Jones' Model and Its Modifications in the Conditions of the Slovak Republic

. One of the key tasks of financial accounting from its beginnings to the present day is to determine the performance of the company. The financial statements should provide users with a true and fair view of the financial position and financial performance of the entity during the period. At present, profit represents the most frequently accepted measure of a company's financial performance. An important prerequisite for profit as a reliable measure of performance is its quality, which can be influenced by various factors or techniques resulting from earnings management. This paper aims to compare the detection capability of the Jones model and its modifications for assessing the occurrence of earnings management in the conditions of the Slovak Republic. We use the regression analysis and comparison method, based on which we compare the detection capability of the Jones model and its modifications for assessing the occurrence of earnings management in the conditions of the Slovak Republic. The contribution of the paper lies in the observation of the Jones model and its modifications to determine a suitable model for assessing the existence of earnings management in companies in Slovakia, which will be the subject of future research.


Introduction
One of the key tasks of financial accounting from its beginnings to the present day is to determine the performance of the company. [19] The financial statements should provide users with a true and fair view of the entity's financial position and financial performance during the period. [10] At present, the company's managers and analysts use a wide range of tools, algorithms, and methods to determine the value of profit representing the most commonly accepted measure of a company's financial performance, to predict the future development of the company's financial health. [12] The relevance of the accrual principle applied in accounting is undeniable. [25] Accrual accounting is, thanks to the time and material allocation of costs and revenues, a good system for evaluating the effectiveness of management, based on which it is possible to measure and detect earnings management. [14] Earnings management is one of the most challenging, debated, controversial and at the same time the most promising topics in finance and financial management. [11] Nevertheless, it is still difficult to provide a uniform definition of earnings management, as evidenced by the fact that the authors of various professional studies very often perceive earnings management differently. The main reasons for the lack of consensus on earnings management are its inconsistency, ambiguity, and problematic measurability. Studies aimed at detecting and measuring earnings management are very often based on discretionary accruals and work with models that estimate the discretionary share of reported profit, starting with simple models that identify discretionary accruals with total accruals to more sophisticated models that try to divide the total accruals into a discretionary and nondiscretionary part. However, systematic evidence regarding the relative performance of these models in detecting earnings management does not yet exist.

Earnings relevance
The question of why profit is such an important indicator of a company [15] that it has become the subject of management and manipulation is answered by Ronen and Yaari [6] explaining three approaches based on the separation of shareholder ownership and management in publicly traded companies. This creates a conflict of interest between the parties, as management decisions may not always be in line with the views of business owners. Depending on the awareness of existing third parties about the true profit and the power to implement the decisions taken, these authors distinguish between a costly contracting approach, a decision-making approach, and a legal approach. The costly contracting approach emphasizes the importance of profit information in creating effective contracts. In the event of unforeseen situations, the parties remain bound by the old contracts and use the earnings management application to fulfill their contractual contracts. In terms of informing third parties about the true profit and the power to make decisions, the first approach presupposes complete information and the power of shareholders to control the management of the company. The decision-making approach describes profit as a valuable source of information for decision-making in a given company. If the shareholders are rational, it is not possible to carry out earnings management without the implicit or explicit consent of investors. The decision-making approach presupposes power, but not shareholder awareness. The legal approach is based on the lack of tools for business owners to effectively control management. However, profit allows owners to use limited tools more effectively, helping to eliminate information asymmetries between owners.

Earnings management
The phenomenon of earnings management belongs to the integral and fundamental part of their business finance. [16] Earnings management is a legal and very effective method of accounting techniques and may be used to obtain specific objectives of the enterprises involving the manipulation of accruals. [23] However, despite the unequivocal legality of earnings management, the creative nature of these activities, which show the financial situation and performance of the company according to management's ideas and not according to the needs of users of financial statements, cannot be denied, as McKee [26] confirms. The first of the definitions relating to earnings management by Degeorge, Patel and Zeckhauser [4] reads Strategic use of managerial judgment in influencing profit values reported by external users of financial statements. It is achieved primarily by timing reported or actual economic events to shift between periods. Other authors define EM as the use of flexibility in the selection of accounting methods that influence managers' decisions about future cash flows. [13] Sosnowski [24] explains that reducing financial limitations of enterprise can cause earnings management because enterprise tries to obtain higher external capital.

