國立中山大學企業管理學系碩士論文 傳統會計衡量法 經濟附加價值法及市場附加價值法對於股票報酬率的解釋能力之研究 : 以泰國證券市場為例

Transcription

1 國立中山大學企業管理學系碩士論文 Department of Business Management National Sun Yat-sen University Master Thesis 傳統會計衡量法 經濟附加價值法及市場附加價值法對於股票報酬率的解釋能力之研究 : 以泰國證券市場為例 The explanatory power of accounting measures, EVA and MVA on stock returns: Evidence from Thailand stock market 研究生 : 李美雪 Ms. Suksom Charoendeesawat 指導教授 : 鄭義博士 Dr. Yih Jeng 中華民國 100 年 8 月 August 2011

2

3 i

4 Acknowledgement It would not have been possible to complete this graduate thesis without the advice, support and encouragement from my advisor, Prof. Yih Jeng. It is such a valuable opportunity for me to pursue my master research under his guidance. Thank you very much I also thank my graduate committees, Prof. Zang Shiwei and Prof. Lin Hsin-Hui for their suggestions to make my thesis more comprehensive. I would like to thank Christopher Clausen for helping me proofread this thesis, Chung Yeh Huang and Carol Tsou for their assistance in the translation and verification of Mandarin abstract. Also I thank National Sun Yat-sen University, College of Management as well as Julia Chan and Debbie Wu, coordinators of MBA program. Last but not least, I thank my friends, Tatchaporn, Apiruk, Kwannapha, Penpan, Pakorn and family members, especially my brother, Thammarat for their supporting and encouraging me throughout my master studies. ii

5 摘要 投資者的投資的目的是為了增加財富, 然而他們的財產往往隨著股票價值和殖利率而變動 因此投資者在進行投資之前, 必須使用財務指標來評估及預測公司的表現 在泰國, 投資者使用會計衡量法來評估上市公司的表現, 典型的財務指標包括 : 每股盈餘 總資產報酬率 股東權益報酬率和純益率 價值基礎衡量法包含 : 經濟附加價值和市場附加價值, 然而此衡量法仍尚未普及 因此本研究之研究目的在於檢視傳統會計衡量法和價值基礎衡量法對於股票報酬率的解釋能力 就每股盈餘及產業而言, 研究對象皆採用具代表性的上市公司 研究期間從 2006 年至 2010 年, 共計 190 間公司 實證結果指出, 相較於經濟附加價值和市場附加價值, 傳統會計衡量法與股票報酬率具有較高關聯性 在會計衡量中, 即使不同產業,ROA 指標對股票報酬率的解釋力最佳 相對地,EVA 在某些產業中, 與過去股票市場之研究結果並不一致, 此結論包含泰國 因此本研究亦延伸探討經濟附加價值與股票報酬率兩者之間具有關聯性的公司, 研究結果發現, 經濟附加價值與股票報酬率具有相關性的公司, 往往擁有較高淨值股價比 ( 成長預測值 ) 在部分投資組合報酬中, 傳統的投資模式如價值及成長策略仍優於 MVA 及 EVA 策略 關鍵字 : 會計衡量法 經濟附加值 市場增加值 股票報酬率 投資報酬 泰國 股市 iii

6 Abstract The primary investment objective of investors is to create their wealth which is reflected in the change of stock market price and dividend yield they receive over the investment period. Thus, investors need financial tools to assess and forecast company performance before making investment decisions. Traditionally, such accounting measures as Earnings Per Share (EPS), Return On Assets (ROA), Return On Equity (ROE) and Return On Sales (ROS) are basic tools for investors in Thailand to evaluate companies performance in the stock market. Value based approaches such as Economic Value Added (EVA) and Market Value Added (MVA) are not widely known among investors yet. Therefore, this study aims to examine the explanatory power of various accounting measures (EPS, ROA, ROE and ROS ) and value based measures ( EVA and MVA ) on the stock returns. This study focuses on 190 sample companies which are representative of all listed companies in the years from in terms of the spread of EPS and industry diversification. The empirical results indicate that accounting measures are more associated with stock returns than MVA and EVA respectively. Among accounting measures, ROA provides highest explanatory power on stock return although the analysis is done separately by sector. In contrast, the results for EVA appear in some sectors and are not consistent with the past research done in other stock markets including Thailand. Thus, the analysis is extended to examine the company characteristics that have relationship between EVA and stock return. The findings indicate that EVA tends to be associated with stock return in companies that have low book to market ratio. In terms of portfolio returns, typical investing styles, such as value and growth strategies still outperform the return from MVA and EVA strategies. Keywords: Accounting Measures, Economic Value Added, Market Value Added, Stock Return, Portfolio Strategy, Thailand Stock Market. iv

7 Table of Contents Thesis Approval Sheet..i Acknowledgement..ii Abstract (Mandarin).. iii Abstract (English) iv Chapter I Introduction Background Information Problem Statement Research Purpose Scope and Limitation Chapter II Literature review The Concept of Economic Value Added The Concept of Market Value Added Related Researches....9 Chapter III Methodology Sample Selection Variables Definition and Valuation Method.20 Chapter IV Empirical Study Descriptive Analysis Correlation Study Regression Analysis Portfolio Returns..37

