree of executives' ownership ii) Contextual factors (avoidance of losses) and iii) Conjectural factors (income smoothing). They suggest that the implementation of IFRS is not identical across all countries, while the domestic authorities of each country are mainly responsible for this purpose. As a consequence, the level of earnings management also varies. To account for the inherent social and economic differences which impact the accounting practices they add value by developing the "Earnings Management Motives (EMM)" model. Specifically they focus on the methods of financing, either equity or debt based the scheme of corporate governance and the legal environment under which the firms operate. They recommend that the companies which are subject to legislation that ties tax to financial reporting and are financed through equity are more prone to manage earnings.
SUN Wensheng & SUN Jie (2008) examine the issue from a different perspective. Except for the known economic factors that influence earnings management they also mention ethical elements as determinant factors. They believe that the ethical standards (or their absence) that dominate a firm and the degree to which this is made explicit to its employees can accordingly constrain (or allow) earnings management practices. Of course, there are conceptual differences of what is or is not allowed and considered as "earnings management" varies depending on one's personal characteristics. Relative to this point of view is the Kaplan's (2001) theory who states that individuals characterize as less unethical a behavior which is to their own benefit. That is, managers are more lenient in judging actions that can be deemed earnings management, compared to shareholders.
Through their own research, Hutchinson & Leung (2007), support the inference that the level of earnings management varies disproportionately with that of the managers' ownership. This means that the augmentation of ownership from minor levels to higher ones restricts earnings management, while, in higher levels of ownership it is more profound. They also examine some features that make earnings management more intense. Concretely, they prove it is more pervasive in little firms, in firms whose operations show unsteady trends - in order to satisfy earnings smoothing purposes - and in low growth firms as well. The same holds for firms which operate in specific industries such as those that engage in cutting edge technologies, due to the need for protection from the prevailing uncertainty of investments in unpredictable resources like intangible assets. Finally, earnings management is more pronounced in firms with insignificant participation of institutional owners -because of the absence of monitoring that would otherwise exist- and the proficiency of auditors depicted by various features like whether the audit firm is a BIG 4 or not, the auditor's independence and the duration of his/her tenure.
Cheng & Warfield (2005) initially define two forms of equity motives for managers stemming from "stock based compensation and stock ownership". They corroborate that managers seek to get rid of the shares they own by selling them, in order to moderate the risk they are subject to and disentangle from the firm's risks. In this regard, they are propelled to use earnings management so as to influence the formulation process of stock prices, which actually means to inflate the prices of the stocks they aim to sell and take advantage of this premium. Their findings substantiate the positive alignment between the executives' stock related incentives and earnings management.
To sum up, we note that there are numerous perspectives through which incentives and factors can affect an individual's behavior and spark earnings management. Now that they are specified, the next section will refer to another critical point, the consequences, both positive and negative involved in IFRS.
3.2 CONSEQUENCES FROM THE ADOPTION OF IFRS
This section is divided in three parts: the first one deals with how the IFRS influence accrual accounting choices as well as the differences between voluntary and mandatory adopters. The second one analyzes the consequences from the adoption of IFRS on earnings quality and the impact they had on countries which are classified as "code-law" accompanied by the relative characteristics mentioned in section 1.2 and further analyzed below. Finally the profits and drawbacks they have incurred are reviewed as well.
3.2.1 Accrual Accounting & Voluntary vs Mandatory Adoption
As mentioned, the adoption of IFRS became obligatory for publicly listed companies within Europe since the fiscal year 2005. One of the changes that this adoption dictated was the emphasis on the use of accrual accounting. As stated: "Accrual basis means a basis of accounting under which transactions and other events are recognized when they occur (and not only when cash or its equivalent is received or paid). Therefore, the transactions and events are recorded in the accounting records and recognized in the financial statements of the periods to which they relate." (www.ifac.org). This practice enables the financial reports to depict a more complete and superior picture of the real economic circumstances that dominate a firm. This is feasible while the financial statements do not only mention the current cash expenses and revenues, but they also mention the expenses to be paid in the future and the revenues to be received as well. Apart from that, a lot of firms decided to operate and report under the IFRS before they were mandated to do so (starting from 1st January 2005). This was a tempting point for research and previous studies probed the reasons for this choice and documented its results with respect to earnings management.
Dumontier & Raffournier (1998) made a preliminary research about what urges the companies to adopt the IFRS voluntarily. Specifically, they examined the case of Switzerland and they concluded that firms adopted the IFRS in order to profit from this option. Due to the peculiarity of Switzerland's accounting regulatory framework, which permits numerous alternative reporting choices, firms would have reasons to report under the IFRS for the following reasons: users of financial statements outside Switzerland were not accustomed to its domestic accounting practices and the acquisition of necessary information is not easy through alternative sources. Also, adoption of IFRS signals more extensive disclosures, as it contains explicit and more severe rules. Last but not least, the foreign markets and the political costs they impose was another reason for this choice.
