Global Economics and Management Review 20 (2015) 6–14

www.elsevierciencia.com/gemrev

Original

The impact of performance measures, leverage and efficiency on

market value added: Evidence from Iran
Mohammad Alipour a,∗ , Mohammad Ebrahim Pejman b
a
Young Researchers and Elite Club, Khalkhal Branch, Islamic Azad University, Khalkhal, Iran
b
Department of Accounting and Finance, Bojnourd Branch, Islamic Azad University, Bojnourd, Iran

a r t i c l e i n f o a b s t r a c t

Article history: The purpose of this paper is to investigate economic value added (EVA), as a performance measurement
Received 4 July 2014 model, as compared to six traditional accounting performance measures vis-à-vis the market value of
Accepted 16 April 2015 firms listed on Tehran Stock Exchange (TSE). The paper also explores the effect of the degree of operating
Available online 27 June 2015
leverage, financial leverage, and efficiency on market value added.
The paper uses a sample of 450 firm-year observations from the Iranian market and applies pooled
JEL classification: ordinary least square and panel data regression. The results indicate that EVA has no superiority over
G30
other performance measures, and that return on sales and return on assets are more powerful than EVA
L25
C23
in explaining firm market value. Due to EVA’s lack of correlation with market value, investors cannot use
it as an internal value creation measure along with the traditional performance measures. This paper is
Keywords: one of the first studies on the relevance of traditional accounting and value-based performance measures
Market value added in explaining TSE market values. The results extend EVA’s role in explaining market values, and address
Economic value added its effect on investors’ decisions in a continental Asian market with characteristics similar to that of Iran.
Traditional measures of performance © 2014 INDEG/PROJECTOS- Inst. para o Desenvolvimento da Gestão Empresarial/Projectos. Published
Efficiency
by Elsevier España, S.L.U. All rights reserved.
Degree of operating and financial leverage
Emerging market

Introduction superior to other measures, and many researchers have supported

EVA as well (Bao & Bao, 1998; Grant, 2003; Lee & Kim, 2009; Lehn &
In today’s competitive world, a main business goal is to create Makhija, 1996; Milunovich & Tsuei, 1996; O’Byrne, 1996; Uyemura,
value and wealth for shareholders. Businesses that focus on share- Kantor, & Pettit, 1996; Worthington & West, 2004). On the other
holder value are much healthier and improve the overall economy hand, some researchers have rejected Stewart’s claim, noting a
(Kim, 2004). As a result, companies need reliable and accurate activ- weak relationship at best between EVA and a firm’s market value or
ity measures. stock returns (Biddle, Bowen, & Wallace, 1997; Chen & Dodd, 1997,
Over the last few years, traditional performance measures have 2001; De Villiers & Auret, 1997; De wet, 2005; Ismail, 2006; Kramer
been criticized due to their inability to incorporate the full cost & Pushner, 1997; Kyriazis & Anastassis, 2007, Kumar & Sharma,
of capital, thus impugning accounting income as a consistent firm 2011a, 2011b; Sparling & Turvey, 2003; Turvey, Lake, Van Duren, &
value predictor (Ehrbar, 1998; Rappaport, 1986; Stewart, 1991). Sparing, 2000; Worthington & West, 2001).
Furthermore, they do not provide strategic value management This study compares EVA to traditional firm performance meas-
guidance (Kim, 2004). In recent years, new measures have been ures, that is, return on assets (ROA), return on equity (ROE), return
introduced, including economic value added (EVA), market value on capital employed (ROCE), return on sales (ROS), earnings per
added (MVA), cash value added (CAV), cash flow return on invest- share (EPS) and dividend per share (DPS). Another purpose is to
ment (CFROI), discounted economic profit (DEP), and shareholder test whether the efficiency and degree of operating and financial
value added (SVA). EVA was first introduced by Stern, Stewart and leverage have any correlation with TSE-listed firms’ MVA. To our
Coin (1991), as a measure that calculated the value added after knowledge, this study is the first to address value-based and tradi-
deducting the cost of capital. Stewart (1991) claimed that EVA is tional firm performance measures in Iran and their effect on MVA.
The study also tests and correlates these variables with MVA as
an external measure of performance so as to find whether EVA
∗ Corresponding author. has any superiority over traditional performance accounting meas-
E-mail addresses: uniclass.alipoor@gmail.com, uniclass.alipoor@aukh.ac.ir ures. Moreover, the relationship of efficiency, degree of operating
(M. Alipour). leverage, and degree of financial leverage with firms’ market value

http://dx.doi.org/10.1016/j.gemrev.2015.04.001
2340-1540/© 2014 INDEG/PROJECTOS- Inst. para o Desenvolvimento da Gestão Empresarial/Projectos. Published by Elsevier España, S.L.U. All rights reserved.
 M. Alipour, M.E. Pejman / Global Economics and Management Review 20 (2015) 6–14 7

