INTERNATIONAL ACCOUNTING AND BUSINESS CONFERENCE 2011 PROCEEDINGS IABC2011 Page 1 THE MARKET VALUATION OF R&D: EMPIRICAL EVIDENCE FROM MALAYSIAN FIRMS SUNARTI BINTI HALID AMIZAHANUM BINTI ADAM NUR ADURA BINTI AHMAD NORUDDIN MASETAH BINTI AHMAD TARMIZI Faculty of Accountancy Universiti Teknologi MARA Seri Iskandar Campus 32610 Bandar Baru Seri Iskandar Perak Malaysia [email protected]Abstract Purpose - The major objective of this study is to understand and recognize the value relevance of research and development (R&D) in market valuation. The firms selected for this study is from Malaysia from the period 2000-2007. This study have examined whether the market perceived R&D information as an important variable in determining the value of a company. Specifically, this study empirically investigated the association between R&D information in determining and explaining the market value. The study also described a relationship between R&D with all other assets. Furthermore, we examined the relationship between the R&D and the sign of earnings items. Design/methodology/approach - An equity valuation model based on the modified balance sheet identity was used to permit R&D and other assets to have separate empirical coefficient values. Findings - This study found weak empirical support at best for the value relevance of R&D at the firm level. However, market was taken into consideration BVNA in determining the firm‟s equity value as compared to R&D. Also, the results showed that the market‟s valuation of R&D is not priced differently from other assets. In addition, our results provided evidence that there is no significant relationship between R&D information and the sign of earnings items. Originality/value - This study employs an approach using the equity valuation model to measure the value relevance of R&D in market valuation. Keywords - Research and Development, Market Valuation, Equity Valuation Model Paper type - Research Paper
38
Embed
The Market Valuation of R&D: Empirical Evidence from Malaysian Firms
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
R/Djt = Research and development of firm j in year t
MVEjt = Market value of shareholders‟ equity of firm j in year t
BVNAjt = Book value of the net assets minus R&D of firm j in
year t
EARNjt = Net profit of firm j in year t
DEARNjt = Dummy variable taking the value of 1 for positive
earnings and 0 otherwise
ejt = Error term
From the above models discussed, error terms are independent, identically normally distributed
with mean 0 and a constant variance, σ2.
Research Hypotheses
Three hypotheses have been developed and will be tested in this study. In fact, the analysis that
is going to be performed will be based on these three hypotheses. The first hypothesis to be
addressed in this study is whether R&D should be considered as an important element when
determining a firm‟s market value. In order to achieve this objective, a3 is the coefficient of main
interest (as in Model 3.2). If the market places value on R&D of a firm, then R&D should be
significant and positively correlated with the firm‟s market value. In order to check for this
relationship the following null hypothesis is tested based on the Model 3.2:
H1: a3 = 0
If the R&D information is significant variable, then further examination should test how the
market perceives R&D in relation to all other assets. In other words, is it priced differently from
other assets? In order to check for this relationship, the following null hypothesis is established
based on the Model 3.2:
H2: a1 = a3
Meanwhile, the third hypothesis examine whether there is any relationship between R&D and the
sign of earnings items throughout the study period. In order to check for this relationship, the
following null hypothesis is tested based on the Model 3.3:
INTERNATIONAL ACCOUNTING AND BUSINESS CONFERENCE
2011
PROCEEDINGS IABC2011 Page 15
H3: There is no relationship between R&D information and the sign of earnings items.
Description of Data Collection
As stated earlier, the main objectives of the study is to investigate empirically the association
between R&D information in determining and explaining the market value and to establish a
relationship between R&D information with all other assets specifically over the period of 2000
until 2007 based on Malaysian firms. Literally, assets are rights accruing to the entity meanwhile
equities represent sources of the assets and consists of liabilities and the stockholders equity.
Thus, income earned is the property of the entity until it is distributed as dividends to the
shareholders. Hence, the firm‟s book value of net assets (excluding R&D), earning and R&D
will be the independent variables in the framework. Consequently, the theoretical model is
presents in Figure 3.1.
