ISSN: (Online)

Transcription

1 ISSN: (Print) ISSN: (Online) THE ROLE OF DISAGGREGATION OF EARNINGS (CASH FLOWS AND ACCRUALS) IN STOCK VALUATION AND EARNINGS FORECASTING AT ACCEPTED COMPANIES IN TEHRAN STOCK EXCHANGE, IRAN MOHSEN HOSSEINI a1 AND AHMAD MODARRES b a Department of Accounting, MA of Islamic Azad University, Qazvin Science and Research Branch, Iran b Department of Accounting, Faculty Member of Tehran University and Khatam Higher Education Institute, Iran ABSTRACT The accounting earnings and its components are considered as the basis for making decision and forecast by the beneficiaries. The reported earnings in financial statements can be divided into two cash and accrued parts. The present research is mainly intended to compare two techniques including earnings disaggregation method (cash flows and accruals) and total earnings method where each of which is based on forecast of earnings and stock market value. In this study, a sample including 79 accepted companies in Tehran stock Exchange (TSE) were selected and the needed data for this investigation have been extracted based on the information from the sampled companies within time range ( ). The correlation coefficient method was employed in this survey and statistical data analysis was conducted at 95% level of confidence as well as testing of hypotheses with multivariate linear regression. With respect to the results derived from the conducted studies and comparison among two methods, the research findings suggested the preferred potential of earnings disaggregation method (cash flows and accruals) in respect to total earnings technique in forecast of earnings and stock market value. These findings are consistent with research theoretical bases and the purposed claim in the given hypotheses. KEYWORDS: Components of Earnings, Cash Flows, Accruals, Earnings Forecast, Stock Market Value The accounting earnings has drawn the most attention from users of financial statements including directors, financial analysts, and investors and either of them typically seeks for their own profits in this field. In order to process and interpret information, financial analysts need to appropriately perceive the components of the reported earnings. Paying attention to earnings Abstract components may contribute to users of financial statements to judge and evaluate better the future value in stock market and eventually the earnings and efficiency of their own investment. The findings indicate that disaggregation models of earnings components should lead to the best forecast for future earnings and the most accurate evaluation of stock market value. But, results and consequences of analysis on total earnings into two cash and accrued parts are unknown to great extent (Wang, 2013). Rather than paying attention to final value and total earnings, analysis on earnings components and specifications of these components are also important in order to analyze accounting earnings and or for its use. Using accruals and current cash flows for prediction of components of future earnings includes some economic concepts. The relevance of current accruals to future cash flows may show that to what extent the management intends to indicate corporative economic value honestly and this possibility that to what extent the future assets can be converted into realized cash flows since the stronger potential for prediction, the greater possibility will exit for their realization. Finally, accounting earnings, which possess more potential for forecasting components of earnings, may probably more reliable and they have stronger relationship with stock return (Modarres & Abbaszadeh, 2008). Whereas there is no perfect information about the relation of forecast and evaluation of earnings components therefore it is valuable to explore empirically into the concepts of information content in two accounting techniques i.e. total earnings (aggregate) method and earnings disaggregation method (cash flows and accruals) to forcast eranings and stock market value. In the current investigation, the role of disaggregation earnings components in forecast of earnings and stock market value and relative performance of both methods and aggregation of earnings into cash flows and accruals and total earnings have been examined. LITERATURE REVIEW To evaluate potential for net earnings from the input cash flows and forecast of future earnings, investors, creditors, and other users of financial statements use the reflected information in accounting earnings. Presentation 1 Corresponding author