is examined. The results of this study could help managers deter- has a significant effect on MVA. Lehn and Makhija (1996) studied
mine which variables contribute most to MVA. A better knowledge 241 large US companies during the period 1987–1993 and found
of the drivers of MVA should also allow investors to improve their that EVA and MVA are positively related to stock returns, and that
strategic and tactical asset allocations. EVA has a slight edge over the traditional performance measures.
The study explores the value relevance of both traditional Uyemura et al. (1996) studied the relationship between MVA and
accounting and value-based performance measures to emerging five performance measures—that is, EVA, net income, ROA, EPS, and
markets. In other words, we would like to test whether the lack of ROE—in a sample of 100 US banks during the period 1986–1995 and
conclusive evidence regarding EVA is confined to mature markets found that EVA has the highest MVA correlation.
or applies also to large, emerging Asian markets such as the Tehran Athanassakos (2007) selected 300 Canadian companies to see
Stock Exchange (TSE). whether they used value-based management methods and investi-
Therefore, we believe it is important to conduct a new study gated the stock price of those companies embracing such methods.
that, for the first time, uses Iranian data and applies more reli- The results suggested that companies that use EVA had a better
able regression techniques for a sample of cross-sectional time stock price than those not using it. Fernandez (2003) analyzed
series data. We test an Iranian sample of 450 firm-year observa- 582 US companies using EVA, MVA, net operating profit after tax
tions for the period 2003–2008, and report evidence that supports (NOPAT), and weighted cost of capital (WACC), and calculated the
Chen and Dodd (1997), Biddle, Bowen, and Wallace (1998), and 10-year correlations between the yearly MVA increase and yearly
Kumar and Sharma (2011a, 2011b), that is, earnings are more asso- EVA, NOPAT, and WACC. Fernandez found that the correlation
ciated with stock return and market value than EVA. Further, we between the increase in yearly MVA and NOPAT was greater than
show that ROS and ROA are more powerful than EVA in explain- its correlation with EVA. De wet and Hall (2004) studied the rela-
ing firm market value. The remainder of the paper is organized as tionship between EVA, MVA, and leverage, and found that debt
follows: Section two discusses the background literature. Section increases lead to increased financial costs, but, to the same extent,
three discusses the research methodology, sample, and variables. reduced capital costs. Thus, financial leverage does not affect MVA.
Section four presents the results and discusses the findings of the Zaima, Turetsky, and Cochran (2005) studied the relationship
study. Finally, section five summarizes the findings and concludes between EVA and MVA while controlling for the economic effect
the discussion. of the market. Results showed that EVA and gross domestic prod-
uct significantly affect MVA vis-à-vis managerial decisions made
after controlling for the systematic economic effect. Kyriazis and
Prior literature and hypothesis development Anastassis (2007) examined the Athens Stock Exchange’s EVA
information content and reported that although it is useful as a
EVA is one of the first value-based management measures and performance evaluator, there is no significant relationship with
has been the subject of many studies. According to Stern, Stewart, MVA, and it has no superiority to other measures. Lee and Kim
and Chew (1996), EVA’s value stems from its ability to be the (2009) selected 353 companies between 1985 and 2004 and com-
main part of an integrated financial management system, lead- pared three new performance measures—that is, refined EVA, EVA,
ing to decentralized decision making. EVA measures the difference and MVA—to three traditional ones, finding that MVA and refined
between the return on a firm’s capital and the cost of that capital EVA are more effective for evaluating hospitality industry company
(Epstein & Young, 1998). performance.
The value of EVA has long been debated. Stewart (1991) argued De wet (2005) compared EVA and traditional accounting meas-
that EVA can be a good predictor of future business performance, ures as drivers of shareholder value in 89 South African companies
while Anderson and Bey (1998) concluded that EVA varies over during the period 1994–2004 and found stronger correlation
time and it is significantly correlated with accounting variables. between MVA and operating cash flow (OCF), but very little cor-
However, Appleby (1997) argued that EVA is a lagging indicator. relation between MVA and EPS and MVA and DPS. Kramer and
Nonetheless, EVA can be used for different internal management Pushner (1997) studied the relationship between EVA and MVA in
activities such as CEO performance (Coles, McWilliams, & Sen, 1000 companies during the period 1982–1992 and found no clear
2001), optimal resource usage (Rompho, 2009; Zimmerman, 1997), evidence to support the contention that EVA is the best internal cor-
and strategic goals (Bahri, St-Pierre, & Sakka, 2011). porate success measure. Misra and Kanwal (2007) studied Indian
MVA is the present value of all future EVA; to create market companies between 2002 and 2006 and argued that traditional
value, firms need to achieve positive EVA (Zafiris & Bayldon, 1999), accounting measures cannot predict corporate performance and
which represents the amount of wealth created for shareholders that EVA is significantly associated with MVA. Biddle et al. (1998)
(a negative MVA represents the amount of capital wasted by man- studied 6174 companies during the period 1984–1993 and rejected
agement) (Kim, 2004). Successful firms add to their MVA and thus Stewart’s claim of EVA superiority, arguing that net income before
increase the value of invested capital. From an investor’s viewpoint, extraordinary items is more effective in explaining stock returns
MVA is the best external measure of a firm’s performance. Also, and firm value. Finegan (1991) studied 450 medium-sized US enter-
MVA can represent the net present value of a company’s current prises and concluded that EVA has greater explanatory power than
and contemplated capital investment projects. Moreover, MVA is such measures as capital growth, ROC, EPS, and cash flow. Using
closely associated with the ability of a business to create value in 1000 firms between 1983 and 1994, Biddle et al. (1997) studied
the future. whether EVA’s relative information content exceeds that of OCF or
Stewart’s 1991 evaluation of 613 American companies in terms earnings, and whether EVA offers any incremental valuation advan-
of average EVA during the period 1987–1988 indicated that com- tage. Biddle et al. found that earnings provide a superior metric of
panies with a positive EVA correlate very highly with MVA, both for relative information, and that EVA provides no superiority in its
the changes in values and the average values used, while in com- association with stock returns or firm value.
panies with a negative EVA, the correlation was quite low. Stern, Grant (2003) confirmed the relationship between EVA and cor-
Stewart, and Chew (1994) argued that accounting measures such porate valuation using a sample of 983 US companies between 1985
as earnings, earnings growth, dividends, dividend growth, return and 1993, with results that supported Stern et al.’s claims. Ismail,
on equity (ROE), and even cash flow are not the key operating in a 2006 study of UK companies, tested EVA’s relative and incre-
measures of firm performance, but EVA is one such measure that is mental information content and other performance measures using
closely linked to a firm’s market value. They also showed that EVA panel data regression. The results failed to support Stern et al.’s
8 M. Alipour, M.E. Pejman / Global Economics and Management Review 20 (2015) 6–14