Figure 3.1: The Framework for the Relationship between Independent and Dependent
Variables
Sample Selection
It is now possible to see how both sampling design and sample size are important to establish the
representativeness of the sample for generalizability (Sekaran, 2000). According to the Sekaran,
if the appropriate sampling design is not used, a large sample size will not, in itself, allow the
Independent Variables
Book Value of Net Assets (Excluding R&D)
R&D
Earnings
Dependent Variable
Market Value of Equity
(Share price x ordinary
shares outstanding)
INTERNATIONAL ACCOUNTING AND BUSINESS CONFERENCE
2011
PROCEEDINGS IABC2011 Page 16
findings to be generalized to the population. Similarly, unless the sample size is adequate for the
desired level of precision and confidence, no sampling design, however sophisticated, can be
useful to the researcher in meeting the objectives of the study (Sekaran, 2000). Thus, sampling
decisions should consider both the sampling design and the sample size.
Black and White (2003) have listed several criteria in choosing sample for their research. These
criteria, as set by Black and White, are used in selecting sample for this research. Thus, the study
population consists of Malaysian firms. The coverage of the study is seven years, starting with
year 2000 until 2007 fiscal year from the listing datastream. Indeed, the data of this study are
extracted from the Balance Sheet and Profit and Loss Statement of the respective firms. Data for
this study were collected from the Thompson One Banker over seven-year period from 2000 to
2007.
A firm-year is included as observation if all such variables (market value of shareholders‟ equity,
book value of net assets, earnings and R&D) are presented for a given fiscal year. Firm-year
from the selected companies with any missing variables is excluded. As a result, the final
sample consists of various sample sizes during the period under study. Table 3.1 summarizes the
sample selection and size used for the study. After excluding the missing observations of
variables market value of equity, book value of net assets, earning and R&D, the final sample for
this study is 387 firm-year observations.
Table 3.1: Sample Selection and Size
Sample Selection
Firm-years
Thompson One Banker 2000-2007
9872
Missing observations of market value of equity
(MVE), book value of net assets (BVNA),
earnings (EARN) and capitalized R&D (R/D)
(9485)
Sample Size
387
Table 3.2: Sample Classified by Years
Year Original
Datastream
Clean Data
(Record R&D)
Non-record
R&D
2000 1234 19 1215
2001 1234 31 1203
INTERNATIONAL ACCOUNTING AND BUSINESS CONFERENCE
2011
PROCEEDINGS IABC2011 Page 17
2002 1234 35 1199
2003 1234 46 1188
2004 1234 53 1181
2005 1234 63 1171
2006 1234 69 1165
2007 1234 71 1163
Variable Definitions
The accounting variables included in the regression model are market value of equity, book
value of net assets, earnings and research and development. A summary of the variables of
interest is presented in Table 3.3. The market value of shareholders‟ equity (MVE) is defined as
the share price multiplied by the number of shares outstanding at the end of the accounting year.
The book value of total assets, research and development (R&D), total liabilities and the earning
figure (EARN) are also taken directly from the Thompson One Banker without any modification,
but with variables combined in some cases as shown. However, the book value of net assets
(BVNA) is derived by deducting the total assets (excluding R&D) with total liabilities.
Table 3.3: Variables Required for Regression from Thompson One Banker
Name variables required for
regression Variables Symbol
Market value of equity
Ordinary share outstanding x share price
MVE
Book value of total assets
Total assets
Book value of total liabilities
Total liabilities
Total sales
Turnover
Earnings
Profit attributable to Shareholders
EARN
INTERNATIONAL ACCOUNTING AND BUSINESS CONFERENCE
2011
PROCEEDINGS IABC2011 Page 18
Name variables required for
regression Variables Symbol
Net assets
Book value of total assets - Book value of total
liabilities - Research and development
BVNA
Research and Development
R&D to sales
R&D
Measurement
Measurement procedures
For purposes of empirical analysis, this study uses descriptive statistics and regressions analysis
as the underlying statistical tests. A descriptive statistics of the data obtained will be conducted
to obtain sample characteristics. In other words, descriptive statistics are used to describe and
summarize the dependent and independent variables. Apart from that, correlation is a statistical
method used to answer questions about relations between variables.
The correlation coefficient is a number that measurers the strength of the relation variables.
Values of the correlation coefficient can range from +1.00 for perfectly positively correlated
variables to -1.00 for perfectly negatively correlated variables. A correlation coefficient of -1.00
and +1.00 indicates perfect correlation. If there is absolutely no relationship between the two
variables, the correlation coefficient will be zero. A coefficient correlation that is close to zero
shows that the relationship is quite weak.