2 of useful information for decision making is one of the important objectives in financial reporting. Given that earnings includes two parts i.e. in cash and accrued one so it is expected that the greater share of cash part in total earnings than its accrued part, the given earnings will be more reliable as well. This issue will cause the companies to have more potential for distribution of their disaggregation earnings and investors and financial analysts evaluate such companies as companies with appropriate performance. These factors cause to increase stock price and thus stock returns in these companies (Dechow et al; 2008). Through admitting the accrued basis for calculation of accounting earnings, it can be claimed that accounting earnings is composed of two cash and accrued parts. The researchers generally believe in that compared to cash flows, the accrued part of earnings may be appropriate parameters for forecast of earnings in the next years. Nevertheless, with respect to controversial nature of accruals it is not clearly obvious that composition of information content from cash and accrued earnings compared to total earnings can have better performance in forecast. In addition, accrued components depend on accounting rules and they are arbitrary components and many accruals may contain of some estimation such as inevitable errors and at the same time it can be probably manipulated by the management. Therefore, it should not be surprising that financial analysts often focus more on forecast of future earnings than on its two components (cash flows and accruals). Despite of this fact, the value and importance of earnings components depend on forecast of future earnings and cash flows (Wang, 2013). There are various attitudes concerning to information content in cash components and accruals of earnings regarding forecast of earnings and determination of corporative stock market value. Some experts argue that the assumption of cumulative nature of earnings components is a latent hypothesis in accrual accounting. This assumption denotes that sum of either of incomes and total income is determined by sum of either of costs and total cost and by subtracting total cost from total incomes the net profit results. Thus, there is no preference among earnings components (cash part and accrual). In contrast, with respect to literature of analysis on financial statements, Sloan (1996) deduces that since accruals are much subjective than cash flows and also it is more possible for them to be affected by manager s arbitrary goals and also to include abnormal accruals therefore they are rarely repeated during next periods. Compared to cash flows, the accruals possess less potential for forecasting (Barth et al, 1999). In the conducted studies the lesser stability of accruals has been reported regarding evaluation and forecast than cash flows and Sloan (1996) showed that stock prices act in such a way that as if investors has no information about less stable accruals in the given earnings so this leads to improper forecast of future earnings and stock valuation (Wang, 2013). Due to the existing ambiguity and various attitudes concerning to information content of earnings components so it is valuable to explore empirically the effect of earnings information content in disaggregation (separate) and aggregate (total) accounting methods to forecast earnings and stock market value. Unlike the previous researches, in this study two following accounting methods have been modeled and compared: 1) Disaggregation earnings (cash flows and accruals) and; 2) Total earnings (aggregate) and review the relative performance in either of the given accounting techniques in forecast of earnings and stock market value with respect to real earnings and observed market value. RESEARCH HISTORY With dividing cash components and accruals of earnings, Wilson (1987) examined the role of information contents in earnings components compared to total earnings. The results of this study showed that earnings accrued components include incremental information content in comparison to its cash items. In a study, In-Mu Haw et al (2001) explored the information content of operational cash flows, earnings, and operational accruals in capital market in China. The results of this study signified the further information content in earnings than in operational cash flows. Similarly, in this investigation, the rising information content in arbitrary accruals was confirmed versus mandatory accruals. In other study, Barth et al (2002) examined the effect of earnings analysis on forecast of abnormal earnings and determination of stock market value. The findings showed that firstly disaggregation of earnings into two cash and accrued components throughout the