Leuz, Daske, Hail & Verdi (2008) also conducted another research about the effects of the sharp adoption of accrual accounting and the voluntary adoption of IFRS. They examined the changes in terms of liquidity, cost of capital and valuation of companies. Initially, they partitioned their sample in mandatory and voluntary adopters to account for the effects of this change. Their inferences suggest that the firms which were obliged to comply with the IFRS revealed higher liquidity amelioration compared to those which had in advance adopted the IFRS voluntarily. On the contrary, the former firms experienced an increase in the cost of capital. Moreover, because of the differentiation in the implementation of the IFRS across countries and the room they leave for applying personal discretion, the significance of the capital market effects varies depending on the severity of their infliction. However, one weakness identified in their paper is that these effects might not be exclusively charged to the IFRS application, but in changes in other factors such as the audit practices, the corporate governance structure and the level of compliance with the IFRS. Finally, they prove that the greater the deviation of national accounting principles from the international standards and the beforehand lack of congruence, the greater the importance and size of this transition is.
In the same line of reasoning, Ipino & Parbonetti (2011), underpin these arguments and provide new insights by documenting the replacement of accrual-based with real earnings management techniques in the aftermath of mandatory IFRS adoption. Not only do they demonstrate the replacement of the former with the latter when the legislative framework is severe, they also demonstrate that the real earnings managements worsens the firms performance and prospects at a greater extent. Thus, the standard-setters' effort to improve the quality of earnings through the IFRS are rendered void unwittingly.
Another research conducted by Guenther, Gegenfurtner, Kaserer & Achleitner (2009), investigates the qualitative characteristics and the differences between mandatory and voluntary IFRS adopters and compares the results for these two groups. Particularly, they find that mandatory adopters were more emphatically characterized by smaller size as well as increased external and bank shareholding. In addition, they note a drop in the extent of income smoothing for early adopters. Regarding the results, the IFRS induced the expansion of conservative practices for mandatory and voluntary groups.
3.2.2 IFRS: Effects on Earnings Management in code-law countries and earnings quality
In this section, the attributes of code-law countries and the associated content and inferences of previous studies about the change in the level of earnings management are mentioned with respect to code-law countries other than Italy after the adoption of IFRS and their effect on earnings quality in general. These countries are, particularly, Germany, Austria and Switzerland. It is shown that the findings differ substantially for each country which is examined and cannot reach an absolutely clear conclusion.
In code-law countries the capital market is less dynamic due to the fact that the firms most of the times derive their financial resources from banks and governmental units. On top of that, investors' protection is weak and their power to claim legal discipline is limited. Besides, the litigation activities are rare in these countries. Considering all these elements, the necessity for solvent financial reporting is quite low. In fact, financial reports' are more prepared in order to respond to demands associated with tax issues or dividend purposes. As Leuz, Nanda & Wysocki (2003) state however, the characteristics of code-law countries and the variation of their effect on earnings management, compared to common-law countries, only have a meaning when the legal enforcement is tight.
The study of Van Tendeloo & Ann Vanstraelen (2005) reveals interesting findings about the association between earnings management and the adoption of IFRS. Particularly, as Germany is a code-law country too, they investigate the case of German public companies and how different the degree is to which they used to manipulate earnings with and without IFRS for the period 1999-2001 as proxied by discretionary accruals. At that time, the adoption of IFRS was not obligatory, however, there were many firms that asked and finally were allowed to adopt the IFRS starting from 1998 because of the reprobation that the domestic German GAAP had received. The criticism was caused by the abundance of managers' options with respect to accounting treatments. It is reasonable to think that since the IFRS inevitably lead to the revelation of necessary information for investors and limit the accounting options of managers compared to German GAAP, the extent of earnings management will be reduced. However, another characteristic of code-law countries is the infrequent lawsuit risks to which they are subject. This means that unlike the previous case, this might incorrectly lead to earnings management techniques. The conclusions they reached suggest that the adoption of IFRS does not necessarily ensure higher reporting quality. In fact, they found that those that have adopted the IFRS actually smoothed earnings much more than those who reported under domestic German GAAP and utilize larger amounts of discretionary accruals. However, they make an important remark regarding the IFRS adopters: when these firms are audited by one of the BIG 4 audit firms the effects of earnings management are minimized significantly. As a final inference, although the authors acknowledge the necessity of IFRS, they state that the international standards, on their own, are not enough to secure the quality of financial reports.
Daske & Gebhardt (2006) examined the change in earnings management before and after the adoption of IFRS in three countries, namely, Germany, Switzerland and Austria. All of them are characterized as code-law countries. For the period between 2000 and 2004, on average, about one fourth of the total number of firms which adopted the IFRS worldwide operated in these countries. The researchers, in order to ascertain the change in the quality of financial reports, used rankings of the disclosure quality of these firms, as derived by the "Best Annual Report". Although they doubt the objectivity of these rankings, they argue that they are reliable because they come from the actual financial reports after the elaboration they have undergone by financial analysts. The criteria used for this evaluation refer to objects such as optional and extra information revealed by the firms, the notes to the financial reports and the future prospects of the firms. Their findings suggest that in all three countries the financial statements of firms that report under IFRS are of higher quality compared to those that report under local GAAP. What is interesting is that this inference is valid not only for voluntary adopters but can be extended to firms that were mandated to report under IFRS as of 2005. This comes to partially contravene the findings of the research mentioned before (Van Tendeloo & Ann Vanstraelen, 2005) as the two studies provide contrasting insights about the effects of IFRS on earnings management.