Past studies have used Ordinary Least Square (OLS) regression of market values on accounting
measurers to examine value relevance. Thus, OLS regression test is performed on the dependent
variable (MVE), to check the relationship between the R&D accounting numbers for the firms
operating in selected countries. OLS is based on a number of assumptions about the variables
and the error term that must be satisfied in order to ensure the interpretations of the regression
estimates are valid. According to Gujarati (1995), under these assumptions, the OLS estimators
of the regression coefficients are the best linear unbiased estimator.
Basically, r-squared measures the movement or changes in a variable that can be explained by
movements in another variable.2 A variable with greater r-squared indicates explanatory power
of that variable in explaining market value of equity. If the r-squared is lower, then the
explanatory variable is less relevant. A 5% significance level was used in this study. This test is
performed using the MICROFIT 4.0 software package. Besides, the analysis of the data is based
on cross-sectional regression. This study also discusses two major statistical problems associated
with the estimation of the models. The two major problems are heteroscedasticity disturbances
2 Literally, r-squared is the proportion of the variance of Y that has been explained by X. For technical reasons (the
total squared error can be decomposed into two squared components: explained and unexplained), the variance (the squared standard deviation) has traditionally been used.
INTERNATIONAL ACCOUNTING AND BUSINESS CONFERENCE
2011
PROCEEDINGS IABC2011 Page 19
and multicollinearity. Heteroscedasticity is the most common statistical problem to be
encountered when estimating a cross-sectional valuation model.
According to Ibrahim et al. (2003a), one of the major econometric problems when estimating
cross-sectional valuation models is heteroscedastic disturbances that appear from the fact that
large (small) firms tend to produce large (small) disturbances. If heteroscedasticity is present,
then the usual OLS estimators, although unbiased, no longer exhibit minimum variance among
all linear unbiased estimators (Gujarati, 1995). In short, they are no longer the best linear
unbiased estimator. Meanwhile, in the case of the two-variable linear model, one common
deflation technique involves transforming the variables by deflating the independent variable
(Landsman, 1986). This procedure implies that the true error variance is proportional to the
square of the independent. Landsman (1986) addressed the heteroscedasticity problem by
estimating the model in deflated form. In this respect, all variables are deflated by total sales.
Besides, the multicollinearity problem will be discussed in depth in Chapter 4.
Finally, the Wald Test is computed in Model 3.2 to measure whether the information provided
by one variable is significantly different from that provided by another. The Wald Test is
performed using the MICROFIT 4.0 software package. In this study, the Wald Test is computed
in order to check how the market perceives R&D in relation to all other assets.
Deflation Technique
Potential statistical problems associated with the estimation of the model were also noted in the
models that are relevant to the present study; as in the examples in Landsman (1986). The major
problem is heteroscedasticity disturbances. In addressing this issue, entire variables are
transformed by deflating them with the independent variable, which in this study is earnings/total
sales, to produce a constant (but still unknown) variance. Through this „deflation technique‟ the
heteroscedasticity problems can be minimized. The „deflation technique‟ has been widely used
by previous researchers, for example Landsman (1986), Shevlin (1991), McCarthy and
Schneider (1995) and Jennings et al. (1996). In these studies, (except for Shevlin, 1991) all data
in the basic models are deflated by total sales in order to reduce the heteroscedasticity problems
as well as to increase efficiency.
Findings & Conclusion
Descriptive Statistics
Estimates of Correlation on the Independent and Dependent Variables
A summary of the estimates of correlation between independent and dependent variables is
reported in Table 4.1. Table 4.1 presents the correlation of market value of equity (MVE) and
independent variables which include the book value of net assets (BVNA), earnings of the firm
(EARN) and research and development (R&D) in Malaysian firms during the period 2000-2007.
From the table, we can see that the correlation between MVE and R&D are considerably
INTERNATIONAL ACCOUNTING AND BUSINESS CONFERENCE
2011
PROCEEDINGS IABC2011 Page 20
moderate (Year 2000 = 0.379; Year 2001 = 0.438; Year 2002 = 0.544; and Year 2003 = 0.198;
Year 2004 = 0.057; Year 2005 = 0.122; Year 2006 = 0.118; and Year 2007 = 0.240) and a
significant positive. In other words, higher correlation between independent and dependent
variables can predict the regression of market value. Apart from that, based on the estimated
coefficients of correlation between other assets in valuing the market shareholder‟s equity in
Malaysian firms for 2000 until 2007 shows a significant positive correlation. Besides, the value
for correlation between BVNA and MVE is somewhat higher than the correlation between R&D
with MVE.