8 Chapter V Conclusions and Suggestions for Further Researches Conclusions Suggestions for Further Researches References Appendix...44

9 List of Tables Table 1: Table for determining sample size from a given population Table 2: Correlation statistic matrix for companies in automotive sector in the years Table 3: Correlation statistic matrix for companies in commerce sector in the years Table 4: Correlation statistic matrix for companies in construction sector in the years Table 5: Correlation statistic matrix for companies in energy sector in the years Table 6: Correlation statistic matrix for companies in industrial sector in the years Table 7: Correlation statistic matrix for companies in food sector in the years Table 8: Correlation statistic matrix for companies in health care sector in the years Table 9: Correlation statistic matrix for companies in information sector in the years Table10: Correlation statistic matrix for companies in tourism sector in the years

10 1. Introduction 1.1 Background Information Maximizing Shareholders wealth has become the main goal for corporate financial management in recent years. Shareholders, the owners of companies, are the firm s most important stakeholders. The financial managers make business decisions in response to the interest of shareholders which means maximizing their wealth through the value of firm. Thus, the financial goal of firm, maximizing shareholder wealth is reflected by the increase of stock market price(baker & Powell, 2005, pp ). From the shareholders perspective, their wealth is measured in term of returns from their investment both in forms of cash dividends and capital gain. Capital gain depends on the changes in the market value of stocks over the period. For shareholders to define the firm value creation, they need some financial tools that enable them to assess and forecast firm s performance. Traditionally, accounting measures like Earning Per Share (EPS), Return on Equity (ROE), Net Operational Profit After Tax (NOPAT) and Earning Before Interest and Tax (EBIT) etc. have been used for evaluating company performance. However, the calculation of those accounting measures derives from financial statements under Generally Accepted Accounting Principles (GAAP) which rely on the accrual basis method. In addition, the accounting treatment can be different depending on the company s accounting policy. Therefore, the performance measures based on the accounting profit are easily manipulated and can t reflect other important factors such as future cash flow and the cost of equity. In consequence to the advent of globalization since the early 1980s, many economic sectors and the capital market had been less controlled by institutions and 1

11 became more accessible for investors. Shareholders grow to be more sophisticated and had more concerns on the company performance to assess their return. As a result, accounting measures based on balance sheet and profit and loss statements were inadequate tools for them. Cash flow had become a more important measure. New performance measurements, the discounted cash flow techniques, such as free cash flow valuation model, had been introduced to reduce the problems associated with the accounting measures. At that time, the concept of maximizing shareholder value had received attention from many firms and also was interested by shareholders. Consequently, Shareholders Value approach was developed in the late 1980 and early 1990 (Maditinos, Zeljko, & Theriou, 2005). This value based approach takes into account the economic value and based its calculation on forecasted cash flow and cost of capital. The value performance measures include Economic Value Added (EVA), Economic Profit (EP), Market Value Added (MVA), Cash Value Added (CVA), cash flow return on investment(cfroi) and total shareholder return(tsr)(fernandez, 2002b, pp ). Thus, this study will examine the explanatory power of the tradition measures: EPS, ROS, ROE and ROS, and value based measures: EVA and MVA with the stock returns. It will also explore further to portfolio returns by basic types of investment strategies: value and growth investing in comparing to MVA and EVA strategies. 2

12 1.2 Problem Statement Generally, investors in Thailand stock market assess the company performances based on accounting measures such as EPS, ROE and ROA since they are provided by Stock Exchange of Thailand (SET) which are easily accessible and understandable. Accordingly, other tools seem not to be in concern for their stock analysis which can lead to wrong investment decision. Thus, this study examines company s performances with additional tools of EVA and MVA in comparison with accounting measures and extends the comparison to the portfolio returns by typical investing strategies and EVA, MVA based strategies. To explicit the explanatory power of those performance measures over the stock return, the closing price used for return calculation is at the month end when the company s financial statement is published by SET. This is to observe the relationship after reflecting with new financial information. 1.3 Research Purpose 1. To study the relationship between stock returns and MVA, EVA, EPS, ROE, ROA and ROS and to investigate which performance measure better explain stock returns. 2. To examine if MVA and EVA can reflect company performance and should be considered as additional measures for investors. 3. To measure portfolio returns by using EVA and MVA strategies in comparing to growth and value strategies that are typically used by investors. 4. To develop an understanding with regard to the relationship between different financial measurement tools and stock returns for investors used in their investment decisions. 3