As far as the effects on the quality of reported earnings and accounting numbers as a whole under the IFRS are concerned, no definitive inferences can be extracted while results are conflicting and inconclusive. Ball (2001) suggests that the international standards ameliorate the quality of financial reporting in general. On the other hand, Paananen (2008) highlights the deterioration of the quality of financial reporting because of international standards. After having taken a closer look at the case of Sweden, she discovered strengthening of income smoothing techniques and overdue detection of losses. Of course this is an isolated incident and it would be risky to generalise these findings for all IFRS adopters. A recent scientific study of Houqe, Zijl, Dunstan & Karim (2011) innovates suggesting that 'earnings quality is a joint function of investor protection and the quality of accounting standards, as proxied by IFRS'. This actually implies that none of these two components is adequate enough to ensure the amelioration of the informativeness and quality of financial reports on its own. It is only the combination of these two that allows for such an improvement. The 'investor protection' concept is approximated by variables like the legal and regulatory environment. A research of Sonderstrom & Sun (2007) is unanimous with respect to the role that the peculiarities of each country plays in the effective (or not) application of IFRS. Specifically, the authors conjencture that the IFRS do not equally enhance the quality of financial reports as this is dependent on country-specific characteristics. Furthermore, each firm prepares its official accounting and financial information in accordance with its needs and motives. Through these motives the quality of reports is influenced either negatively or positively in correlation with the discretion allowed by the country's regulations and laws and its political system too. This actually means that the appropriateness and the robustness of accounting standards is not the only crucial aspect that determines accounting quality.
3.2.3 Advantages and disadvantages from the adoption of IFRS
The last paragraphs of this part are devoted to the advantages and disadvantages arising for the implementation of IFRS. Nobes (2004) scrutinizes the international standards and stresses the importance of the existence of the suitable principles that should be embodied in the explicit rules. Of course he admits the advantages of these rules namely the transparency and comparability of financial reports that they can incur, but he also adds that the necessity for them exists due to the absence of such principles. Apart from that, he suggests that this is the reason why there is enough space for discretionary accounting treatments and as a consequence, earnings management. His main contribution is that he makes clear the categories of accounting that are subject to such limitations-distortions and proposes modifications and the right principles for their improvement such as the correct definition of balance sheet items (assets). These categories relate to Leasing, Financial Assets, Post-employment Benefits, Subsidiaries, Equity issues and Government Grants. Finally, he concludes that the inclusion of the right principles could both diminish the bulk of standards and improve the quality of reporting at the same time supporting in general a principles-based approach. Paul & Burks (2010) point out the negative and positive effects of the transition for US GAAP to IFRS, some of which can be generalized for the rest of the countries that have adopted the IFRS too. For example, as benefits they consider the lowering in costs for enterprises, especially for making investments abroad, the fact that more hierarchical management layers will be preoccupied with the financial reports thus reducing the possibility of mistakes and earnings management attempts and the elasticity of rules which, if used appropriately, leads to revelation of more in-depth and useful information. Couto, Cordeiro & Silva (2009) investigated the case of Portugal and found that although the initial intention of the implementation of the IFRS was purely to improve the quality of financial reporting, it triggered some discrepancies in the short run. These refer to individual elements of the Balance Sheet as well as the Income Statement. However, they did not record an absolute trend in these alterations.
Right at the commencement of the IFRS implementation, Ball (2005), founded on a theoretical level his opinion about the consequences of the switch from the application of local GAAP in each individual European country to the application of the common IFRS as of 2005. He categorized the advantages into direct and indirect and noted that the latter could prove more profitable than the former. In the first category he sets the cost reduction as well as the elimination of prior arbitrary assessments made by analysts in order to homogenize international financial reports. In addition, he sets the enhancement of the position of minor investors in comparison with the major and most skillest ones, because they would both derive their information from the same sources and he also puts emphasis on the "Efficient Market Hypothesis (EMH)", stating that stock prices would reflect with a greater preciseness the available information. Regarding the second category, he illustrates the managers would be more cautious when deciding about the realization of investments, if they are aware of the fact that they are closely observed and they would, therefore, act in the interest of shareholders more and more. Another subject that he goes through is the "fair value accounting". Although he admits that this method, as promoted by the IFRS, can be useful and informative, he also opposes to the argument that it can render the prediction of future earnings very difficult, considering the potential estimation errors included in it. This is why he doubts the merit that this practice could breed.
As far as the drawbacks of the implementation of IFRS are concerned, they consist of the irregular and unequal implementation of standards among the countries that have adopted them. He first makes a distinction between the homogeneity of reporting standards and the real reporting activity, which may vary even under the same reporting regulation (IFRS). Then, he explains that although the standards are common internationally, the degree of their diversified implementation is attributed to "political and economic reasons". This actually means that the cross-country differences in monitoring, enforcement bodies and either common or code-law status are deterministic factors that shape the level of inconsistency. As a consequence, investors are tricked, while left to perceive financial statements of companies from different countries as equally reliable.