hypothesis as net operating income after taxes and net income was available. As of the end of 2003, there were 328 firms listed on
outperformed EVA and residual income. TSE, which is one of the oldest and biggest capital markets in the
Kim (2006) provided empirical evidence on the relative and Middle East. Because of the specific nature of their activities, firms
incremental information content of EVA and traditional perfor- related to banking and financial institutions were excluded. More-
mance measures, earnings, and cash flow in the US hospitality over, companies with long periods without transactions (more
industry. EVA’s information content and other explanatory vari- than two months) were excluded. The final sample consisted of 75
ables indicated that earnings are more useful than cash flow firms, and a balanced panel set of 450 firm-year observations was
in explaining hospitality firms’ market value. Kramer and Peters obtained.
(2001) examined EVA’s ability to serve as a proxy for MVA across 53
US industries and found that the, “marginal costs of using economic
Measures
value added as a proxy for market value added are not justified by
any marginal benefits”. Riahi-Belkaoui (1993) examined the rela-
Independent variables
tive and incremental content of value added, earnings, and cash
The theoretical model of this research considers EVA, EPS, DPS,
flows in the US context. The results indicated that the information
ROA, ROE, ROCE, ROS, SOA, SOCE, DOL and DFL as independent
content of value added is a major market return determinant, pro-
variables. These variables are expected to be related to the depend-
viding incremental information content beyond both net income
ent variable. The method of the present research is similar to that
and cash flow.
of Kim (2006), Irala (2007), Stewart (1991), Milunovich and Tsuei
Kumar and Sharma (2011a) compared EVA with traditional per-
(1996), O’Byrne (1996), and Uyemura et al. (1996). EVA attempts to
formance measures in 873 non-financial Indian companies during
capture the true economic profit of a company (Maditinos, Sevic,
the period 2000–2008 and found that NOPAT and OCF are more
& Theriou, 2009), and can be calculated as follows where EBIT is
effective than EVA for evaluating firm performance; regarding mar-
earnings before interest and tax:
ket value, EVA contributes less than NOPAT, OCF, EPS, and RONW.
Kumar and Sharma (2011b) studied 608 Indian companies and
EVA = EBIT − (Cost of Capital × Capital Employed)
found that traditional measures do not have a significant effect on
companies’ MVA and that EPS is negatively related to market value. In order to calculate EVA, we made certain adjustments as
Moreover, the results of this research indicated that NOPAT and OCF proposed by Stern et al., such as non-cash expenses due to account-
are superior to EVA in explaining company market value. ing records (except depreciation). To remove the effect of such
Considering the literature, the following hypotheses are formu- expenses, several reserves such as allowance for bad debts and
lated: inventory obsolescence, investment and accrued expense reserves,
employees’ termination benefits, and deferred tax are added to cap-
H1: EVA is superior to the traditional measures of performance ital employed; any increase in these reserves is added to profit after
in determining the value and market value of Iranian firms. tax. What is meant by capital employed in the above relation is the
• H1a: There is a positive relationship between EVA and MVA. sum of interest-bearing debts and equity, which is modified by the
• H1b: There is a positive relationship between earnings per share following adjustments:
and MVA.
• H1c: There is a positive relationship between dividend per share • Adding accounting reserves such as allowance for doubtful
and MVA. accounts
• H1d: There is a positive relationship between return on assets and • Adding research and development and marketing costs
MVA. • Adding abnormal accumulation of losses minus abnormal earn-
• H1e: There is a positive relationship between return on equity and
ings after tax
MVA.
• H1f: There is a positive relationship between return on capital
employed and MVA. Earnings per share (EPS) are calculated by dividing net income
• H1g: There is a positive relationship between return on sales and by the average number of common shares outstanding. Dividend
MVA. per share (DPS) is calculated by dividing dividends by average num-
H2: There is a positive relationship between efficiency and MVA. ber of common shares outstanding. Return on assets (ROA) is a
• H2a: There is a positive relationship between sales on fixed assets profitability ratio and is calculated by dividing net income by aver-
and MVA. age total assets. Return on equity (ROE) is defined as the ratio of net
• H2b: There is a positive relationship between sales to capital income over the book value of equity. Return on capital employed
employed and MVA. (ROCE) is equal to earnings before interest and tax divided by
H3: There is a negative relationship between leverage and MVA. average capital employed. Return on sales (ROS), is determined
• H3a: There is a negative relationship between degree of operating by scaling operating (net) profits by total sales recorded over the
leverage and MVA. same period. Sales on fixed assets (SOA) are calculated by dividing
• H3b: There is a negative relationship between degree of financial sale by the total fixed assets, and sales on capital employed (SOCE)
leverage and MVA. are equal to sales divided by average capital employed. Finally, the
H4: EVA, along with the traditional accounting performance impact of different levels of operating and financial leverage on
measures, has information content. MVA is evaluated. Degree of operating leverage (DOL) calculated
by dividing change in operating income by change in sales. A busi-
ness would benefit if it can estimate the DOL as the impact of the
Methods leverage on the percentage of sales can be quite striking if not taken
seriously. Degree of financial leverage (DFL) is calculated in a ratio:
Sample and data percent change in net profit before tax divided by percent change
in sales. This ratio also helps in determining the suitable leverage
The sample company financial data was collected using soft- for achieving business goals. The higher leverage of the company,
ware provided by the Tehran Stock Exchange (TSE). This study uses the more risk it has, with businesses subsequently needing to treat