3 composed data might lead to reduced error in forecast but analysis of earnings into cash flows and four accrual components might minimize forecast error upon dividing of industries. By conducting a study based on derived evidences from Oman and USA securities and exchange center (bourse), Bassam (2002), examined on the one hand the potential of cash and accrued components of earnings in determination of corporative value and on the other hand the potential of these components in forecast of abnormal earnings, profit, cash flows, dividends, and rturn(efficiency). The results of this investigation indicated that accrued items and cash flows have explanatory potential to forecast abnormal earnings, future earnings, and in determination of corporative value. Moreover, using accruals and cash flows simultaneously enjoys more potential in forecast of future abnormal earnings compared to each of these components individually. Arab Mazar Yazdi et al (2006), explored the information content of cash flows and accruals in Iranian capital market by the relationship among stock return with earnings and its components. Their findings reflected that compared the operating cash flows, the earnings includes more information content and on the other hand the conducted investigations in this study signified the rising information content in accruals compared to operating cash flows. Yoder (2006), examined the potential of forecast for future cash flows in accrued models versus cash flows related models. The results of his study indicated that compared to cash flows related models, the accrued models have not more potential for forecast of future cash flows. He argues that as fluctuation is reduced in earnings and sale rate, the potential of forecast increases for the accrued models. Modarres and Abbaszadeh (2008), examined the potential for forecast of earnings and cash flows as well as their mutual effect on each other and concluded that cash flows had more potential for forecast of equities and one could forecast cash flows by means of earnings and for equities by the aid of cash flows as well. Habib and Al-Hamavi (2009), examined the information content of cash and accrued components of earnings in Egypt. They found that information content of both cash and accrued components affect on forecast of stock value as well as forecast of abnormal earnings. Similarly, there is positive correlation between cash flows with stock market value and also there is negative correlation among accruals and stock market value. Mohseni et al (2011), surveyed the impact of earnings disaggregation in potential for forecast of stock price in accepted companies in TSE (bourse). The results of this study showed that employing accounting earnings as a united figure in accounting based evaluation method might not be verified with this presupposition that components of accounting earnings had the same weight in forecast of stock price. In other words, the components of accounting earnings namely cash flow and total earnings accruals as well as the main components of accrued components have different information contents in which this incremental information content will increase potential for forecast and interpretation of corporative stock prices. In his investigation, Wang (2013), explored the potential for forecast of earnings and corporative value by dividing earnings components into cash and accruals and comparing them with total earnings. The given results from this study indicated that compared to total earnings system, the earnings disaggregation system includes generally further information content and it enjoys more potential to forecast earnings and stock market value. RESEARCH GOALS AND HYPOTHESES With respect to the aforesaid issues, the main goal of this study included the comparison of two method of earnings disaggregation (cash flows and accruals) and total earnings method (aggregation) and review on potential for forecast of earnings and stock market value in either of two methods and it has been tried to answer to this question that: Could it be increased in potential of forecast of earnings and stock market value by disaggregation of earnings into cash flows and accruals or not? With respect t the aforesaid objective and some of evaluation models and findings from the similar conducted studied including by Wang (2013), it is identified that earnings and its constituent components are some of effective factors on this value and they can forecast it; therefore, with dividing of accounting earnings into two parts i.e. cash flows and accruals, it has been expected in this investigation that two cash and accrued parts of earnings have more potential to forecast earnings

4 and stock market value versus total earnings method. Therefore, research hypotheses have been formulated to answer research question as follows: First hypothesis: As the earnings is disaggregated into cash flows and accruals, the potential is increased for forecast of earnings. Second hypothesis: As the earnings is disaggregated into cash flows and accruals, the potential of forecasting stock market value increases. METHODOLOGY The present study is of applied type in terms of goal and it is descriptive in terms of method and given that it searches for finding the significant relationship among research variables and also with respect to variance of variables, the limits of variance of dependent variable is independent so it is of correlation type. Multivariate linear regression has been utilized as statistical model in this investigation. The needed information about companies has been extracted via information banks including Rahavard Novin Software and TSEofficial website and then panel data technique was employed to test hypotheses by conclusion and the needed computations in Excel spread sheet. STATISTICAL POPULATION AND SAMPLING The research statistical population comprises of the accepted companies in TSE under the covered time period including 8-year term that ranged from the beginning of 2005 (21st March) to the beginning of 2013 (20 March). The statistical sample was selected by omission technique and based on the following criteria: 1) The financial information of the studied enterprise should be available. 2) To increase potential for comparison, the end of corporative fiscal year (Iranian calendar) is stipulated as 21 st March while the given company should not change fiscal year. The reason for selection of this criterion was this point that to selected similar time periods as possible and seasonal conditions and factors not to affect on selection of other factors and variables. 3) The enterprise should not be included in banks and financial institutes (investment companies, financial brokers, holding and leasing companies). Choosing this criterion is due to different structure of loss and profit statements in these companies compared to other companies. 4) In order to homogenize statistical sample during the studied years, the accepted companiesin TSE up to the end of Iranian calendar year 1383 (20 March 2005) and they have not exited TSE during years ( ) as well. 5) The transactional pause should not occurred in these companies for more than three months since corporative efficiency (return) depends on stock contracts and if corporative stock is subjected to transactional pause and interruption for a long period this may lead to inability for measurement of corporative efficiency with adequate accuracy. 6) Those companies should not incur any loss during the studied years. After enforcement the aforesaid conditions in this study, 79 companies were selected from the accepted companies in TSE (bourse), which were qualified concerning to research requirements. MEASUREMENTMODEL To measure research concepts in this section, the research variables and abbreviations and way of calculation of research variables are shown briefly in Table (1). It should be explained that research variables have been homogenized based on the total assets.