3.3 EARNINGS MANAGEMENT & THE ROLE OF AUDITORS: BIG 4 & OTHERS
This section grapples with the previous studies that bring to the surface the prominence of the auditors with respect to the influence they can exert in order to restrain earnings management attempts, pinpointing especially the BIG 4's activity.
3.3.1 Audit quality & the usefulness of auditors
First of all, the quality of audit determines to a high level how valuable the financial statements are. With the term "audit quality" is meant the "the joint probability that the external auditor detects an anomaly in financial statements, and then reveals it to the users of these statements" (DeAngelo, 1981). Of course, this is just a brief description and as it is understood, it contains components such as the auditor's fame, the existence or not of an audit committee inside an organization, as well as the degree to which it is free of managerial influence and the duration of the auditor's occupation by the firm. The audit committee has a double role both in the surveillance during the formulation of financial reports and the conduct of audit procedures. Moreover, in case the auditors discover suspicious financial anomalies, the audit committee shall safeguard their right and possibility to mention these anomalies to administrative layers above the ones that might have induced them and, thus, discharge them from any managerial stress to conceal potential abuses. The significance of all these components of "audit quality" is supported by prior literature and was proven by Piot & Janin (2005). The importance of auditors in deterring earnings management is also underlined by Healy & Wahlen (1998) but they clarify that the audit process is not flawless and it is probable that failures can occur which foster the application of earnings management practices.
Nelson, Elliott & Tarpley (2000) aided by executives of one of the BIG 5 audit firms who participated in the research, probed the activity of auditors and the state under which they manage to restrain earnings management. The BIG 5 were initially formulated by the BIG 4 plus Arthur Andersen LLP which collapsed in 2002 after the revelation of Enron scandal. Arthur Andersen was responsible for deficient audits of Enron and did not manage to reveal the manipulation of its financial reports and prevent the scandal. As a consequence, the firm lost its customers and reputation. They recorded what they called "Earnings Management Attempts (EMA)" and based on these, they concluded that the auditors were more hesitant or inefficient to demand from the firm to restate its financial statements or to forestall earnings management in the following situations: i) there is not adequate proof of erroneous financial statements, because of ambiguous legislation ii) the magnitude of the errors detected was insignificant because as they say auditors' main task is to "identify and require adjustment of material misstatements, and many EMAs could be considered misstatements, so it is reasonable that auditors would be less likely to waive misstatements that they consider material." iii) the audited company is sizeable, as there is a positive relation of the auditors' salary and the magnitude and reputation of the firm they audit. As a consequence, auditors have the incentive to circumvent improprieties that cause inconveniences to the firms, under the fear of losing their customer. iv) EMAs that lessen the income of the present year.
3.3.2 BIG 4 vs "Second - Tier" firms
Lately, except for the known BIG 4 audit firms, other notable audit firms have come to the fore and act as their deputies. This could be expected, because in the passage of time, the volume of tasks that need to be performed by the auditors has been multiplied. Francis, Maydew & Sparks (1999) have documented the supremacy of BIG 6 (subsequently BIG 4) against the rest audit firms in detecting and confining earnings management attempts. This inference is extracted from the lower degree of abnormal accruals found in enterprises that are audited by one the BIG 6. The BIG 6 audit firms consisted of Deloitte & Touche, Ernst & Young, Arthur Andersen, KPMG, Price Waterhouse and the Coopers & Lybrand which merged with Price Waterhouse in 1998 and nowadays formulate the PwC (formerly Price Waterhouse). After this merge, they were established as BIG 5 and subsequently as BIG 4 as described in section 3.3.1. This merger and the collapse of Arthur Andersen are likely to have affected the quality of audit. Non-BIG 4 were given the opportunity to absorb the "redundant" auditors of the overlapping positions who, however, had acquired valuable knowledge and experience from the BIG 4 firms. In this way, the prestige of a smaller firm is enhanced and the quality of audit that it provides can be improved as well. Particularly referring to the merger of Price Waterhouse and Coopers & Lybrand, the employees could improve their level of expertise through the day-to-day interaction with other colleagues, gain a profound understanding of the new markets and therefore affect positively the quality of the provided audit because of the highly skilled personnel and the special attention that would be given by higher-level employees. As a result, the financial statements of BIG 4 clients are more representative of the genuine condition of the firm and are of higher quality. Apart from these, they also document the BIG 6's higher expenditures for acquiring in depth audit specialization and improve their practices in general. Supportive of these findings is the research conducted by Becker, Defond, Jiambalvo & Subramanyam (1998). Also, DeFond and Jiambalvo (1993), prove the effectiveness of bigger audit firms to object to the force of managers and issue the proper audit opinion.