all TSE-listed firms during the period of 2003–2008 for which data it as similar to having a debt.
 M. Alipour, M.E. Pejman / Global Economics and Management Review 20 (2015) 6–14 9

Table 1
Definition of the research variables.

Variables Definition Adapted from

Dependent variable
Market value added (MVA) Market capitalization – book value of equity Nappi-Choulet, Missonier-Piera, and Cancel (2009); Kumar
and Sharma (2011a, 2011b); Lee and Kim (2009); De wet
(2005)

Independent variables
Economic value added (EVA) Earnings before interest and tax − (cost of Chen and Dodd (1997), Biddle et al. (1998), Kumar and
capitala × capital employed) Sharma (2011a, 2011b); De wet (2005)
Earnings per share (EPS) Net income/average common stock De wet (2005); Kumar and Sharma (2011a, 2011b);
Finegan (1991); Milunovich and Tsuei (1996)
Dividend per share (DPS) Dividend/average common stock De wet (2005)
Return on assets (ROA) Net income/total average assets Lehn and Makhija (1996); Turvey et al. (2000); De wet
(2005)
Return on equity (ROE) Net income/stockholder equity Lehn and Makhija (1996); Turvey et al. (2000); De wet
(2005)
Return on capital employed (ROCE) Profit before interest and tax/average capital employed Worthington and West (2004); Zafiris and Bayldon (1999)
Return on sales (ROS) Net income/sales Lehn and Makhija (1996); Turvey et al. (2000)
Sale on fixed assets (SOA) Sales/total average fixed assets Herrmann, Inoue, and Thomas (2003)
Sales on capital employed (SOCE) Sales/average capital employed Firth (1979)
Degree of operating leverage (DOL) % change in operating income/% change in sales Chiou and Su (2007); De wet and Hall (2004)
Degree of financial leverage (DFL) % change in net profit before tax/% change in sales Chiou and Su (2007); De wet and Hall (2004)
a
Cost of capital = ((earning per share/share price) × (equity at the beginning of the period/capital at the beginning of the period)) + ((interest bearing debts/capital at the
beginning of the period) × (financing costs/average interest bearing debts).

Dependent variable Hausman tests are presented. If the Lagrange test gives a significant
MVA, popularized by Stewart (1991), is used as the dependent result, then the panel results are preferred over the pooled results.
variable and represents the value-added created for the share- If the Hausman test gives a significant result, then the fixed effect
holder investments supplied during a certain period. In other results are preferred to the random effects. As pointed out by Hsiao
words, MVA is the difference between a company’s market value, (1986), simple least squares estimation of pooled cross-section and
as reflected primarily in its share price, and the book value of the time series data may be seriously biased.
equity. Table 1 summarizes variable definitions and descriptions.
Results and discussion
The methodology and models
Descriptive statistics and correlation
This paper employs both pooled and panel data techniques
to estimate the regression models. To assess the variables’ rela- Table 2 independent and dependent variable descriptive statis-
tive information content, the coefficient of determination (R2 ) of tics show that all the research variables have positive mean and
various performance measures is examined and analyzed. Com- median. In addition, the average EVA of these companies is IRR
parison of the R2 of various performance measures will provide a 54,867.79, and their average MVA is IRR 718,968.30, indicating that,
direct test of Stern et al.’s claim about EVA’s performance measure overall, these companies are able to earn higher returns than cost
superiority. One interesting finding related to the literature is that of capital. On average, these companies earned 18.11 percent profit
most researchers have used R2 and panel data regression model to by employing their assets and the earned profit in return for equity
measure value relevance (Garvey & Milbourn, 2000; Ismail, 2006; and capital employed is 63.53 percent and 165 percent respec-
Kramer & Pushner, 1997; Sparling & Turvey, 2003). Our method- tively, which are considerable values. Moreover, these companies
ology is based on the similar work of Kim (2006), and Kumar and have earned 28.44 percent profit from their sales. According to this
Sharma (2011a, 2011b). table, these companies have had IRR 1682.10 EPS during six years
Panel data models are powerful research instruments that and have distributed IRR 1154.30 of dividends. Furthermore, MVA
account for the effects of cross-sectional data. This in turn may has a higher mean and median than EVA. Moreover, the sales of
help us estimate the appropriate empirical model. We use general these companies have been 7.22 times their capital employed and
models for panel data that enable an empirical estimate of the rela- they have earned 83 percent profit by employing their assets, sug-
tionship between independent variables and MVA. We formulate gesting the high efficiency of these companies’ managers. Finally,
behavioral differences between the various cross-section elements the degree of financial and operating leverage of the companies are
as follows: 2.39 and 2.58, respectively.
Moreover, a test of normality used in this research is
Yit = ˛ + ˇXit + εit (pooled model) (1)
Jarque–Bera, which examines whether data skewness and kurto-
Yit = ˛i + ˇXit + εit (fixed effects model) (2) sis match a normal distribution. The results of this test at the 0.05
significance level show that all the data is normally distributed
Yit = ˛i + ˇXit + (εit + i ) (random effects model) (3)
(Table 2). The results of correlation analysis presented in Table 3
where Yit is the MVA of firm i in year t, ˛i is the intercept coefficient indicate that there is no multicollinearity between independent
of firm i, ˇ is a row vector of slope coefficients of the regressors, and variables.
Xit is a column vector of firm-specific variables for firm i in year t,
which represents the explanatory variables reported in Table 1, εit Results of regression analysis
is the error term. According to Green (2003), the panel data sug-
gests the use of fixed or random effects that control for unobserved Table 4 shows the results for every independent variable’s
firm and/or year effects. In order to distinguish the preferable set relative information content tests. The assessment is made by con-
of results statistically, the results of the Lagrange multiplier and ducting four separate regressions for each performance measure
10 M. Alipour, M.E. Pejman / Global Economics and Management Review 20 (2015) 6–14