Table 4.1: Estimated Correlation Matrix of Variables for Malaysian Firms
Variables
BVNA
EARN
RD
2000
MVE
BVNA
EARN
0.654
-0.110
-0.600
0.379
0.018
0.133
2001
MVE
BVNA
EARN
0.670
-0.323
-0.460
0.438
0.111
-0.191
2002
MVE
BVNA
EARN
0.483
0.388
-0.033
0.544
0.255
0.254
2003
MVE
BVNA
EARN
0.359
0.342
0.285
0.198
0.405
-0.017
2004
MVE
BVNA
EARN
0.329
0.190
0.419
0.057
0.580
0.033
2005
MVE
BVNA
EARN
0.362
0.581
0.539
0.122
0.300
0.077
2006
MVE
BVNA
EARN
0.552
0.543
0.574
0.118
0.192
0.118
2007
MVE
BVNA
EARN
0.626
0.304
0.233
0.240
0.399
-0.292
Descriptive Statistics on the Independent and Dependent Variables
In general, descriptive statistics summarizes and describes the observation of the data used in this
study. As mentioned in the Chapter 3, one potential econometric problem when estimating cross-
sectional valuation models is the problem of heteroscedastic disturbances, which arise from the
fact that large or small companies tend to produce large or small disturbances. Thus, to address
this issue, the whole variables are transformed by deflating all variables with the total
INTERNATIONAL ACCOUNTING AND BUSINESS CONFERENCE
2011
PROCEEDINGS IABC2011 Page 21
sales/earnings in order to produce a constant variance but still unknown. Besides, this „deflation
technique‟ is hoped to eliminate the heteroscedasticity problem. A summary of the overall
descriptive statistics obtained from the sample size is presented in Table 4.2. The table provides
the mean, standard deviation and coefficient of variation for selected variables: market value of
equity, book value of net assets, earnings and R&D, during the period 2000-2007.
Table 4.2 presents the mean, standard deviation and coefficient of variation for selected variables
in Malaysian firms during the period 2000-2007. Besides, all the variables were deflated by the
total sales. From the table, the results show that the values vary significantly throughout the
study period. The mean values of R&D in Malaysian companies are considerably moderate
throughout the study period (Year 2000 = 0.956; Year 2001 = 0.770; Year 2002 = 0.762; Year
2003 = 1.396; Year 2004 = 2.024; Year 2005; 2.011; Year 2006 = 2.369; and Year 2007 =
2.897). It concludes that R&D is a significant activity from the sample of Malaysian firms.
Apart from that, the mean values of variable MVE in Malaysian companies are also considerably
moderate (Year 2000 = 1.698; Year 2001 = 1.343; Year 2002 = 1.433; Year 2003 = 1.626; Year
2004 = 1.620; Year 2005 = 1.432 ;Year 2006 = 1.816; and Year 2007 = 1.536). Meanwhile, in
2001, the mean value of earnings was negative, which constitute of -0.052 (as reported in Table
4.2). The negative value shows that a few firms in Malaysia are suffering from losses in Year
2001. Standard deviation is a measure of the spread of data in relation to the mean. The standard
deviation is the traditional choice and is the most widely used. It summarizes how far an
observation typically is from the average. It is the most common measure of the variability of a
set of data. If the standard deviation is smaller, it means the probability of distribution is tighter.
From Table 4.2 it can be seen that standard deviation values for R&D is somewhat higher than
other selected variables during the period 2000-2007 (Year 2000 = 1.322; Year 2001 = 1.226;
Year 2002 = 1.036; Year 2003 = 3.785; Year 2004 = 8.502; Year 2005 = 8.239; Year 2006 =
8.874; and Year 2007 = 10.461). Meanwhile, the coefficient of variation is defined as the
standard deviation divided by the average and is a relative measure of variability as a percentage
or proportion of the average. The result also shows that the coefficient of variation for earnings
(Year 2000 = 9.641; Year 2001 = 11.508; Year 2002 = 9.828; Year 2003 = 3.810; Year 2004 =
75.112; Year 2005 = 3.817; Year 2006 = 7.901 and Year 2007 = 15.049) is higher than
coefficient of variation for R&D (Year 2000 = 1.381; Year 2001 = 1.591; Year 2002 = 1.360;
Year 2003 = 2.711; Year 2004 = 4.199; Year 2005 = 4.095; Year 2006 = 3.745 and Year 2007 =
3.609). This measure (coefficient of variation values for earnings) which is the ratio of standard
deviation of earnings to mean earnings capture the volatility for earnings for a given mean dollar
amount of earnings. In other words the earnings are more volatile than R&D in Malaysian firms
throughout the study period.