5 Table 1: Operational definition of research variables Type of variable Dependent Dependent Dependent Independent Independent Independent Dependent Dependent Title of variable Return on Equity Cash return on equity Accruals Return on equity Sum of Accruals operating cash flow book value of equity Operating income Stock Market Value Abbreviation ROE ROCF ROACC ACC CFO BV OI MV Way of calculation METHOD OF DATA ANALYSLS In order to analyze data in this study and comparison of two accounting methods i.e. 1)Earnings disaggregation method (cash flows and accruals) and 2)Total earnings method, three phases have been carried out by means of the conducted research by Wang (2013), as follows: First step: Before conducting test of the predicted models at second and third steps in this study, initially values of a1, a2, and a12 in two method of (2) earnings disaggregation (cash flows and accruals) and total earnings have been estimated respectively by means of Jack-Knif statistical technique (re-sampling) and execution of regressions (1) and (2) where their results can be seen in Table (5). After estimation of coefficients a1, a2, and a12, in order to forecast long-term mean Return on Equity (µ), Cash return on equity (µ1), Accruals Return on equity (µ2) and regressions (3) and (4) for earnings disaggregation method (cash flows and accruals) and regression (5) have been computed for total earnings method as follow: (1) µ (3) µ (4) µ= (5)

6 Second step: At this step, the expected earnings and stock market value of company (i) in year (t+1) have been forecasted in earnings disaggregation method (cash flows and accruals) and total earnings method respectively by execution of regressions (6), (7), (8), and (9) by means of estimated parameters derived from first step (CFACC superscript denotes forecast by means of earnings disaggregation method and superscript (OI) stands for forecast by total earnings method): a) Forecast of expected earnings by two methods of earnings disaggregation(cash flows and accruals) and total earnings: (µ µ µ +(1- µ = µ µ µ (6) = µ (7) b) forecast of stock market value by means of methods of earnings disaggregation (cash flows and accruals) and total earnings: (8) (9) Third step: After taking the first and second steps, with respect to the resultant coefficients and finally with execution of regressive model, first and second hypotheses have been tested by means of Eq. (10) and (11) respectively and this point may be implied that after regressions (10) and (11), reasoning and conclusion have been done by the derived results of the aforesaid regressions at various modes as it described in Table (2): (10) (11) Table 2: The way of conclusion from testing of hypotheses First hypothesis If γ 11 value differs statistically from zero and γ 12 value does not significantly differ from zero. If γ 11 value does not significantly differ from zero and γ 12 value statistically differs from zero. Other conditions Second hypothesis If γ 21 value differs statistically from zero and γ 22 value does not significantly differ from zero. If γ 21 value does not significantly differ from zero and γ 22 value statistically differs from zero. Other conditions The preferred technique Earnings disaggregationmethod (verification of hypothesis) Total earnings method No method is preferred to other one. The preferred technique Earnings disaggregationmethod (verification of hypothesis) Total earnings method No method is preferred to other one.