Table 2
Descriptive statistics.

Variables Obs. Mean Median Minimum Maximum Std. dev. Kurtosis JB statistics

EVA 450 54,876.79 7500.267 −1,559,183 5,341,148. 392,750.7 119.3191 164,118.8

EPS 450 1682.137 1372.000 21.00000 9445.000 1401.675 11.73871 1195.329
DPS 450 1154.316 958.0000 6.000000 6088.000 896.6407 7.435686 373.3801
ROA 450 0.181150 0.161439 −0.048808 0.537375 0.110010 2.937176 25.36832
ROE 450 0.635359 0.577245 −0.425244 2.893375 0.437294 6.906811 289.5248
ROCE 450 1.658579 1.355005 0.014158 9.415000 1.378990 10.32478 880.8615
ROS 450 0.284442 0.247728 0.001213 2.163353 0.195520 32.64937 11,122.11
SOA 450 0.830590 0.770902 0.021677 2.904223 0.363070 7.010449 279.9068
SOCE 450 7.220198 5.104050 0.115319 170.6526 11.24836 158.4964 293050.0
DOL 450 2.398147 0.872043 −179.8356 497.4044 31.77217 212.0458 526592.6
DFL 450 2.537872 0.910335 −26.19032 204.8651 15.12364 119.2053 164747.3
MVA 450 718968.3 152079.5 −147,889.0 11,484,434 1,593,588. 22.63292 5392.971

Note: EVA (economic value added), EPS (earnings per share), DPS (dividend per share), ROA (return on assets), ROE (return on equity), ROCE (return on capital employed),
ROS (return on sale), SOA (sales on fixed assets), SOCE (sales on capital employed), DOL (degree of operating leverage), DFL (degree of financial leverage), MVA (market value
added).

(EVA, EPS, DPS, ROA, ROE, ROCE and ROS). This estimation is done beta coefficient is (ˇ = 104.15) and (p > 0.05); thus, there is no rela-
by regressions based on Eqs. (1)–(7). Then, regression Models 8, 9, tionship between EPS and MVA. Therefore, Hypothesis H1b is not
10, and 11 are used to test the second and third hypotheses. Finally, supported. This is inconsistent with the findings of Uyemura et al.
regression Models 12 and 13 are used to examine the informa- (1996), Finegan (1991), and Milunovich and Tsuei (1996). However,
tion content of EVA and other measures (Hypothesis 4). The first our results are consistent with the findings of Kumar and Sharma
hypothesis demands examining whether EVA and the traditional (2011a, 2011b).
performance measures are related to MVA prior to testing. To exam- In contrast to De wet (2005), DPS is found to have a posi-
ine these relationships, seven sub-hypotheses are developed and tive (ˇ = 309.53) impact on Iranian firms’ MVA. As can be seen in
tested. According to Tables 4 and 5, the Lagrange Multiplier test is Table 4, the calculated p-value is 0.001, which is less than the con-
statistically significant, suggesting the suitability of panel models fidence level (0.05). In other words, it can be concluded that DPS
over the pooled model. However, the Hausman tests suggest that is positively associated with MVA. Therefore, Hypothesis H1c is
the fixed effects method is an adequate parameterization of Mod- accepted. Consequently, Table 4 and the coefficients obtained for
els 1 and 12, whereas the random effects method is required for Model 4 (ˇ = 563,150, p < 0.05) show that ROA is positively associ-
Models 2–9 and 13. ated with MVA. Hence, Hypothesis H1d is also accepted. The results
Hypothesis H1a states that EVA is positively associated with of regression analysis for ROE in Table 4 and Model 5 show that
MVA. Table 4 presents the regression results for the dependent its relationship with MVA is positive (ˇ = 97,385, p < 0.05). This is
variables MVA. Table 4 and Model 1 (ˇ = 0.0478, p > 0.05) show that consistent with our expectations and Hypothesis H1e is supported.
there is a positive, yet not significant relationship between EVA and ROCE has a positive relationship with MVA, but this relation-
MVA. Consequently, Hypothesis H1a is not confirmed. This result is ship is not statistically significant (Model 6: ˇ = 65,118, p > 0.05);
consistent with the findings of Biddle et al. (1997), De Villiers and therefore, Hypothesis H1f is rejected. The results show that ROS
Auret (1997), Kramer and Pushner (1997), Chen and Dodd (1997, produces a positive effect on MVA (Model 7: ˇ = 35,607, p < 0.05);
2001), Turvey et al. (2000), Worthington and West (2001), Sparling therefore, Hypothesis H1g is supported. This finding is consistent
and Turvey (2003), De wet (2005), Ismail (2006), and Kyriazis and to Uyemura et al. (1996), Lehn and Makhija (1997), De wet (2005)
Anastassis (2007). Moreover, our results about the relationship and Kumar and Sharma (2011a, 2011b). Moreover, Table 4 shows
between EVA and MVA are inconsistent with many previous stud- the results of the independent variable relative information con-
ies, such as Stewart (1991), Stern et al. (1994), Kumar and Sharma tent tests. The assessment is made by analyzing seven separate
(2011a, 2011b), Milunovich and Tsuei (1996), and Uyemura et al. regressions for each performance measure, that is, EVA, DPS, EPS,
(1996). Hypothesis 1b predicts a positive link between EPS and ROA, ROE, ROCE, and ROS. This estimation is done by pooled and
MVA. The data provided in Table 4 and Model 2 show that the EPS panel data regressions based on Eqs. (1)–(7). Table 4 provides the

Table 3
Correlation matrix.