13 1.4 Scope and Limitation This study focuses on listed companies on the Stock Exchange of Thailand by excluding securities, insurance, and banking industries since their operating activities and financial structures are different from other industries. Thus, the value from factor calculation can not be used for comparison across all industries. After excluding the financial industry, there are 7 major industries consisting of 22 sectors. A sample group of 10 sectors are selected from each major industry. Therefore, this study examines the explanatory power of performance measures and stock returns from these 10 sectors which are representative of all major industries. Also, these 10 sectors are defined by a range of EPS values that approximate the proportion of the population of companies. Accounting measures (EPS, ROA, ROE and ROS), EVA and MVA are calculated from all companies under those 10 sectors and the information used for calculation is taken from financial statements, stock prices and dividend payments published by the Stock Exchange of Thailand in the years Stock returns are composed of two parts. The first part is a dividend yield paid each fiscal year. Another is capital gain which is calculated by the closing price at the end of each February month as the SET provides a 60-day extension period for all listed companies to submit their financial statements. However, there is one limitation of this study. As the assumption based for investors, the calculation method used in the analysis should not be complicated for investors to understand and apply in their investment decision. Therefore, basic formulas that are generally acceptable are employed in the calculation. In computing EVA components, basic EVA without adjustment from GAAP operating profits and balance sheet is applied in this study (Shil, 2009). 4

14 2. Literature review This chapter is composed of two parts. The first part is to explain the concept of Economic Value Added and Market Value Added. Another part covers the relevant researches regarding the explanatory power of accounting and value based performance measures over stock returns from various stock markets including Thailand. 2.1 The concept of Economic Value Added Stern Stewart & Co. who own the trademark of EVA defined the definition of Economic value added via their website as follows: Economic Value Added is a measure of economic profit. It is calculated as the difference between the Net Operating Profit After Tax and the opportunity cost of Invested Capital. This opportunity cost is determined by the weighted average cost of Debt and Equity Capital ("WACC") and the amount of Capital employed. 1 According to Ehrbar (1998), EVA can be calculated by the following formula. EVA = NOPAT C%(TC) Where NOPAT is net operating profits after taxes C% is the percentage cost of capital TC is total capital EVA is not a new concept but it is one variation of residual income with adjustment on the calculation of income (NOPAT) and capital. From the formula, EVA includes total cost of debt and equity capital whereas accounting measures contain only the cost of debt capital in part of the interest expense (Chen & Dodd, 1997). 1 source: 5

15 There is no definite way to make an adjustment on EVA. Any company can decide the adjustment according to its business and strategy. Ehrbar (1998) describes level of adjustments for EVA as shown in the figure below. From the left side, EVA can be computed by Basic EVA which is the unadjusted EVA from GAAP balance sheet and profit and loss statement. Moving to the right, Disclosed EVA is used by Stern Stewart in its published MVA/EVA rankings and is computed by a dozen standard adjustments to publicly available accounting data. For Tailored EVA, each company can develop a tailored EVA definition in response to its organizational structure, business mix, strategy, and accounting policy. At the extreme right is True EVA. It is the most theoretically correct and accurate measure of economic profit after considering all relevant adjustments to accounting data (Ehrbar,1998,p.164-5). Figure 1 The EVA spectrum (Ehrbar,1998,p.165) Biddle, Bowen, & Wallace (1999) noted that Stern Stewart developed 164 possible adjustments to GAAP and to internal accounting treatments for improving the measure of both EVA combination: operating profit and capital. Some of the common adjustments are shown in table 2.1. These adjust items are the EVA leading components. For example, under GAAP accounting, items such as R&D and operating leases are charged to expenses which mislead the amount of stated capital. 6

16 Instead, EVA treatment capitalizes these items to the balance sheet and amortizes them for the appropriate period. Thus, EVA adjustments can eliminate the distortion of accounting convention introduced by standard GAAP accounting. Table 2.1 Accounting items under GAPP and adjustment in EVA method.2 Adjust Items Under GAPP EVA Method Marketing and R&D Expense Record as asset and amortize expense Deferred taxes Record as asset or liability Reverse recording of asset or liability to reflect cash basis reporting Purchased goodwill Record as asset amortize over up Reverse amortization to reflect original asset amount to 40 years Operating leases Expense Record asset and amortize; record liability and related interest Bad debts and warranty cost Estimate accrual Reverse accruals to reflect cash basis reporting LIFO inventory costing LIFO permitted Convert to FIFO Construction in progress Record as asset Remove from assets Discontinued operations Include in assets and earnings Remove from assets and earnings According to Stern Stewart, EVA integrates the framework of financial management and incentive compensation. For financial management, EVA provides a measure of performance consistent with the objective of maximizing shareholder wealth. Wealth is created when the company can cover all operating costs and the cost of capital. This makes EVA more associated with stock return and firm value than traditional accounting metrics. In addition, EVA is not only a performance measure but also a tool that link corporate objectives with the implementation process and strategy of each functions. 2 source: Biddle, G. C., Bowen, R. M., & Wallace, J. S., (1999). Evidence on EVA. Journal of Applied Corporate Finance, 12(2),