Table 4.2: Descriptive Statistics for Malaysian Firms (Deflated Form - Total Sales as
Deflator)
Variables N Mean Std.
Deviation
Coef. Of
Variation
2000
INTERNATIONAL ACCOUNTING AND BUSINESS CONFERENCE
2011
PROCEEDINGS IABC2011 Page 22
Variables N Mean Std.
Deviation
Coef. Of
Variation
Market value of equity
Book value of net assets
Earnings
Research and development
19 1.698
1.088
0.025
0.956
1.237
1.226
0.243
1.322
0.728
1.126
9.641
1.381
2001
Market value of equity
Book value of net assets
Earnings
Research and development
31
1.343
1.193
-0.052
0.770
1.473
1.020
0.608
1.226
1.096
0.855
11.508
1.591
2002
Market value of equity
Book value of net assets
Earnings
Research and development
35
1.433
1.239
0.024
0.762
1.179
0.860
0.235
1.036
0.822
0.693
9.828
1.360
2003
Market value of equity
Book value of net assets
Earnings
Research and development
46
1.626
1.137
0.078
1.396
1.488
0.801
0.299
3.785
0.915
0.704
3.810
2.711
2004
Market value of equity
Book value of net assets
Earnings
Research and development
53
1.620
1.002
0.006
2.024
2.152
0.886
0.524
8.502
1.328
0.884
75.112
4.199
2005
Market value of equity
Book value of net assets
Earnings
Research and development
63
1.432
0.969
0.065
2.011
1.890
1.086
0.250
8.239
1.320
1.121
3.817
4.095
2006
Market value of equity
Book value of net assets
Earnings
Research and development
69
1.816
0.913
0.037
2.369
2.404
1.249
0.294
8.874
1.324
1.368
7.901
3.745
2007
Market value of equity
Book value of net assets
Earnings
Research and development
71
1.536
1.015
0.037
2.897
1.809
0.968
0.563
10.461
1.177
0.953
15.049
3.609
Empirical Results
INTERNATIONAL ACCOUNTING AND BUSINESS CONFERENCE
2011
PROCEEDINGS IABC2011 Page 23
The Relationship between R&D and Firm’s Market Value
Regression analysis is used to predict the relationship between one variable (dependent variable)
on the basis of other variables, also known as independent variables. Table 4.3 summarizes the
statistics from the basic regression model that have defined the market value of equity (MVE) as
the share price times number of shares outstanding at the end of accounting year. There are
several outstanding general findings associated with the results appearing in the table below. All
the variables have coefficients of the correct sign. The intercept term (a0) is systematically non-
zero for the years 2000 to 2007 at the 5% significant level.
As mentioned, a1, a2 and a3 are the slope coefficients for book value of net assets, earnings and
research and development respectively. The main interest of this study is on a3, the slope
coefficient for R&D. If the market places value on the reported R&D of a firm, then a3 should be
significant positively correlated with the firm‟s market value. An examination of Table 4.3
reveals that the intercept term (a0) is systematically non-zero. Specifically, the value of intercept
varies for overall result except for Year 2001 (a0 = -0.118). Other than that, the intercept term a0
is significant at 5% level except in Year 2000 (a0 = 0.414, OLS t = 1.330, p = 0.203); Year 2001
(a0 = -0.118, OLS t = -0.377, p = 0.709); Year 2002 (a0 = 0.399, OLS t = 1.509, p = 0.141); Year
2004 (a0 = 0.652, OLS t = 1.319, p = 0.193); and Year 2007 (a0 = 0.431, OLS t = 1.737, p =
0.087). The presence of a statistically significant intercept suggests that the empirical intercept
may be picking up some omitted variable.
In this context, following the argument presented by Kane and Unal (1990), the intercept would
be interpreted as unbooked assets and liabilities. Kane and Unal believed accountants‟
misvaluations of portfolio positions that accounting principles designate, as on-balance sheet
items and the systematic neglect of off-balance sheet sources of value not formally booked
become sources of hidden capital. In other words, they interpreted the estimated intercept as a
net source of (drain on) unbookable assets and liabilities.