Variable EVA ROA ROE ROCE ROS EPS DPS SOCE SOA DOL DFL MVA

EVA 1.0000
ROA 0.0074 1.0000
ROE −0.0205 0.6777** 1.0000
ROCE −0.0002 0.4321** 0.5744** 1.0000
ROS −0.0240 0.6653** 0.4922** 0.2485** 1.0000
EPS −0.0293 0.4231** 0.5255** 0.1913 0.2335** 1.0000
DPS −0.0427 0.5128** 0.5897** 0.6581** 0.2914** 0.7540** 1.0000
SOCE −0.0026 −0.1073 0.0708* 0.5706** −0.1803** 0.5402** 0.4086** 1.0000
SOA −0.0101 0.0715 0.0181 0.1567 −0.0839** 0.1109 0.0958 0.2904** 1.0000
DOL −0.0055 0.0698 0.0388 0.0369 0.0339* 0.0487 0.0110 −0.0123 −0.0687 1.0000
DFL 0.0038 0.0149 −0.0103 −0.0123 0.0014 −0.0146 −0.0246 −0.0240 −0.0757 0.6197** 1.0000
MVA 0.0591* 0.4414** 0.3228** 0.0587** 0.4422** 0.0952** 0.1979** −0.1029 −0.2067** −0.0085 −0.0171 1.0000

Note: EVA (economic value added), ROA (return on assets), ROE (return on equity), ROCE (return on capital employed), ROS (return on sale), EPS (earnings per share), DPS
(dividend per share), SOA (sales on fixed assets), SOCE (sales on capital employed), DOL (degree of operating leverage), DFL (degree of financial leverage), MVA (market value
added).
*
Correlation is significant at the 0.05 level (2-tailed).
**
Correlation is significant at the 0.01 level (2-tailed).
Table 4
Test results of the relative information content of EVA, EPS, DPS, ROA, ROE, ROCE and ROS.

This table reports results from pooled and panel data model. The dependent variable is Market Value Added(MVA)

Models Pooled model Fixed effects model Random effects model Lagrange multiplier test Hausman test

Constant Beta (p value) Adj. R2 Constant Beta (p value) Adj. R2 Constant Beta (p value) Adj. R2

Model 1 (EVA) 70,580 (0.000)* 0.23982 (0.005)* 0.003 71,634 (0.000)* 0.0478 (0.843) 0.06 71,700 (0.000)* 0.1614 (0.4954) 0.003 45.797 (0.0000)* 5.851763 (0.0156)*
Model 2 (EPS) 51,388 (0.000)* 115.69 (0.000)* 0.011 55,037 (0.000)* 94.610 (0.143) 0.063 53,385 (0.002)* 104. 15 (0.098) 0.005 44.591 (0.0000)* 0.407975 (0.5230)
Model 3 (DPS) 32,566 (0.000)* 323.24 (0.000)* 0.039 35,741 (0.014)* 296.825 (0.002)* 0.086 342,146 (0.048)* 309.53 (0.001)* 0.031 42.733 (0.0000)* 0.214531 (0.6432)
Model 4 (ROA) −31,807 (0.000)* 54,533 (0.000)* 0.172 −46,493 (0.007)* 62,292 (0.000)* 0.343 −35,180 (0.053) 563,150 (0.000)* 0.165 29.969 (0.0000)* 0.476848 (0.4899)
Model 5 (ROE) 49,374 (0.237) 10,031 (0.000)* 0.093 10,205 (0.516 92,360 (0.000)* 0.122 68,755 (0.678) 97,385 (0.000)* 0.083 34.117 (0.0000)* 0.416298 (0.5188)
Model 6 (ROCE) 58,171 (0.000)* 77,357 (0.000)* 0.004 61,954 (0.000)* 55,247 (0.397) 0.059 60,265 (0.000)* 65,118 (0.306) 0.001 45.203 (0.0000)* 0.451539 (0.5016)

M. Alipour, M.E. Pejman / Global Economics and Management Review 20 (2015) 6–14
Model 7 (ROS) −30,749 (0.000)* 35,188 (0.000)* 0. 192 −30,749 (0.035)* 35,188 (0.000)* 0. 191 −31,964 (0.063) 35,607 (0.000)* 0. 184 33.769 (0.0000)* 0.093212 (0.7601)
Model 8 (SOA) 14,540 (0.000)* −88,424 (0.000)* 0.041 13,643 (0.000)* −77,713 (0.005)* 0.081 14,129 (0.000)* −83,518 (0.001)* 0.030 42.155 (0.0000)* 0.322367 (0.5702)
Model 9 (SOCE) 81,490 (0.000)* −14,049 (0.000)* 0.009 79,036 (0.000)* −10,629 (0.194) 0.062 801,862 (0.000)* −12,232 (0.130) 0.004 42.680 (0.0000)* 1.468847 (0.2268)
Model 10 (DOL) 71,513 (0.000)* −444.6 (0.596) −0.0002 71,533 (0.000)* −529.64 (0.8554) 0.05 715,257 (0.000)* −495.60 (0.8633) −0.003 44.749 (0.0000)* 0.006687 (0.9348)
Model 11 (DFL) 71,841 (0.000)* −1754 (0.319) −0.0002 71,735 (0.000)* −1326.4 (0.8280) 0.05 71,778 (0.000)* −1501.5 (0.8037) −0.003 44.637 (0.0000)* 0.038385 (0.8447)

Notes: EVA (economic value added), EPS (earnings per share), DPS (dividend per share), ROA (return on assets), ROE (return on equity), ROCE (return on capital employed), ROS (return on sale), SOA (sales on fixed assets),
SOCE (sales on capital employed), DOL (degree of operating leverage), DFL (degree of financial leverage). Pooled and panel data regression analysis model are performed based on Eq. (1): MVAi,t = ˇ0 + ˇ1 (EVAi,t ) +-Ci,t ; (2):
MVAi,t = ˇ0 + ˇ1 (EPSi,t ) +-Ci,t ; (3): MVAi,t = ˇ0 + ˇ1 (DPSi,t ) +-Ci,t ; (4): MVAi,t = ˇ0 + ˇ1 (ROAi,t ) +-Ci,t ;(5): MVAi,t = ˇ0 + ˇ1 (ROEi,t ) +-Ci,t ; (6): MVAi,t = ˇ0 + ˇ1 (ROCEi,t ) +-Ci,t ; (7): MVAi,t = ˇ0 + ˇ1 (ROSi,t ) +-Ci,t ; (8): MVAi,t = ˇ0 + ˇ1 (SOAi,t ) +-Ci,t ;
(9): MVAi,t = ˇ0 + ˇ1 (SOCEi,t ) +-Ci,t ; (10): MVAi,t = ˇ0 + ˇ1 (DOLi,t ) +-Ci,t ; (11): MVAi,t = ˇ0 + ˇ1 (DFLi,t ) +-Ci,t.
*
Regression significant at 5 percent level of significance. Durbin–Watson statistics (D–W) of the residuals report 1.92–2.22, respectively, for regression equations (1)–(11). We present p-values for significance in parentheses.