17 Regarding incentive compensation, EVA is tied to the incentive compensation plan that helps motivate managers and employees to work more efficiently because their incentive paid depends on the company EVA performance. Still, there are two concerns in implementing EVA. First, it can lead managers to under-invest since they may try to boost company s EVA by limiting investment in order to receive a higher EVA bonus. Second, a concern is related to the measurement of EVA at a divisional and a sub-divisional level. Conceptually, any unit with operating profits and assets can be evaluated performance by EVA. However, in practice, there is synergy among divisions. Thus, the issues of facility and asset sharing, cost allocation and transfer pricing become the obstacle in implementing EVA for individual division (Young & O'Byrne,2001,p.96-97). 2.2 The concept of Market Value Added Stern Stewart & Co. define Market Value Added via their website as follows. Market Value Added measures the difference between the market value of the firm (Debt and Equity) and the amount of Capital invested. Firms that trade at premiums to invested capital have positive MVA, while those trading below invested capital have negative MVA. 3 MVA = market value-total capital 4 MVA is a measure of wealth creation. If the total market value of a company is greater than the capital invested (positive MVA), the company has created shareholder wealth. Conversely, negative MVA means the company has destroyed shareholder wealth (Stewart, 1991). 3 source: 4 See Ehrbar, Al, EVA: The real key to creating wealth, p.44 8

18 MVA indicates the difference between cash in and cash out on what investors put in as capital and what they should get by selling at market price. Under financial objective of shareholder wealth creation, the goal of a company is to create high MVA as much as possible. However, there are some disadvantages associated with MVA application. First, MVA can change according to the fluctuation of stock market price in the short run. Therefore, MVA is not suitable as a guide for day to day decision making. Second, MVA can be calculated only for public companies which are traded on the stock market. Third, MVA can be calculated only at consolidated level. There is no MVA for a division, business unit, subsidiary or product line. Consequently, to assess the performance of each unit, managers should instead focus on internal performance measures that are closely linked to MVA (Ehrbar,1998,p.52). 2.3 Related Researches In consequence of the invention of the EVA trademark, Stern Stewart & Co. stated that EVA can be used instead of earning or cash flow from operations as a measure of both internal and external performance. Abandon earnings per share (Stewart, 1991, p. 2) Earnings, earnings per share, and earnings growth are misleading measures of corporate performance (Stewart, 1991, p. 66) The best practical periodic performance measure is economic value added (EVA) (Stewart, 1991, p. 66) Forget EPS, ROE and ROI. EVA is what drives stock prices (Stern Stewart advertisement in Harvard Business Review, November-December,1995,p.20). In order to support his claims that EVA is a better performance measures than the others. Stewart (1994) cited in-house research indicating that EVA stands well out from the crowd as the single best measure of wealth creation on a contemporaneous 9

19 basis and EVA is almost 50% better than its closest accounting based competitor in explaining changes in shareholder wealth (p.75) Thus, EVA gained great attention both in academic and practitioner publications. A large number of studies have investigated the relationship between accounting measures, EVA, MVA and stock returns. Lehn & Makhija (1993) concluded that EVA outperforms accounting measures ( ROA, ROE, ROS) in explaining stock returns for a study of 241 US companies over the years De Medeiros (2005) examined the relationship between EVA and stock return in Brazilian firms. The outcome shows that stock returns are influenced by past behavior of EVA. Maditinos, et al (2005) revealed that the results of a study of the Greek stock market indicates EPS outperforms and is associated with stock returns more than other performance measures ( ROI, ROE, EVA and SVA ). EVA alone is not the best performance measure but its explanatory power increases when it is combined in the EPS model. The evidence suggests that more determinants should be considered in firm valuation assessment. Ismail (2006) observed the explanatory power of EVA on stock returns in comparision to other accounting measures. The results for the study of 2,252 samples in the UK stock market over years states that net operating profit after tax and net income outperform EVA and residual income in explaining stock return. Dodd & Chen (1996) explored the correlation between stock return and residual income, ROA, EPS, ROE and EVA. The result indicates that ROA provides the best explanation on stock returns, followed by EVA, residual income, EPS and ROE. For Hartono (1999), a study of the relationship between ROA and EVA in the Indonesian stock market reveal that ROA outperform EVA in explaining stock return 10

20 Acheampong & Wetzstein (2001) compared traditional and value added measures in the food industry. The result reveals that value added measure are not significantly different from traditional measures but can be used along with traditional measures. West (1999) revealed the study results for 110 Australian companies over years that stock returns are more associated with earnings than net cash flow, residual income and EVA respectively. However, the analysis of EVA components, capital changes and GAAP related adjustments, indicates significant explanatory to the returns. Biddle, Bowen, & Wallace (1997) studied the relationship between stock returns and operating cash flow (OCF), earnings before extraordinary items(ebei) and residual income (RI). The result reveals that earnings are more associated with stock returns than EVA, residual income or cash flow from operation. Only some EVA components can offer incremental information to explain stock returns, when considered together, EVA is not related to stock returns. Chen & Dodd (1997) concluded that R 2 in operating income regression is higher than R 2 of residual income and EVA regression respectively in analyses of their relationship to stock returns. They suggested that companies should use a different measurement such as cycle times, market share growth, product quality etc. if they want to align organizational metrics with stock value. Fernandez (2002a) analyzed the relationship of EVA, MVA, NOPAT and WACC on 582 US companies. The result indicates that the correlation in the increase of NOPAT with MVA is greater than the increase of EVA with MVA. In addition, EVA shows negative correlation to MVA. The relationship is further analyzed between shareholder value and Economic Profit and EVA by examining a sample of 28 Spanish companies. Only 2 companies that had a high correlation between EVA and shareholder value creation. Therefore, the study concludes that accounting measures 11