However, the study of Boone, Khurana & Raman (2010) opposes to these pretensions. Firstly, they support that there is some degree of resemblance between the two categories of audit firms (BIG 4 and non-BIG 4) derived from their common concern of suffering lawsuits and status damage, as side effects of poor audits. Nevertheless, they admit that they differ in the degree of reliance on their major clients. This practically means that the BIG 4 are more resistant and can discard more easily high levels of managerial coercion. Afterwards, they find that there are no considerable discrepancies between the audit quality that the audit firms of both of these categories offer using as a benchmark the ease with which the audit firm is eager to compile an ominous statement. Nonetheless, the most important aspect of their study refers to the perspective of investors who view the financial statements that are audited by one of the BIG 4 as qualitatively superior with respect to the accounting information that they contain and can be used by investors, even though the results of the research do not justify this behavior. Except for the fame of the BIG 4 that affects the investors' point of view, this is also explained by the reasoning that in case of an audit default, and keeping in mind the BIG 4's prosperous financial condition, they will have better chances to gain large amounts of compensation.
The study of Francis & Yu (2009) proceeded one step further. They illustrated that there is a fluctuation across each of the BIG 4 audit firms, with respect to the quality of audit they provide, which is proportionate to the stature of the regional bureau in which they operate. They support the previously mentioned argument that just like bigger audit firms, greater bureaus within the same company are less contingent on single clients too. This happens because the supervision is stronger in these bureaus and do not easily permit audit misrepresentations and because the relative impact of losing a customer, even a major one is lesser compared to another relatively small bureau. In addition, the bigger a bureau is, the deeper the expertise of the auditors working in it. It is important to note that this characteristic is not transferable because it is developed based on each bureau and its customers. This is attributed principally to the numerous customers and colleagues who mutually exchange their business knowledge and experience through their interpersonal relationship and improve their skills.
3.4 EARNINGS MANAGEMENT LITERATURE REVIEW
This part of the literature review makes a retrospect in the most important past studies about earnings management. Conceptually, there are two broad categories of earnings management defined by the means that are used in order to manage earnings: Real and Accruals-based earnings management as distinguished by Cohen, Dey and Lys, (2007).
In the first category three figures are included that are thought to reflect the degree of earnings manipulation. These are the cash from operations, the discretionary expenses as well as the costs of production. In addition, the authors mention the three most common and powerful ways through which the earnings can be managed and influence the above measures. These are the hastening of sales, the presentment of small cost of goods sold and the decline in discretionary expenses. In the first case (hastening of sales), such a practice takes place through either the slackening of credit conditions for the materialization of transactions or extreme discounts provided to buyers. As a result, the earnings of the present year will rise. On the other side, however, the cash flows will decline. In the second case (small cost of goods sold) managers coordinate the production process in a way that more units than actually needed or can be sold are produced. Therefore, they manage to distribute the standard amount of fixed cost into a bigger number of product units thus expanding artificially the profit margin. Taking into consideration the ordinary level of sales and the increased cost associated with the various aspects of the production process (not only the fixed cost), it is reasonable that cash flow from operations will be reduced. The last method (decline in discretionary expenses) concerns these types of cost that managers can influence according to their discretion, such as advertisement costs or R&D costs. In this way they manage to augment the earnings and potentially the cash flows.
In the second category (accruals based earnings management) earnings management takes place as follows: When firms sell their products on credit, a revenue accrual is created by this transaction. However, firms can also create accruals (through earnings management) that change the income. These are known as the discretionary accruals and can either increase or decrease income. Activities of this category include for example the undervaluation or overvaluation of warranty costs and/or inventories.
Calculating the discretionary accruals and reaching conclusions about earnings management has been widely used in prior earnings management literature. Many models have been developped for this reason. The most important ones are presented in the next section. However, models for the examination of real earnings management in prior literature are limited. This is reasonable because it is difficult to examine earnings management which takes places with real and not accrual-based activites, since real earnings management activities, such as the overproduction of inventory, are not directly observable in the financial statements. Zang (2012) examines the two categories of earnings management and documents that these two can serve as substitutes. When managers manipulate earnings they make the choice between real and accruals-based earnings management depending on the relative costs they will incur for each one of them. He makes use of a model initially developed by Roychowdhury (2006) in order to find the ordinary amount of production level and discretionary costs (such as advertisement costs). With this model, he can assess any deviation from the usual level and investigate real earnings management. Particularly, the magnitude of the residuals of the following equation denotes the level of overproduction and subsequently the inflation of earnings:
PRODt/At-1 = a0 + a1 (1/At-1) + a2 (St/At-1) +a3 (Î”St/At-1) + a4 (Î”St-1/At-1) + et
PRODt is the sum of the cost of goods sold in year t and the change in inventory from t-1 to t and Î”S is the change in net sales.
Similarly, the residuals of the equation below, multiplied by (-1) signal decline in discretionary expenses and as a result, inflate earnings:
DISXt/At-1 = a0 + a1 (1/At-1) + a2 (St-1/At-1) + et
Where DISX means discretionary expenses.