Earnings management testing
The issue of measuring and detecting earnings management is a relatively frequently researched problem. Despite a large amount of professional literature, there is no uniform opinion on the most effective testing methods. In the researched literature, the method using residues from aggregated accrual models, which is based on the classical Jones model, is the most used. The second most used method is models based on specific accrual accounts. In research, the most used residual model from aggregated accrual accounts in most cases provides results that indicate evidence or confirmation of the use of earnings management. This model is also cited by Dechow, Sloan and Sweeney [17] as the model with the strongest and best results. Although he is not talking exactly about residues from aggregated accruals, he is talking about the modified Jones model, which is the basis for it, so they can be considered identical. On the other hand, Dechow, Richardson, and Tuna [18] use the modified Jones model to test accrual accounts as explanatory variables for the skewed, non-standard profit distribution that is identified, for example, by Burgstahler and Dichev [2].

Jones model and its modifications
The present review of the scientific literature related to earnings management points to diversity [1] because the Jones model and its modifications use the different intensity of earnings management testing. [7] Jones's model mitigates the assumption of nondiscretionary accrual constant. This model takes into account the impact of changes in the economic situation of the company on the non-discretionary accrual with the help of tangible fixed assets and changes in sales as independent variables of the model. Jones's model divides the time series of companies' profits into two stages, which are estimation periods and event periods. The estimation period represents zero discretion accrual. The event period assumes that the discretionary accrual is not zero, resulting in earnings management. The main assumption of Jones' model is that companies do not implement earnings management during estimation periods. This assumption is very unlikely to be met, which may result in bias in the measurement and detection of earnings management. [5] Jones' model includes the assumption of non-discretion in sales. If a company manages profit through discretionary revenues, then Jones's model removes part of the profit from the discretionary accrual estimate, which the author herself acknowledges as a limitation of her model. [8] The consequence of the limitation is the fact that the given model detects earnings management, even in its absence. [27] Dechow, Sloan, and Sweeney [17] sought to streamline Jones' original model and therefore proposed to adjust the variable change in sales (ΔREV) to the variable change in receivables (ΔREC). The purpose of such a modification was to reduce the measurement error of the discretionary accrual if the discretionary accrual was generated through revenues. Jones' modified model points to the assumption that any change in sales revenue from a supplier credit is the result of earnings management. However, it is uncertain in the literature whether a given modification improves the original Jones model. The authors dealing with this issue, Jeter and Shivakumar [3] consider such an assumption to be unrealistic and therefore tried to design their model. As part of the creation of their model consisting in the modification of the original Jones model, they proposed to include cash flow from operating activities in the non-discretionary part of earnings management, due to its more effective detection within a group of companies with extreme cash flow reporting. The author Kasznik [20] also tried to improve Jones' original model, incorporating cash flow (FOCFO) as the third explanatory variable in his model. Another of the authors on the subject, Key [9] added a new variable to the original Jones model, which is intangible assets, and justified this transformation by using the relationship between intangible assets and depreciation, which are one of the components of the non-discretionary accrual. Teoh, Welch, and Wong [21] also attempted to modify Jones' model and focused on short-term accruals. They defined short-term accrual as the difference between changes in non-monetary current assets and changes in operating current liabilities. They determined the non-discretionary part of the short-term accrual by increasing the company's sales. Kothari et al. [22] presented a modification of the original Jones model, which is also known as "performance matching" while offering 2 approaches. The first is the pairing of similar companies, which alleviates the need for the least-squares DA estimation. The second, linear approach, is to consider the return on assets (ROA) to regulate the performance of the organization.

Methodology
The conference contribution aims to compare the detection capability of the Jones model and its modifications for assessing the occurrence of earnings management in the conditions of the Slovak Republic. For achievement relevant data we use the financial database, Amadeus. In our research, we analyze the Jones model and its modifications in the conditions of the Slovak Republic. Our analysis is applied to a sample of 1253 enterprises in the condition of the Slovak republic. In research, we use a selection method and comparison method, based on which we compare the detection capability of the Jones model and its modifications for Slovak enterprises according to the adjusted coefficient of determination, standard deviation, and the level of statistical significance of the earnings model as such and individual independent variables entering the analyzed earnings models. Within the analysis of statistical significance, we use the F -the test by the medium of setting hypotheses H0 and H1.

Results and discussion
We analyze profit models using criteria proposed by the literature, which are adjusted coefficient of determination, standard deviation, and the level of statistical significance of the profit model as such and individual independent variables entering the analyzed profit models. We apply individual profit models to a sample of 1253 Slovak companies, within which we selected companies according to the following criteria.