21 and EVA are not the determinant of a firm s value creation. The increase in firm value is basically assessed by the changes in expectations of the firm s cash flow and the changes in firm s risk Turvey, Lake, Van Duren, & Sparling (2000) examined the relationship between EVA, ROA, ROE and stock market performance for 17 companies in the Canadian food industry. The result indicates no significant relationship between EVA and shareholder value. Taufik (2008) investigated accounting approaches ( ROE and ROA ) and EVA with the effect on stock returns of banks listed on the Jakarta Stock Exchange. The result shows that both approaches have a significant influence on stock returns, but in comparison, EVA is associated more with stock return than accounting approaches. Sparling&Turvey (2003) examined the relationship of EVA and shareholder return for 33 food companies and the entire sample are composed of companies in Canada. The outcome indicated that there is no significant relationship between the changes in EVA and shareholder returns. Kim (2006) observed the explanatory power of earning(nopat), cash flow and EVA with the market value of hospitality firms in the US. The findings indicate that earnings provide a better explanation of market value of stock. EVA itself has little explanatory power. Athanassakos (2007) investigated the performance of 39 Canadian companies which adopt value a based management method. The statistical analysis indicates that companies that used EVA had better stock price performance than those not using EVA. There is also a study about the accounting measures, EVA and MVA in the Thailand stock market. Jitvorrrapun (2003 ) tests the relationship between stock returns and EVA, MVA, ROA, ROE and ROS in 203 listed companies for the years

22 Results reveal that ROE can provide a better explanation for stock returns than EVA and MVA. Also, for industry analysis, the results are shown in the same way except three industries. EVA which is calculated from the energy and commercial sectors and MVA, calculated from property development sectors have significant correlation with stock returns which are higher than accounting measures. In addition, a study on portfolio return, EVA and MVA strategies generate lower returns than the value strategy that is commonly use by investors. Kuwongbundit (2006) explored the relationship between accounting earning, EVA and stock price in energy and communication industries. The result indicates that accounting earning ( EBITDA) are more associated with stock market price than EVA. However, when combining data of these two industries, EVA shows negative value while earnings still generate positive profit number to the company. 13

23 3. Methodology 3.1 Sample selection In the year 2010, there were a total of 474 listed companies in the Stock Exchange of Thailand. However, to compare historical data from year , 62 companies are excluded from the observation due to an incomplete dataset. Therefore, the total companies for the initial observation are 412 companies. All 412 listed companies can be divided into 8 major industries in 25 sectors which can be summarized into below table. Table 3.1 Number of listed companies in Thailand stock market year 2010 Industry Sector No. of Company Agro & Food Industry Agribusiness 15 Food and Beverage 24 Consumer Products Fashion 22 Home & Office Products 10 Personal Products & Pharmaceuticals 6 Financials Banking 10 Finance and Securities 31 Insurance 15 Industrials Automotive 16 Industrial Materials & Machinery 22 Packaging 13 Paper & Printing Materials 2 Petrochemicals & Chemicals 11 Property & Construction Construction Materials 26 Property Development 54 Resources Energy & Utilities 20 Mining 2 Services Commerce 13 Health Care Services 13 Media & Publishing 24 Professional Services 2 Tourism & Leisure 13 Transportation & Logistics 15 Technology Electronic Components 10 Information & Communication Technology 23 Total

24 However, the financial industry is not included in this study since this industry has a different financial structure and operating activities. Thus, the value of those performance measures in the financial industry has different result which can not be used in comparison to the other industries. After excluding the financial industry, 7 major industries with 356 listed companies remain for further analysis. Then, the sample size is determined by the population number of 356 companies. According to Krejcie & Morgan (1970), they create a summary table for the sample size of a given population (as shown in appendix in table 1). To be an efficient representation of 356 companies, the required sample size should be around 190. The sample companies are included according to sector and range of EPS values. The sample group should be representative of each industry and has a spread of EPS values close to the population. According to this selection criteria, the sectors which are selected for the sample group of this study are listed below Table 3.2 The number of sample group by industry Industry Sector Number of Company Agro & Food Industry Food and Beverage 24 Consumer Products Fashion 22 Industrials Automotive 16 Industrial Materials & Machinery 20 Property & Construction Construction Materials 26 Resources Energy & Utilities 20 Services Commerce 13 Health Care Services 13 Tourism & Leisure 13 Technology Information & Communication Technology 23 Total

25 Table 3.3 The range of EPS value in population and sample group EPS Range Number of Population Number of Sample Total Thus, the study examines the relationship between performance measures and stock returns in these 10 sectors with total sample size of 190 companies. 3.2 Variables definition and valuation All variables are calculated for 190 listed companies for the years The information for each component of variables is extracted from financial statement and stock price data published by SET. The definition and the formula of each variable can be described as below Accounting Measures 1. Earnings Per Share ( EPS ) Net income of a firm divided by the number of its outstanding shares held by the shareholders. EPS = Net income Common shares outstanding 2. Return of asset (ROA) Ratio measuring the operating profitability of a firm, expressed as a percentage of the operating assets. ROA indicates a firm's ability to efficiently allocate and manage its resources. 5 Definition from and 16