As he states, the difference between real and accruals-based earnings management is that while real earnings management activities change the actual transaction, the accruals-based earnings management activities change the way a transaction that has taken place is presented in the financial reports. This is feasible with the choice of the "appropirate" accounting choice. The results of his research reveal an exchange between real and accrual-based earnings management. In fact, they use both techniques to some extent. However, he makes an important remark: "real activities manipulation has to occur during the fiscal year, but accrual manipulation can occur after the fiscal year end". This gives managers the opportunity, at year end, to manipulate accruals in a way that is inversely related to the extent of real earnings management. This means that if real earnings mangement is exercised to a great extent within the fiscal year, at year end it will be counterbalanced by a low accruals-based manipulation. He also finds that in cases where there is deep monitoring and tight accounting framework with not so many alternatives, the companies prefer real earnings management activities which are harder to detect. On the contrary, accruals manipulation is preferred in companies with low competitiveness and higher tax costs.
The senior studies dealt mainly with two categories of incentives, - those that stem from the need to comply with rules and regulations and those that stem from the need to abide by the agreed covenants, referring to them as the main causes of earnings management. Dechow & Skinner (2000) state that: "as stock market valuations increased during the 1990s, especially in conjunction with the increased importance of stock-based compensation, managers have become increasingly sensitive to the level of their firms' stock prices and their relation to key accounting numbers. Consequently, their incentives to manage earnings to maintain and improve those valuations have also increased, which arguably explains why earnings management has received so much recent attention". The last decade though, the concept of accrual accounting attracted the researchers' attention as a leading contributor in earnings management practices. As Dechow & Dichev (2002) punctuate, when an economic transaction takes place, it is quite possible that the respective cash inflows or outflows associated with it, will not occur at that same time, but at a later point in time. The answer to this problem, in order to provide a complete view of the aggregate economic activity, is the use of accrual accounting, which aligns the moment that the transaction is recorded with that of the occurrence of the economic effects resulting from it. It is logical that this procedure requires personal evaluation of these transactions. Since these evaluations are not always either reliable or honest, it is expected that occasionally assessment errors will occur which distort the accruals quality. As a consequence, such false estimations could be considered as earnings management attempts, when made purposefully. Furthermore, Dechow & Dichev prove that there is an inverse relationship between the accruals quality and the size of this kind of errors.
Burgstahler & Dichev (1997) advert to two of the most ordinary purposes for which earnings manipulation is employed. These are the companies' intent to shy away from announcing either a decline in earnings or negative earnings (loss). Given the fact that stakeholders fixate on the attainment of financial goals, expressed in pure accounting figures, both cases provide the motivation to managers to do so, due to high expenses they would have to suffer in business dealings with them in the opposite case. In the first condition, as the duration of sequential raises in earnings is prolonged, the strength of motivation is enhanced. However, this inference does not hold for the second condition. Overall, earnings manipulation is more widespread for loss aversion compared to declines in earnings.
The study of Degeorge, Patel & Zeckhauser (1999) could be characterized as the sequel of Burgstahler & Dichev as they also underscore the role of accomplishing the benchmarks that have been set. Except for the two already mentioned ones, they also refer to the need for achieving the analysts' previsions. Both studies find that when earnings management occurs, it is highly probable to notice reported figures that merely reach the targets and highly improbable to see respective figures that merely miss the targets. Another contribution of their research is that they classify the three benchmarks according to their magnitude and the caution given to them by managers. Most important is ranked the statement of positive earnings, then, the statement of earnings which exceed a benchmark, such as previous year's earnings, and least important is the attainment of analysts' previsions.
Dechow & Skinner (2000) report the difference in earnings management perspectives between the regulatory bodies and the academic community. Specifically, academics are more reassuring about the importance of earnings management and the damage it can incur to the related parties, compared to regulatory bodies which are more worried about it. The authors admit that this happens due to the inadequacy of research techniques used by academics to detect earnings management, the deviation regarding the importance they attribute to various species of managers' motives for earnings management and, last but not least, the existing discrepancy about how reasonably the members of capital markets behave. Broadly, they argue that academics undervalue the significance of market-related motives. The authors' phrase: "our natural tendency to assume investor rationality has caused us to ignore capital market incentives for earnings management" is indicative of this. On top of this, they face difficulties in the calculation of earnings management levels and they rely on investors' ability to identify and confront earnings management. Regulators from their side, see a fraction of earnings management as necessary evil because of the freedom provided to managers. Also, they do not assume that investors are always adequately reasonable when making decisions. Contrariwise, they are convinced that they cannot perceive earnings manipulation at a timely basis and this inability is amplified by the fact that the access to useful information is quite expensive.
Cohen, Dey & Lys (2007) found a diachronic and gradual increment of the aggregate level of earnings management. Prior to the previous decade's scandals, like Enron, a drop in earnings management associated with the manipulation of accruals was observed, but the authors assert that this did not happen exclusively because of the modification in the SOX legislation. Perhaps other causes led to this change too, like the prudence shown by auditors in the wake of recent scandals. In fact this drop confessed an exchange, not a decline, between real and accruals earnings management, as the practices that belong to the category of real earnings management showed an upward drift. This kind of practices are thought to be more expensive for the firm but are more favorable for the managers as they are not easily tracked. They also wonder about the results of regulations in limiting earnings management: "some argue that these frauds occurred after 70 years of ever-increasing securities regulation" implying indirectly that plurality of laws might prove ineffective.