Table 5
Regression analysis.

Test results of the incremental information content of EVA, EPS, DPS, ROA, ROE, ROCE and ROS

Variables Pooled model Fixed effects Random effects

Model 12 Model 13 Model 12 Model 13 Model 12 Model 13

C −524,344.4 (0.0000)* −421,831 (0.0000)* −663,286.1 (0.0000)* −567,730.8 (0.0017)* −536,757.4 (0.0020)* −497,597.1 (0.0102)*
EVA 0.291051 (0.0000)* – 0.157374 (0.4592) – 0.276706 (0.1796) –
EPS 174.1860 (0.0003)* 186.4599 (0.0001)* 168.3947 (0.3157) 187.8003 (0.2493) 173.2484 (0.2924) 187.1974 (0.2461)
DPS 93.63031 (0.0320)* 50.91331 (0.2204) 149.9843 (0.3426) 93.46883 (0.5297) 99.64343 (0.5081) 74.27580 (0.6104)
ROA 3,777,749 (0.0000)* 2,969,223 (0.0000)* 4,287,829 (0.0029)* 3,557,700 (0.0080)* 3,831,984 (0.0057)* 3,287,977 (0.0125)*
ROE 257,920 (0.0076)* 223,434 (0.0104)* 116,523.6 (0.7403) 68,568.06 (0.8293) 241,435.4 (0.4688) 133,574.5 (0.6656)
ROCE −387,619 (0.0000)* −371,416 (0.0000)* −350,170.1 (0.0264)* −341,183.7 (0.0232)* −383,511.1 (0.0122)* −354,107.4 (0.0169)*
ROS 2,183,016 (0.0000)* 2,333,883 (0.0000)* 2,275,449 (0.0001)* 2,443,956 (0.0000)* 2,188,753 (0.0001)* 2,389,048 (0.0000)*

Adjusted R2 0.264 0.243 0.284 0.268 0.245 0.224

F-statistic prob. (F-statistic) 176.62 (0.000)* 193.99 (0.000)* 6.346430 (0.000)* 6.482048 (0.000)* 14.19028 (0.000)* 15.427 (0.000)*
R2 changes % 2.1% – 1.6% – 2.1% –
Durbin–Watson statistic 1.81 2.01 2.12 2.34 1.89 2.22
Lagrange multiplier test 18.843 (0.0001)* 25.529 (0.0000)* – – – –
Hausman test 17.884413 (0.0125)* 4.525127 (0.6060) – – – –

Notes: EVA (economic value added), EPS (earnings per share), DPS (dividend per share), ROA (return on assets), ROE (return on equity), ROCE (return on capital employed), ROS (return on sale), SOA (sales on fixed assets), SOCE
(sales on capital employed), DOL (degree of operating leverage), DFL (degree of financial leverage). Pooled and panel data regression analysis model are performed based on Eq. (12): MVAit = ˇ0 + ˇ1 (EVAit ) + ˇ2 (EPSit ) + ˇ3
(DPSit ) + ˇ4 (ROAit ) + ˇ5 (ROEit ) + ˇ6 (ROCEit ) + ˇ7 (ROSit ) + εit ; (13): MVAit = ˇ0 + ˇ1 (EPSit ) + ˇ2 (DPSit ) + ˇ3 (ROAit ) + ˇ4 (ROEit ) + ˇ5 (ROCEit ) + ˇ6 (ROSit ) + εit .
*
Regression significant at 5 percent level of significance. We present p-values for significance in parentheses.

11
12 M. Alipour, M.E. Pejman / Global Economics and Management Review 20 (2015) 6–14

coefficients and adjusted R2 values for each variable. First, we find Since adjusted R2 has increased by 2.1 percent, it can be concluded
that some of the regressions are significant. Similarly, the coeffi- that EVA, along with other performance measures, can explain Ira-
cient results suggest that, in terms of relative information content, nian firms’ market value. However, EVA’s explanatory power is very
some performance measures are statistically significant at the 0.05 scant and the relationship is not statistically significant. Therefore,
level. the results show that we cannot accept Hypothesis H4. Considering
The R2 values of the variables presented in Table 4 (3.1 percent the obtained beta coefficient, it can be argued that ROA and ROS are
for DPS; 16.5 percent for ROA; 8.3 percent for ROE; 18.4 percent most effective determinants of Iranian firms’ market value. More-
for ROS) reject the first hypothesis, that is, EVA is not superior to over, the regression results (Table 5) reveal that only 24.5 percent
the traditional performance measures. Considering these values, (adjusted R2 ) of changes in the MVA is explained by the regression
we can argue that ROS is superior to other measures in explaining model, leaving the majority unexplained. This means that, apart
the market value of Iranian firms. After this measure, ROA has the from these variables, one should consider other variables such as
highest R2 coefficient, and ROE, and DPS come next. Of greater con- operating cash flow and residual income to capture the exact vari-
cern is the finding that there is absolutely no relationship between ation in the sample firms’ MVA.
EVA and MVA. This is surprising, and it can be argued that the meas- According to Tables 4 and 5, we also tested the first order serial
ures based on accounting profit are superior in explaining these correlation and multicollinearity in our data. Serial correlation
firms’ market value. Therefore, the hypothesis that the relative was analyzed by examining the Durbin–Watson (D-W) statistics.
information content of EVA is superior to traditional measures (H1) The D-W statistics of the residuals report 1.81–2.34, respectively
is rejected, and we can conclude that ROS and ROA are better pre- for regression equations (1)–(13) and range in value from 0 to
dictors of Iranian firms’ market value change. Our results about the 3 with an ideal value of 2, indicating that errors are not corre-
variables’ relative information content are consistent with many lated (Field, 2000). Moreover, in all models, the level of F-statistic
international studies (e.g., Biddle et al., 1998; Chen & Dodd, 1997; is significant, suggesting the regression estimation validity. To
De wet, 2005; Kumar & Sharma, 2011a; Kumar & Sharma, 2011b; detect the presence of multicollinearity, variance inflation factor
Ray, 2001; Worthington & West, 2001). (VIF) was also analyzed. A general rule is that the VIF should not
The third hypothesis states that there is a significant positive exceed 10 (Belsley, Kuh, & Welsch, 1980). VIF values of all inde-
relationship between efficiency and firms’ market value. Mod- pendent variables were in range with the highest value of 7.492,
els 8 and 9 were examined to assess the relationship between indicating a low degree of multicollinearity among the variables.
these variables, with Table 4 providing the results. First, we found Moreover, to determine the absence of multicollinearity problems,
that the coefficient on SOA is negative and significant (p < 0.05). the Pearson’s correlation coefficients between explanatory vari-
It can thus be concluded that there is a significant negative ables were tested. Studies have suggested that multicollinearity
relationship between sales on fixed assets and MVA. Second, should be considered a serious problem only if the correlation coef-
we discovered that SOCE was negatively associated with MVA ficient between explanatory variables is more than 0.8 (Kennedy,
(ˇ = −12,232, p > 0.05). Since this relation was not significant, nei- 1985) or more than 0.9 (Tabachnick & Fidell, 1996). As shown
ther Hypothesis H2a nor Hypothesis H2b was supported. SOA in Table 3, the correlation coefficients between explanatory vari-
represents the efficiency of the management in increasing sales ables are not high. As a result, we can ignore any multicollinearity
using fixed assets, and SOCE represents the efficiency of the man- problems.
agement in increasing sales using capital employed. The negative
correlations found between SOA and MVA are quite enigmatic and
Conclusion
unexpected.
Hypothesis H3a predicted a negative link between degree of
At the beginning of this paper, we examined the informa-
operating leverage and MVA. The relationship was negative and
tion content of EVA and six traditional accounting performance
insignificant (ˇ = −495.60, p > 0.05). Based on Panel 4 and Model
measures—that is, EPS, DPS, ROA, ROE, ROCE, and ROS—in explain-
11 (ˇ = −1501.5, p > 0.05), there is a negative relationship between
ing the market value of TSE-listed firms. Then, the relationship
degree financial leverage and MVA; however, the lack of a sta-
between efficiency, degree of operating and financial leverage, and
tistically significant relationship means there is no significant
market value of these firms was examined. Using a data set of 75
relationship between leverage and MVA in TSE-listed firms and
Iranian manufacturing companies for the period 2003–2008, we
degree of operating and financial leverage cannot explain the
tested the relative and incremental information content of EVA
dependent variable. Therefore, the models show that we cannot
and conventional performance measures. Regarding incremental
accept Hypotheses H3a and H3b.
information content, EVA is not superior to other performance eval-
Hypothesis H4 states that EVA and the traditional accounting
uation measures; moreover, the results show that ROS and ROA
performance measures have information content. In order to deter-
have a greater explanatory power than EVA. Further, the results
mine the incremental information content of EVA, we used two
show that one of the efficiency measurement variables is negatively
regression models (Eqs. (12) and (13)), with all the variables and
associated with MVA. The results also indicate that TSE-listed firms
another regression model excluding EVA. The overall model sug-
show no significant relationship between leverage and MVA. The
gests that both are significant with F-values (6.34 and 14.19) that
empirical results of the study do not support the claim that EVA is a
are statistically significant at the 0.05 level.
better performance indicator than traditional accounting measures
The results of Table 5 and the obtained coefficients reveal that
in explaining market value. We find evidence supporting the earlier
only ROA, ROCE, and ROS are statistically significant and can be
work of Biddle et al. (1997), Chen and Dodd (2001), Kim (2006), and
included in the model. It can be argued that in both models ROA,
Ismail (2006), suggesting that traditional accounting-based meas-
ROCE and ROS are significantly associated with the dependent vari-
ures are more associated with MVA than EVA. Our results, in sum,
able. ROCE has negative association, whereas ROA and ROS are
do not support the claim of Stern et al. that EVA is superior to other
positively related with MVA.
measures in explaining MVA. The present findings can have the
Moreover, the R2 of Model 12 is 28.4 percent, indicating that
following implications:
26.8 percent of changes in Iranian firms’ market value can be
explained by these variables. R2 of Model 13 is 24.5 percent, and
since EVA has been incorporated in Model 12, but not Model 13, 1. Due to the weak correlation between EVA and MVA,
it can be concluded that Model 12 has higher explanatory power. investors cannot consider EVA along with the traditional
 M. Alipour, M.E. Pejman / Global Economics and Management Review 20 (2015) 6–14 13