26 ROA = Net income Average total assets 3. Return on Equity ( ROE ) Ratio measuring shareholders' profitability, expressed as a percentage of the firm's net worth. ROE indicates a firm's efficiency in applying common-stockholders' money. ROE = Net income Average common stockholder 4. Return on Sales ( ROS ) Ratio measuring the operating performance of a firm, expressed as a percentage of sales revenue. ROS = Net income Sales The EPS, ROA and ROE information is published by SET except the ROS ratio which needs to be calculated separately Economic Value Added ( EVA ) After-tax profit that exceeds the required minimum return on capital. Computed by deducting the cost of capital (both debt and equity) from the after-tax profit EVA = NOPAT + WACC * Invested Capital 1. NOPAT = EBIT ( 1- tax) Where: NOPAT is net operating profit after tax, equal to operating income times (1 minus the tax rate) EBIT is earning before interest and tax Tax is the corporate tax rate of 30% 17

27 2. Invested Capital = Total asset Current liabilities Invested Capital is the capital that a company has invested or can invest in itself. 3. WACC (Weighted average cost of capital) A firm's weighted average cost of capital including debt, preferred stock, and common stock that are issued by the corporation to finance its operations and investments. The WACC equation is calculated according to the components Where: WACC = Kd( 1-t )D + Ke*E (D+E) (D+E) Kd is cost of debt, the rate that a company pay to obtain long tern debt capital which can be calculated by Interest / Long term Debt D is interest-bearing debt E is the book value of equity Ke is cost of equity, the total return that a firm s investors expect to earn in exchange for owning the asset and bearing the risk of ownership. Ke is calculated from the CAPM equation. CAPM : Ke = Rf + β( Rm-Rf ) Where: β = Cov( Ri,Rm) σ 2 Rm β is an indicator of the volatility of a stock in relation to a benchmark. It represents the ratio of the covariance between the return on stock price ( Ri) and the return on the market ( Rm) over the variance of the market return during the years Rm-Rf is market risk premium, the difference between the risk-free rate and expected returns on the market. Rf is Average return on one year government bond 18

28 Regarding market risk premiums, the SET market index is not the appropriated indicator for determining expected market return since the SET index had fluctuation trends during the research year. Thus, Rm is estimated by an earning-based approach, referring to the reverse of market P/E (Hitchner, 2006, p. 166). The expected market return from year can be calculated as shown in the table below. Table 3.3 Market Risk Premium from the years Market P/E Rm(%) Rf (%) Market risk premium (%) Market Value Added ( MVA ) The difference between the market value of a company and the capital contributed by investors for both bondholders and shareholders, calculated by the following formula. MVA = Market Value Invested Capital Where: market value is calculated from outstanding share* closing price Stock return The stock return is measured by dividend yield and the change of stock price over a year as if investors hold stock for one year period. The closing price applied in this model is the closing price 60 days after the end of year. Since SET allows listed companies to submit their financial statement 60 days after the year end and as checked, all sample companies submit financial statements for SET publish by the end of February. 6 Market P/E is published by SET, and Rf is from the report of the Thai Bond Market Association, 19

29 To reflect an effect from the disclosure of financial data, stock price at t+1 is taken from the closing price at the end of February and subtracted from the previous year s closing stock price at the end of February. Regarding dividend, it is an annual paid amount for the fiscal year of financial statement. Stock return = Log [( Stock price t+1 + dividend] Log ( stock price t ) 3.3 Method The statistical tests include descriptive, correlation and regression analysis. However, to define the explanatory power of performance measures on the stock return, the result of relative information content from regression model is used to assess on which performance measure is more associated with stock return. In addition, the study further explores to the portfolio returns from two typical strategies: growth and value investing, comparing to MVA and EVA strategies. The methodology is described in detail as follows Regression analysis The regression model is employed to test the relationship between performance measures: accounting measures, EVA, MVA and stock returns. The test is done by pair wise comparison regression of the accounting measures: EPS, ROA, ROE, ROS and value based measures: EVA and MVA. The explanatory power of each measure is determined by the value of R 2 since R 2 is defined as the proportion of variance in the dependent variable which can be explained by the independent variable. 20

30 The regression model can be described as follows. Return i,t = α 0t + β 1t X i,t + ε it Where α 0t is the constant value of each regression equation β 1t is the coefficient of X ( the slope of regression line) X it is independent variable of firm i at time t ε it is the error in predicting dependent value The equation of each performance measure which tests with stock return is shown as below. Equation (1) Return i,t = α 0t + β 1t EPS i,t + ε it (2) Return i,t = α 0t + β 1t ROA i,t + ε it (3) Return i,t = α 0t + β 1t ROE i,t + ε it (4) Return i,t = α 0t + β 1t ROS i,t + ε it (5) Return i,t = α 0t + β 1t (EVA i,t / Invested Capital i,t) + ε it (6) Return i,t = α 0t + β 1t (MVA i,t / Invested Capital i,t ) + ε it In this study, EVA and MVA are used for the analysis are adjusted by invested capital in order to reduce the effect of the difference in firm size (Grant, 1996). These regression models are applied for the analysis in two ways. First, it tests the relationship between performance measures and stock return for a total sample group of 190 companies and then examines the relationship further to in each sector. Second, those regression models are used to test the relationships of grouped companies classified by determinants that can affect the stock return. The selected determinants include beta, market capitalization and Book to Market ratio (BTM) According to Nittayagasetwat, Aumeboonsuk, & Nittayagasetwat (2005), they conclude that these determinants can affect the stock return in the study of the cross section of expected stock returns in the stock market of Thailand. 21