7 RESEARCH FINDINGS variables in the model does not lead to creation of false regression. Hardy test has been utilized for this purpose in Reliability of research variables was examined which null hypothesis denotes lack of reliability of before analysis of research data. The reliability of variable and the opposite hypothesis signifies reliability research variables denotes that mean and variance of of variables. As the results of this test indicate in Table variables over the time and covariance of variables are (3), p-value is smaller than 0.05 for all variables. Thus, all fixed during several years. As a result, employing these research variables are reliable during the studied period. Table 3: The given results from Hardy test Hardy test for each of variables Variables Test statistic P-value OI MV In the followings, as it explained in section of data analysis at first step, initially the needed coefficients for forecast of long-term mean Return on Equity (ROE), Table 4: The estimation coefficients at throughout sample size Cashreturnon equity (ROCF) and Accruals Return onequity (ROACC) have been estimated by means of Models (1) and (2) as it described in Table (4): Symbol Estimation Standard T-value P-value deviation Intercept C CFO/B ACC/B C ACC/B C B C OI/B Walt tests: a1 = a2; chi- square = , P-value = 0.000<0.05 Wald tests: a1 = a12 + a2; chi-square = , P-value = 0.000<0.05 As it characterized in Table (4), the assumption of equality of coefficients و ) was tested under the conditions where its null hypothesis denotes equality of coefficients and the opposite hypothesis signifies inequality of the coefficients by means o Wald test where the given results suggest rejection of null hypothesis. In other words, the deduction of Sloan (1996) had indicated that when cash flows and accruals have the same coefficients in earnings, total earnings will be the appropriate earnings for forecast so this issue is rejected in this study. Similarly, with respect to the given results from Table (5), it is and this denotes the more consistency in cash flows than in accruals while we have and this also means that compared to accruals, cash flows may present further information regarding prediction of future cash flows (Sloan, 1996; & Wang, 2013).

8 Finally at third step and calculation of long-term mean Return on Equity (ROE), Cash return on equity (ROCF) and Accruals Return on equity (ROACC) as well as prediction of the expected earnings and stock market value and also at second step by means of two methods of earnings disaggregation (cash flows and accruals) and total earnings, the research hypotheses have been tested as follows. THE GIVEN RESULTS FROM TESTING FIRST HYPOTHESIS The first hypothesis in this study expresses that with disaggregation of earnings into cash flows and accruals, the potential is increased for forecast of earnings. The derived results from test of regression (10) have been reflected in Table (5). Table 5: The given results from testing first hypothesis throughout the total sample Estimation of Standard deviation T-value P-value coefficients C Correlation Durbin-Watson 1.96 coefficient statistic Determination F-statistic coefficient Adjusted determination coefficient Probability value With respect to the given results in Table (5), positive sign of correlation coefficient indicates the direction relationship among variables. The adjusted determination coefficient value (0.581) signifies the explanatory potential of the model to describe dependent variable; in other words, this means of variance in dependent variable (operating income) has been explained by independent variables. According to Table, value of Durbin-Watson statistic is 1.96 (between 1.5 and 2.5) and this figure shows that errors are independent from each other and there is no correlation among errors as a result assumption of correlation among errors is rejected and regression can be used for this purpose. Given that F- statistic (Fisher) is significant at this level (pvalue=0.000<0.05), thus according to the conducted analysis, this model is totally significant at 99% level of confidence and also with respect to this fact that value of coefficient γ12 is and it statistically differs from zero (p-value= 0.000<0.05) and at the same time since value of coefficient γ12 is greater than significance level (p-value = 0.170>0.05) and it does not significantly differ from zero, the null hypothesis is rejected and the opposite hypothesis is verified that expresses the potential for forecast of earnings will be increased by disaggregation of earnings into cash flows and accruals. THE GIVEN RESULTS FROM TESTING SECOND HYPOTHESIS The second hypothesis of the study implies that potential for forecast of stock market value increases with disaggregation of earnings into cash flows and accruals. The derived results from the test have been reflected in Table (6).