The adjusted coefficient of determination
The adjusted coefficient of determination expresses the explanatory power of the profit model. We performed a statistical regression analysis to determine the values of the adjusted coefficient of determination. Using the adjusted coefficient of determination, we then explain the variability and relevance of the analyzed profit models, by comparing the obtained results over time and between models, with higher values indicating a higher explanatory power of the model and vice versa.

Standard deviation
Through the analysis of the standard deviation, we explain the variability of the values of the variables of the selected profit models. If the standard deviation becomes low, it means that the given profit model more effective measures and detects earnings management in a selected sample of companies operating in the conditions of the Slovak Republic. The following tables provide a summary of each profit model.

ROA
H o H o H o 0.00 % Source: Own processing.
Based on the results from the table, we see that the Jones model achieves a high percentage expression of the statistical significance of the variables contained, especially in the case of the PPE variable, which appears to be statistically significant throughout the analyzed period. However, the variable ΔREV is recorded for statistical significance only in the last year of 2017, and therefore the percentage expression of the statistical significance of this variable is at the level of 33.33%. The modified Jones model achieves a high percentage of the statistical significance of all independent variables. The variable ΔREV -ΔREC is statistically insignificant only in the first year of 2015, the variable PPE is statistically significant in the whole analyzed period. The Jeter and Shivakumar model also achieves a high percentage of the statistical significance of the variable only in the case of PPE. The remaining variables appear to be statistically significant and their percentage of statistical significance reaches the level of 33.33%. The Kasznik model has a statistically significant variable, which is PPE in all three years, and a ΔREV variable, which is statistically significant in only one of the three analyzed periods. The variable ΔCFO is statistically significant only in two of the three analyzed periods, while its percentage of statistical significance reaches the level of 66.67%. The Teoh, Welch, and Wong model contains only one variable ΔSALE -ΔREC, which, however, is statistically significant only in the last year of 2017, while its percentage of statistical significance reaches the level of 33.33%. The variables contained in the remaining models appear to be statistically insignificant for all three years, with the result that they do not explain the manipulation of the discretionary accrual. Although the key model achieves a high percentage expression of the statistical significance of the PPE variable, the ΔREV variable is statistically insignificant, reaching a level of statistical significance of 33.33% and the IA variable is statistically significant, while its statistical significance is at 0%. Kothari's model also contains a statistically insignificant variable ROA, whose level of statistical significance reaches 0%.

Final evaluation of the achieved results
We evaluated selected analyzed profit models based on three criteria, which adjusted the coefficient of determination, standard deviation, the level of statistical significance of individual profit models, and statistical significance of variables contained in the analyzed profit models. The following table provides an overview of the criteria met by the given profit models for the examined period of three years.

Conclusion
The analysis was focused on the measurement and detection of earnings management in the conditions of the Slovak Republic using the Jones model and its modifications. To find out whether the Jones model and its modifications can capture and detect earnings management in the conditions of the Slovak Republic, we set four criteria, based on which we assessed the effectiveness of selected models. The four main criteria included the adjusted coefficient of determination, the standard deviation, the statistical significance of the profit models, and the statistical significance of the variables contained in the analyzed profit models. The highest explanatory power represented by the average value of the adjusted coefficient of determination was achieved by the Modified Jones model at the level of 0.0973, which means that the model explains the variability of the total accrual to 9.73%. The second in order is Kothari's model with a value of 0.0966. The following are Kasznik's model with a coefficient value of 0.0736, Jeter, and Shivakumar with a coefficient value of 0.0648, Jones's model with a value of 0.0601, Key model with a value of 0.0590. The Teoh, Welch and Wong models show the lowest explanatory power expressed by the average value of the adjusted coefficient of determination at the level of 0.0383. The second criterion was the standard deviation of the variables contained in the analyzed profit models. In terms of this criterion, the Modified Jones model was the most effective, mitigating the high standard deviation of the first variable of the original Jones model. Its modification into the variable ΔREV -ΔREC caused a decrease in the average value of the standard deviation to 31% of the average value of the coefficient. The average standard deviation of the PPE variable reached almost 25% of the average value of the coefficient. The Teoh, Welch, and Wong models also achieved low values of the mean standard deviation. The statistical significance of the profit model itself was met by several models, including the Jones model, the Modified Jones model, both cash flow models, the Key model, and the Kothari model. The percentage expression of statistical significance in the examined period reaches the level of 100% for all analyzed models. In terms of the statistical significance of individual variables contained in the analyzed profit models, the Modified Jones model was the most effective. This model achieves a high percentage of statistical significance for both. The Kasznik model satisfies the last criterion. Based on the above findings, we consider the Modified Jones model to be the most effective in the conditions of the Slovak Republic.