Cohen & Zarowin (2008) relied on the previous study and developed it, documenting the consequences of both real and accruals earnings management for the firms employing it. Regardless of which of the two was used, they found corroborative evidence that earnings management impacted the firm as follows: the level of investments augmented in the relative years of earnings manipulation and shrunk afterwards. Moreover, the joint effect of these practices was stronger than that of just a single of them. Apart from this, the implications incurred by the firm were of similar magnitude irrespective of the species of earnings management.
3.5 EARNINGS MANAGEMENT ESTIMATION MODELS
According to Beneish (2001), the methods applied by researchers in order to track earnings management practices are sorted into three main classes. The first one examines the discontinuities in the flow of earnings. However, Durtschi and Easton (2005) doubt the validity of this method. Specifically, they prove that the discontinuities in the distribution of earnings might potentially signal earnings management activities, but this is not always the case. In order to be sure that such discontinuities are responsible for earnings management deeper research has to be conducted. They document that factors such as the bias introduced in the data selection process, deflation and the restatement of quarterly data are liable for the form of earnings distribution, instead of earnings management. The second one deals with particular accruals, the manipulation of which is thought to possess explanatory power of earnings management, such as forecasts for bad debts. As Healy and Wahlen (1999) document the method of study of specific accruals is mainly used for firms that operate in the financial industry, like banks and insurance companies. Regarding the current study, the financial industry is excluded from the research as stated in section 5.3.Apart from that, they also state that these tests are weak because they do not take into consideration the conditions in which managers are more motivated to manipulate earnings, such as to beat the forecasts of financial analysts. The last one relies on the computation of discretionary accruals and their separation from non-discretionary, after having computed total accruals. For the above mentioned reasons, a model of the last category is considered more appropriate for the current empirical research. First of all, as stated earlier, earnings is the main figure used the evaluation of a firm's performance. However, for a limited number of years - and not the whole lifetime of a firm - there are timing and matching problems associated with cash flows. Such issues can be resolved if revenues are recorded only when a product or a major part of it is sold and its payment is considered relatively certain. In addition, this problem requires the mapping of revenues with the corresponding costs that relate to them. As a result, cash flows need to be "modified" in order to mirror the true performance of the firm. These modifications are operationalized through accruals. The figure of earnings consists of the sum of accruals plus cash flows. Managers, in turn, are responsible for the adjustments to the level of accruals. The use of managerial discretion can benefit the proper measurement of performance if applied correctly. However, such discretion sometimes results to purposeful inaccurate estimates of accruals or errors for earnings management objectives such as earnings smoothing. Consequently, the operation of accruals is vital for the estimation of a firm's performance the true picture of which can be distorted from the manipulation of accruals. This last category of earnings management recognition process involves various estimation models the most known of which are the following: the DeAngelo Model, the Healy Model, the Jones and the Modified Jones Model and the Industry Model. These models are assessed in Dechow, Sloan, and Sweeney (1995).
3.5.1 Healy Model
The model developed by Healy anticipates that earnings manipulation takes place in every year that is examined. It attempts to detect earnings management with the use of average total accruals divided by lagged total assets. The procedure starts with the distinction of the total sample into three categories. In one of them, earnings are anticipated to be inflated and in the other two categories of the total sample(?) earnings are anticipated to be deflated. Then, the mean total accruals of the former category (considered as the estimation period) are compared with each of the mean total accruals of the latter category (event period). The mean total accruals are finally employed to proxy for the non-discretionary accruals. The Healy Model is illustrated by the following formula:
NDAÏ„ = Î£(Î¤Î‘t) / T
NDAÏ„ = non-discretionary accruals
TA = total accruals divided by lagged total assets
t = 1, 2, â€¦ T years of the estimation period, where T denotes the number of years of the estimation period.
Ï„ = a year in the event period
3.5.2 DeAngelo Model
The DeAngelo Model is a specific category of the Healy Model and uses total accruals as a proxy for non-discretionary accruals. In this occasion, the non-discretionary accruals are calculated as last year's total accruals divided by lagged total assets. The DeAngelo Model aims at detecting earnings management by calculating total accrual first differences (total accruals in year t minus total accruals in year t-1). The model is portrayed by the following formula:
NDAÏ„ = TAÏ„-1
NDAÏ„ = non-discretionary accruals
TAÏ„-1 = total accruals divided by lagged total assets
Ï„ = a year in the event period
When a firm is in a steady condition, the amount of non-discretionary accruals in a particular year t equals the corresponding amount of non-discretionary accrual in year t-1. As a consequence, a potential discrepancy of the non-discretionary accruals of these consecutive years is associated with earnings management activities. However, the DeAngelo Model, like the Healy Model, can provide accurate and reliable estimations of earnings management only when non-discretionary accruals show stability for the period that is examined and the discretionary accruals have a mean value of zero.