9 Table 6: The given results from testing of second hypothesis throughout total sample Estimation of coefficients Standard deviation T-value P- value C Correlation coefficient Durbin-Watson statistic Determination coefficient F-statistic Adjusted determination coefficient Probability value With respect to the given results from Table (6), the positive sign of correlation coefficient indicates the direct relationship among variables. The adjusted determination coefficient is and this shows the explanatory potential of this model to describe dependent variable. Namely, 65.8% of variance in dependent variable (stock market value) has been interpreted by independent variables. According to this table, the value of Durbin- Watson statistic is 1.86 and this figure indicates that the errors are independent from each other and there is no correlation among errors as a result the hypothesis for correlation among errors is rejected so regression can be used. Given that F-statistic is significant at this level (p-value=0.000<0.05); therefore, according to the conducted analysis, the model is totally significant at 99% level of confidence and whereas value of coefficient γ 21 is and it statistically differs from zero (pvalue=0.000<0.05), but value of coefficient γ 22 significantly differs from zero (p-value= 0.341<0.05), thus, null hypothesis is totally rejected while the opposite hypothesis is verified which expresses that potential of forecast of stock matket value increases with disaggregation of earnings into cash flows and accruals. CONCLUSION With respect to type of information requirements for investors, it necessitates preparation of some information that enables investors to forecast the expected value for their own investment and making proper decision regarding them and one could improve potential for forecast of accounting information through developing various models by conducting furthers researches in this field and finding the significant relations among accounting information, components of earnings, and stock returns regarding requirements of investors and other users. This study was intended to compare two method of disaggregation of earnings (cash flows and accrual) and total earnings and to examine the potential of each of these methods for forecast of earnings and stock market value. The results showed that cash flows have been more consistent than accruals. Similarly, compared to accrued items, cash flows present further information regarding forecast of future cash flows. In general, the evident and resultant findings from this testing of hypotheses (first and second hypotheses) signify the preferred explanatory potential of the disaggregated models to aggregated ones in respect of forecast and finally in can be claimed that compared to total earnings method, earnings disaggregated method (cash flows and accruals) in terms of forecast of earnings (first hypothesis) and forecast of stock market value (second hypothesis) and it has more potential and preference than it. In other words, better dividing of earnings into its constituent components has led to more accurate forecasts of future earnings in companies and consequently given that forecast of corporative future equities is one of the effective factors on determination of stock market value thus this event may also lead to more accurate and better forecast regarding stock market value in companies. Likewise, the above results are complied with the given findings from studies done by Bassam (2002), Habib and Al- Mahavi (2009), Mohseni et al (2011), and Wang (2013). SUGGESTIONS FOR NEXT STUDIES 1. It is recommended to repeat this study with aggregation of sum of accruals with all their constituent components and through comparing the

10 given results about the review of improving forecast and explanatory potentials of this model. 2. It is suggested to repeat this study and to compare its results by means of net earnings instead of the used operational profit. REFERENCES Arab Mazar Yazdi M., Mashayekhi B., and Rafiei A., Information content in cash flows and accruals in Iranian capital market. Quarterly of accounting & auditing reviews., 4 : Barth M., Beaver W., Hand J., and Landsman W., Accruals, Cash Flows and Earnings Values. Review of Accounting Studies.,4(3-4 : Barth, M., Beaver W.and J., and Landsman W., Constraints on Accrual Components of Earnings in Earnings Valuation., Bassam M., The Role of Accruals and Cash Flows in Earnings Valuation: Evidence From Amman and The U.S Stock Exchanges. PHD Dissertation. Kent State University. Dechow, P.M, Scott A.R., and Richard G.S., 2008, The persistence and pricing of the cash component of earnings, Habib, W. M. and Elhamawy S.M.,2009. On the Value Relevance of Finance's Cash Flow and Accounting's Accruals: Empirical Evidence from Egypt. Journal of Economic Studies 2009., 3:1-22. Haw, In-Mu, Daqing Qi., woody Wu., 2001, The nature of information in accruals and cash flows in an emerging capital market: The case of China. The international journal of accounting., 36 : Mohseni, Abdul Reza,Osmani Q., Karbasi Yazdi M., The effect of earnings aggregation in predictability for stock price in the for accepted companies in Tehran stock Exchange (TSE). Journal of financial management and accounting perspective., 3 : Modarres A. and Abbaszadeh M., The analytical survey in effect of predictability for accrual items and cash flows on quality of the predicted earnings. Journal of knowledge & development., 24: Sloan r.g., Do stock Prices Fully Reflect Information in Accruals and Cash Flows About Future Earnings. The Accounting Review., 71: Wang p., The role of disaggregation of earnings in stock valuation and earnings forecasting, xfi centre for finance and investment., available from Wilson G. P., The incremental information content of the accrual and funds components of earnings after controlling earnings. The Accounting Review., 62: Yoder T. R., The Incremental Cash Flow Predictive Ability of Accrual Models, Working Paper, The Pennsylvania State University.