31 Beta Beta refers to systematic risk of stock in comparison to the whole market at a period of time. The company with high beta has more risk than the company with low beta. Beta(risk) can imply return. Stocks with a high beta should generate a higher return than the market. Market Capitalization The market capitalization of company is measured by the stock market price multiplied by share outstanding. A higher market capitalization company means for a larger size of company. The lower one is the smaller size of company. The company with small market capitalization is considered riskier than the large cap company. However, small cap company does not always generate higher return than large cap company even it has more risk 7. Book to Market Ratio Book to Market ratio can be calculated by Book to Market = Book value of firm Market value of firm Book value of equity represents the cost the company incurred in the past to generate value. Market value represents investors' collective estimate of the value that the company will generate in the future. A low book-to-market ratio means that investors believe the company will generate more value in the future compared to the past. A high book-to-market ratio indicates a belief that future growth will not be as great 7. This can be interpreted that a company with low book to market ratio has more growth expectation than a company with high book to market ration 7 Source: 22

32 Consequently, all 190 companies are divided into two groups by referring to the median value of each determinant since the median is the middle number to separate the higher and lower half of data. Then, the test is done with two categories of companies that generate high and low value of beta, market capitalization and BTM. This is to examine the characteristics of companies for which MVA and EVA are associated with stock returns Portfolio returns This study explores further the portfolio performance computed by two typical strategies like growth and value investing in comparing to EVA and MVA strategies. The return of SET 50 Index is used as the benchmark for portfolio returns and the portfolio return of each strategy is measured by Sharpe ratio. The formula can be described as follows (Strong, 2009, p. 518). Sharpe measure = R-R f σ Where: R = average return R f σ = risk free rate = standard deviation of returns The portfolios are constructed from all 190 companies by referring to their value of earning growth, P/B, MVA and EVA. The detail of portfolio construction for each strategy is defined as below. Growth Strategy Growth investing is the strategy where an investor attempts to buy the stock that have high expected growth rates even if stock price is relatively high in the market. A growth stock is on of a company whose sales, earnings and share of the 23

33 market are expanding faster than the general economy and faster than the average for the industry (Cohen, Zinbarg, & Zeikel, 1987, p. 21). Thus, in this study, the growth of earnings is used to define the growth stock. All sample companies are divided equally into 4 quartiles by their growth of earnings. The highest rank of earning growth companies is included in quartile 1 and the lowest rank of earning growth companies is in the quartile 4. Value Strategy Value investing is opposed to growth investing style. Investor chooses a stock that is undervalue in relation to its underlying fundamentals. Stock with low price relative to book value are considered to be value stocks. Thus, value portfolio is formed by the rank of P/B value. The stock with lowest P/B value is grouped in quartile 1, then quartile 2, 3 and the highest group for quartile 4 respectively. EVA Strategy This portfolio is formed by the rank of EVA/outstanding shares categorized according to 4 quartiles. The highest EVA companies are grouped in quartile 1 and the lower ones are in quartile 2, 3 and 4 respectively. MVA Strategy The portfolio is formed by the rank of MVA/outstanding shares categorized into 4 quartiles. The highest MVA companies are grouped in the quartile 1 and the lower ones are in quartile 2, 3 and 4 respectively. 24

34 4 Empirical Study This chapter focuses on the empirical results of statistical tests and portfolio returns. Regarding statistical tests, the analyses below are applied for all sample companies and separated by industry. 1. Descriptive analysis 2. Correlation analysis 3. Regression analysis The results of regression analysis are used for determining the explanatory power of accounting measures, EVA, MVA over the stock returns which is done in two ways. First, the test is done without any control variable and second, the sample group is divided into 2 categories by the determinant of beta, size, book-to-market ratio. The final part covers the detail of portfolio return in comparison between typical two investing styles, MVA and EVA strategies. 4.1 Descriptive Analysis The frequency distribution of minimum, maximum, average and standard deviation values for each variable is summarized in the tables below by 10 industries and overall sample companies. 25

35 Table 4.1 Descriptive statistics of variables by industry and all sample companies from the years Industry EPS ROA (%) ROE (%) ROS(%) MVA/Invested EVA/Invested Return (%) Capital Ratio Capital Ratio Automotive (n=80) Min (23.59) (27.60) (65.28) (172.97) (0.66) (159.04) Max , Mean STD Commerce (n=65) Min (4.57) (24.52) (73.31) (79.10) (0.39) (185.24) Max , Mean , STD , Construction (n=130) Min (7.14) (28.06) (100.44) (178.80) (1.47) (228.28) Max , Mean (4.18) STD , Energy (n=100) Min (1.01) (22.54) (64.29) (96.95) (0.37) (164.18) Max , , Mean , STD