accounting-based measures in their investment decisions. EVA Fernandez, P. (2003). EVA, economic profit and cash value added do not measure
alone cannot predict and explain the market value of the firms shareholder value creation. Journal of Applied Finance, 9(3), 74–94.
Field, A. (2000). Discovering statistics: Using SPSS for windows (1st ed.). London: Sage.
listed in the Tehran Stock Exchange and investors should pay Finegan, P. T. (1991). Maximizing shareholder value at the private company. Journal
little attention to this measure in their decision making. of Applied Corporate Finance, 4(1), 30–45.
2. Generally, the results of the present research identified and Firth, M. A. (1979). The effect of size, stock market listings, and auditors on voluntary
disclosure in corporate annual reports. Accounting and Business Research, 9(36),
introduced the measures that can help managers improve com- 273–280.
panies’ internal performance and increase shareholder wealth. Garvey, T. G., & Milbourn, T. T. (2000). EVA versus earnings: Does it matter which
is more highly correlated with stock returns? Journal of Accounting Research, 38,
209–245.
The results of the research must be interpreted with respect to Grant, J. (2003). Foundations of economic value added. Hoboken, NJ: Wiley.
the following limitations: Green, W. H. (2003). Econometric analysis. Upper Saddle River, NJ: Prentice-Hall.
Herrmann, D., Inoue, T., & Thomas, W. B. (2003). The sale of assets to manage earnings
in Japan. Journal of Accounting Research, 41(1), 89–108.
1. The independent variables that may affect the dependent vari- Hsiao, C. (1986). Analysis of panel data. Econometric Society Monographs Cambridge:
able MVA and were not examined in the present research (that Cambridge University Press.
Irala, L. R. (2007). Corporate performance measures in India: An empirical anal-
is, product quality, community satisfaction and inflation); ysis (Working papers No. WP-2007/01/A). Available from www.Papers.ssrn.
2. The possible effects of sampling limitations, such as exclusion of com/sol3/Delivery/SSRN ID967664 code709346.pdf
some samples due to lack of required data; Ismail, A. (2006). Is economic value added more associated with stock return
than accounting earnings? The UK evidence. International Journal of Managerial
3. Time limitation, which involved the financial information Finance, 2(4), 343–353.
between 2003 and 2008. Kennedy, P. (1985). A guide to econometrics (2nd ed.). Cambridge, MA: MIT Press.
Kim, K. S. (2004). Strategic planning for value-based management an empirical
examination. Management Decision, 42(8), 938–948.
It is recommended that other research be carried out that Kim, W. G. (2006). EVA and traditional accounting measures: Which metric is a
accounts for size, inflation, and industry classification. Researchers better predictor or market value of hospitality companies? Journal of Hospitality
& Tourism Research, 30(1), 34–49.
who study the information content of profit, residual income, and
Kramer, J. K., & Peters, J. R. (2001). An inter-industry analysis of economic value added
operating cash flows should compare these variables’ correlation as a proxy for market value added. Journal of Applied Finance, 11(1), 41–49.
with MVA to identify which of these measures is more powerful. In Kramer, J. K., & Pushner, G. (1997). An empirical analysis of economic value added
addition, the data can be tested using alternative dependent vari- as a proxy for market value added. Financial Practice and Education, 7(1), 41–49.
Kumar, S., & Sharma, A. K. (2011a). Association of EVA and accounting earnings with
ables (such as stock returns) with the same or a larger data set market value: Evidence from India. Asia-Pacific Journal of Business Administration,
of a similar or greater time span. Many international studies have 3(2), 83–96.
used stock returns instead of MVA to examine various performance Kumar, S., & Sharma, A. K. (2011b). Further evidence on relative and incremental
information content of EVA and traditional performance measures from select
measures’ value relevance (e.g., Ismail, 2006; Kim, 2006; Maditinos Indian companies. Journal of Financial Reporting & Accounting, 9(2), 104–118.
et al., 2009; Worthington & West, 2001). Kyriazis, D., & Anastassis, C. (2007). The validity of the EVA approach: An empirical
application. European Financial Management, 13(1), 71–100.
Lee, S., & Kim, W. G. (2009). EVA, refined EVA, MVA, or traditional performance meas-
References ures for the hospitality industry. International Journal of Hospitality Management,
28(3), 301–484.
Anderson, A. M., & Bey, R. P. (1998). An empirical analysis of the consistency, sensitivity, Lehn, K., & Makhija, A. K. (1996). EVA and MVA: As performance Measures and Signals
and validity of economic value added (Working paper). Financial Management for strategic change. Strategy and Leadership, 24, 34–38.
Association Meetings, 14–17 October. Lehn, K., & Makhija, A. K. (1997). EVA, accounting profits, and CEO turnover: an
Appleby, C. (1997). The new lingo of value added. Hospitals & Health Networks, 7(3), empirical examination, 1985–1994. Journal of Applied Corporate Finance, 10,
50–52. 90–97.
Athanassakos, G. (2007). Value-based management, EVA and stock price perfor- Maditinos, D. I., Sevic, Z., & Theriou, N. G. (2009). Modelling traditional accounting
mance in Canada. Management Decision, 45(9), 1397–1411. and modern value-based performance measures to explain stock market returns
Bahri, M., St-Pierre, J., & Sakka, O. (2011). Economic value added: A useful tool for SME in the Athens Stock Exchange (ASE). Journal of Modeling in Management, 4(3),
performance management. International Journal of Productivity and Performance 182–201.
Management, 60(6), 603–621. Milunovich, S., & Tsuei, A. (1996). EVA in the computer industry. Journal of Applied
Bao, B.-H., & Bao, D.-H. (1998). Usefulness of value added and abnormal economic Corporate Finance, 9(1), 104–115.
earnings: An empirical examination. Journal of Business Finance and Accounting, Misra, A., & Kanwal, A. (2007). EVA as the most significant measure of financial
25(1-2), 251–265. performance: A study of select Indian firms. Journal of International Business and
Belsley, D. A., Kuh, E., & Welsch, R. E. (1980). Regression Diagnostics: Identifying Economics, VII(1), 76–85.
Influential Data and Sources of Collinearity. New York, NY: Wiley. Nappi-Choulet, I., Missonier-Piera, F., & Cancel, M. (2009). Value creation and the
Biddle, G. C., Bowen, R. M., & Wallace, J. S. (1997). Does EVA beat earnings? Evidence impact of corporate real estate assets: An empirical investigation with French
on associations with stock returns and firm values. Journal of Accounting and listed companies. Journal of Corporate Real Estate, 11(2), 78–90.
Economics, 24(3), 301–336. O’Byrne, S. F. (1996). EVA and market value. Journal of Applied Corporate Finance,
Biddle, G. C., Bowen, M. R., & Wallace, J. S. (1998). Economic value added: Some 9(1), 116–125.
empirical evidence. Managerial Finance, 24(11), 60–71. Rappaport, A. (1986). Creating shareholder value. New York: The Free Press.
Chen, S., & Dodd, J. L. (1997). EVA: An empirical examination of a new corporate Ray, R. (2001). Economic value added: Theory, evidence, a missing link. Review of
performance measure. Journal of Managerial Issues, 9(3), 318–333. Business, 22, 66–70.
Chen, S., & Dodd, J. L. (2001). Operating income, residual income, and EVA: Which Riahi-Belkaoui, A. (1993). The information content of value added earnings and cash
metric is more relevant? Journal of Managerial Issues, 13, 65–86. flows. International Journal of Accounting, 28, 140–146.
Chiou, C. C., & Su, R. K. (2007). On the relation of systematic risk and accounting Rompho, N. (2009). Application of the economic value added (EVA) protocol in a
variables. Managerial Finance, 33(8), 517–533. university setting as a capital budgeting tool. Journal of Financial Reporting &
Coles, J. W., McWilliams, V. B., & Sen, N. (2001). An examination of the relation- Accounting, 7(2), 1–17.
ship of governance mechanisms to performance. Journal of Management, 27, Sparling, D., & Turvey, C. G. (2003). Further thoughts on the relationship between
23–50. EVA and stock market performance. Agribusiness, 19(2), 255–267.
De Villiers, J., & Auret, C. J. (1997). A comparison of EPS and EVA as explana- Stern, J. M., Stewart, G. B., III, & Chew, D. H. (1994). The EVA financial management
tory variables for share price. Studies in Economics and Econometrics, 22(2), systems. Journal of Applied Corporate Finance, 7(2), 32–46.
47–63. Stern, J. M., Stewart, G. B., & Chew, D. H. (1996). EVA® : An integrated financial
De wet, J. H. (2005). EVA versus traditional accounting measures of performance management system. European Financial Management, 2(2), 223–245.
as drivers of shareholder value – A comparative analysis. Meditari Accountancy Stewart, G. B., III. (1991). The quest for value. New York, NY: Harper Business.
Research, 13(2), 1–16. Tabachnick, B. G., & Fidell, L. S. (1996). Using multivariate statistics. New York, NY:
De wet, J. H., & Hall, J. H. (2004). The relationship between EVA, MVA and leverage. Harper Collins Publishers.
Meditari Accountancy Research, 12(1), 39–59. Turvey, C. G., Lake, L., Van Duren, E., & Sparing, D. (2000). The relationship between
Ehrbar, A. (1998). Economic value added: The real key to creating wealth. New York, EVA and the stock market performance of agribusiness firms. Agribusiness, 16(4),
NY: Wiley. 399–416.
Epstein, M. J., & Young, D. (1998). Improving corporate environmental perfor- Uyemura, D. G., Kantor, C. C., & Pettit, J. M. (1996). EVA for banks: Value creation,
mance through economic value added. Environmental Quality Management, 7(4), risk management, and profitability measurements. Journal of Applied Corporate
1–7. Finance, 9(2), 94–109.
14 M. Alipour, M.E. Pejman / Global Economics and Management Review 20 (2015) 6–14

Worthington, A., & West, T. (2001). The usefulness of EVA (EVA) and its compo- Zaima, J. K., Turetsky, H. F., & Cochran, B. (2005). The MVA-EVA Relationship: Sep-
nents in the Australian context. Accounting, Accountability and Performance, 7(1), aration of Market Driven versus Firm Driven Effects. Review of Accounting and
73–90. Finance, 4(1), 32–49.
Worthington, A. C., & West, T. (2004). Australian evidence concerning the informa- Zimmerman, J. L. (1997). EVA and divisional performance measurement: Cap-
tion content of economic value-added. Australian Journal of Management, 29(2), turing synergies and other issues. Journal of Applied Corporate Finance, 10(2),
201–224. 98–109.
Zafiris, N., & Bayldon, R. (1999). EVA and MVA: A simple version and application.
Journal of Applied Accounting Research, 2(2), 83–105.