3.5.3 Industry Model
This model was formulated by Dechow and Sloan in 1991. According to this model, the notion the non-discretionary accruals are constant in time is relaxed. Also, the assumption that the factors which influence the non-discretionary accruals are identical for various firm that belong in the same industry is prevalent. However, these factors are not stated explicitly. The formula used is the following:
NDAÏ„ = Î³1 + Î³2 median1(TAÏ„)
NDAÏ„ = non-discretionary accruals
median1(TAÏ„) = the median volume of total accruals divided by lagged assets for all non-sample companies with the same 2-digit SIC code.
Î³1 and Î³2 are estimated with the Ordinary Least Squares method form the sample of the estimation period.
This model has the ability to moderate the measurement errors. This is attributed primarily to two elements. The first one is that it eliminates the fluctuation of discretionary accruals which is correlated among firms which operate in the same industry. The second one is that it also eliminates differences in non-discretionary accruals that is common for companies that operate in the same industry
3.5.4 Jones Model
Like the Industry Model, in the Jones Model, the notion that non-discretionary accruals are constant in time is relaxed. Moreover, it is indirectly assumed that the revenues are non-discretionary. This implies that if earnings are recognized through discretionary components, the model will partially mitigate earnings that come from the discretionary accruals. The privilege of this model is that it controls for the consequences of the shifts in economic conditions that affect the non-discretionary accruals. The formula which has been developed is the following:
NDAt = Î±1 (1/At-1) + Î±2 (Î”REVt) + Î±3 (PPEt)
NDAt = non-discretionary accruals in year t
At-1 = total assets at t-1
Î”REVt = revenues in year t minus revenues in year t-1, divided by lagged total assets
PPEt = the gross property, plant and equipment divided by lagged total assets
Î±1, Î±2, Î±3 = firm specific parameters
From the following equation the estimates of these parameters can be derived, with the use of the Ordinary Least Square (OLS) method
Î¤At = b1*(1/At-1) + b2*(Î”REVt) + b3*(PPEt ) + Îµt
Where TA = Total accruals divided by lagged total assets.
3.5.5 Modified Jones Model
The function of this model is based on the distinction between the discretionary and non-discretionary accruals, just like the Jones model. However, there is a remarkable difference between the two. A limitation of the Jones Model is that it underestimates earnings management and subtracts part of the managerial discretion exercised on revenues because it is indirectly assumed that.all revenues are considered nondiscretionary. The Modified Jones Model cures this inefficiency correcting the propensity of the Jones Model to bias the estimation of accruals. While the Jones Model hypothesizes that all revenues are non-discretionary, the Modified Jones Model predicts that every single change in sales based on credit is considered earnings management. This is a logical recommendation since it is simpler to exercise discretion over revenues with respect to credit sales than it is for sales associated with cash. This difference between the two models is operationalized through the inclusion of a variable which controls for changes in receivables.
The structure of the model and the three requisite stages which will take place in order to test the two hypotheses are the following:
The first stage consists of the calculation of the aggregate (total) accruals through the equation:
Î¤At = b1*(1/At-1) + b2*(Î”REVt ) + b3*(PPEt) + Îµt
In this equation the symbols mentioned above denote the following:
ƒ˜ TA = the aggregate (total) accruals in year t, divided by lagged total assets
ƒ˜ At-1 = the total assets in year t - 1
ƒ˜ Î”REVt = revenues in year t minus revenues in year t - 1, scaled by lagged total assets
ƒ˜ PPEt = the gross property, plant and equipment in year t, scaled by lagged total assets
ƒ˜ b1 , b2 , b3 = estimates of the parameters Î²1 , Î²2 , Î²3 (explained below) originating from the process of ordinary least squares (OLS)
ƒ˜ Îµt = residuals of the equation
The second stage involves the computation of the non-discretionary accruals with the use of the following equation:
NDAt = Î²1*(1/At-1) + Î²2*(Î”REVt - Î”RECt) + Î²3*(PPEt)
The new variables included are:
ƒ˜ NDAt = non-discretionary accruals in year t and,
ƒ˜ âˆ†RECt which denotes the difference of net receivables between years t and t-1, divided by lagged total assets in year t-1
ƒ˜ Î²1 , Î²2 , Î²3 which are firm-related parameters
In the final stage we deduct the NDAt from the TAt and the result gives us the discretionary accruals (DA), which is the index of earnings management we ask for:
Î¤At - NDAt = DA
The absolute magnitude of the last equation depicts the extent of earnings management and given its inverse association with the quality of financial reporting, a higher value means less qualitative earnings.
Chapter 3, which completes the theoretical part, presents an extensive literature review of the determinants and other influential factors of earnings management, the consequences from the adoption of IFRS as well as the effects from voluntary versus mandatory adoption. Moreover, it includes the implications of IFRS on earnings quality and earnings management in code-law countries and the associated advantages and disadvantages. Also, the role of auditors in general and the distinction between the BIG 4 audit firms from "Second - Tier" audit firms are included too. Its last part refers to earnings management estimation models of prior literature. Finally, literature review tables which summarize the sample, method and findings of previous researches can be found in the Appendix.