36 Table 4.1 Descriptive statistics of variables by industry and all sample companies from the years ( cont.) Industry EPS ROA (%) ROE (%) ROS(%) MVA/Invested EVA/Invested Return (%) Capital Ratio Capital Ratio Fashion (n=110) Min (4.53) (16.62) (86.24) (45.32) (0.33) (123.29) Max , Mean STD Food (n=120) Min (4.21) (7.66) (94.79) (7.96) (1,108.08) (0.24) (209.97) Max , Mean , STD , Health (n=65) Min (1.29) (5.98) (7.06) (6.02) (0.05) (64.36) Max , Mean , STD Industrial (n=100) Min (4.90) (37.37) (189.41) (64.11) (319.09) (1.35) (279.51) Max , Mean (0.22) (2.96) STD

37 Table 4.1 Descriptive statistics of variables by industry and all sample companies from the years ( cont.) Industry EPS ROA (%) ROE (%) ROS(%) MVA/Invested EVA/Invested Return (%) Capital Ratio Capital Ratio Information (n=115) Min (3.77) (80.95) (4,642.17) (332.15) (28.23) (284.49) Max , Mean (38.46) (0.12) 1, (0.13) (10.50) STD , Tourism (n=65) Min (2.48) (30.67) (126.89) (126.69) (4,958.07) (0.56) (171.21) Max , Mean (11.56) STD , Total (n=950) Min (23.59) (80.95) (4,642.17) (332.15) (4,958.07) (28.23) (284.49) Max , , Mean , STD ,

38 The descriptive analysis is done for all 190 companies and also according to the 10 industries. From the table, the results for total sample companies in part of accounting measures indicates that EPS and ROA have a lower value of standard deviation than ROS and ROE. This means EPS and ROA for overall companies are not spread out too much and the values are close to their mean values. When comparing to MVA and EVA, MVA shows the highest standard deviation at 1, whereas EVA is at 1.00 which is the lowest one among all performance measures. All performance measures and returns show a negative number at the minimum value and the average return for all sample companies is at 1.42%. Looking at accounting measures for each industry, EPS and ROA still present low standard deviations except in food and health care services which indicates that the ROS has a lower standard deviation than EPS. However, for all industries, ROA is the only variable that maintains the lowest standard deviation. In part of MVA and EVA, MVA still shows the highest standard deviation and the lowest among all performance measures belongs to EVA. In addition, all industries present negative values of EPS, ROA, ROE, ROS, EVA and also for the returns. For MVA, only three industries including food, industrial and tourism that have a negative MVA. Considering the mean values of each variable, the energy industry shows the highest average EPS and ROS value whereas industrials has the lowest average EPS and ROS. The health industry appears to have the highest mean value for ROA, ROE and return. Their lowest mean values belong to the tourism industry for ROA and the return and information industry for ROE. Regarding MVA, the fashion industry presents the highest average value and the lowest belongs to the commerce industry. For EVA, the highest mean value is in the tourism industry and the information industry has the lowest one. 29

39 After considering the data description, the extreme outlier observations are deleted to transform the population data into a normal distribution. Then, the normality of data is assessed before further analysis in the next statistical test. 4.2 Correlation Study This study employs correlation statistical method to measure the strength of relationship between two variables. The correlation coefficient can indicate their directions and strength of relationship. The correlation is applied to all and for each industry. The correlation matrix below illustrates whether each variable has a relationship with stock returns or not at a significant level of 0.01 and Table 4.2 Correlation statistic matrix for all sample companies in the years The correlation result for 190 sample companies shows that EPS, ROA, MVA and EVA have a relationship with stock returns in the same direction. Also, for MVA which 30

40 is computed based on the stock market price indicate significant correlation to those variables. However, EVA has a negative relationship with MVA although it shows positive direction with the return. At the industry level, the relationship for each variable and stock return can be summarized per the table below. Table 4.3 The relationship of variable with stock return by industry years Industry EPS ROA ROE ROS MVA EVA Automotive Commerce Construction Materials Energy & Utilities Fashion Food and Beverage Health Care Services Industrial Materials & Machinery Information & Communication Technology Tourism & Leisure All Companies The relationship between accounting measures are found in most industries except commerce and health care. Also MVA is correlated with stock return except in health care and information industry. For EVA, it is associated with stock returns in three industries which are industrial, information and tourism. According to correlation result, the regression analysis is applied in the next section observe the explanatory power of each variable on stock return for all 190 companies and also for grouping companies classified by beta, company size and book to market ratio. 4.3 Regression Analysis Simple regression analysis is used to define the relationship between a pair wise of independent and dependent variables. A pair wise comparisons technique is used. If the p-value is less than or equal to 0.05, the independent variables (MVA, EVA, EPS, ROE, ROA and ROS ) are said to affect the dependent variable, which is stock return. 31