In this study, the a3 coefficient for R&D throughout the study period is found to be significantly
non-zero as reported in Table 4.3 [Year 2000 (a3 = 0.293, OLS t = 2.015, p = 0.062); Year 2001
(a3 = 0.451, OLS t = 2.972, p = 0.006); Year 2002 (a3 = 0.413, OLS t = 2.697, p = 0.011); Year
2003 (a3 = 0.042, OLS t = 0.700, p = 0.487); Year 2004 (a3 = -0.050, OLS t = -1.164, p = 0.250);
Year 2005 (a3 = 0.015, OLS t = 0.592, p = 0.556); Year 2006 (a3 = 0.002, OLS t = 0.101, p =
0.920); and Year 2007 (a3= 0.013, OLS t = 0.708, p = 0.481)]. Besides, an examination of Table
4.3 reveals that only two years is found to be significant at the 5% level [Year 2001 (p = 0.006)
and Year 2002 (p = 0.011)]. This suggests that, from the investors‟ point of view, R&D
information represents an economic resource for the year 2001 and 2002. It can be concluded
that R&D investments by Malaysian firms have been significantly lower than their counterparts
from developed countries.
INTERNATIONAL ACCOUNTING AND BUSINESS CONFERENCE
2011
PROCEEDINGS IABC2011 Page 24
Apart from that, Malaysia‟s R&D intensity still indicates that it is still below that of other
developing countries such as South Africa, Pakistan, India, China, Brazil and Venezuela. A
major reason for Malaysia‟s low R&D intensity is that many Malaysian enterprises are oriented
towards the domestic market and thereafter remain less pressure to be innovative, through R&D,
to be competitive in the international markets (White Paper - Comparative Analysis of R&D
Developments in Malaysia, 2002). In fact, annual surveys carried out by the United Kingdom
Department of Trade & Industry, none of Malaysian companies ranked in the top 500
international companies that undertook R&D investment in the year 2000 and 2001. This shows
that firms in Malaysia do not spent as much on R&D as compared to other firms in developed
countries. Apart from that, R&D activities among firms in Malaysia were very limited.
Compared to firms in other countries such as the United States, Japan and Germany, the amount
spent on R&D by Malaysian firms was very much less (Alfan, 2003).
Apart from that, investors more perceive on BVNA in market valuation. The result (as reported
in Table 4.3) is as follows: [Year 2000 (a1 = 0.878, OLS t = 4.512, p = 0.000); Year 2001 (a1 =
0.938, OLS t = 4.654, p = 0.000); Year 2002 (a1 = 0.549, OLS t = 3.074, p = 0.004); Year 2003
(a1 = 0.440, OLS t = 1.488, p = 0.144); Year 2004 (a1 = 1.066, OLS t = 2.346, p = 0.023); Year
2005 (a1 = 0.077, OLS t = 0.337, p = 0.737); Year 2006 (a1 = 0.685, OLS t = 2.952, p = 0.004);
and Year 2007 (a1= 1.025, OLS t = 4.972, p = 0.000)]. Consequently, these findings confirm the
belief that the market was taking into consideration BVNA in determining the firm‟s equity
value as compared to R&D.
As explained in Chapter 3, r-squared measures the movement or changes in a variable that can be
explained by movements in another variable. A variable with greater r-squared indicates
explanatory power of that variable in explaining market value of equity. In general, the higher
the r-squared value, the better the model fits the data. However, if r-squared is lower, then the
explanatory variable is less relevant. Table 4.3 reports the result of r-squared value for Malaysian
companies across the year 2000-2007 (Year 2000 = 0.648; Year 2001 = 0.585; Year 2002 =
0.512; Year 2003 = 0.200; Year 2004 = 0.136; Year 2005 = 0.345; Year 2006 = 0.381; and Year
2007 = 0.423). On top, we can conclude that the Model 3.2 has its explanatory power due to a
higher value of r-squared for year 2000, 2001, 2002 and 2007. Therefore, there is an evidence to
infer that a linear relationship exists between the dependent and independent variables for whole
year of study.
INTERNATIONAL ACCOUNTING AND BUSINESS CONFERENCE
2011
PROCEEDINGS IABC2011 Page 25
Table 4.3: Market Value Predictions for Malaysian Firms (Basic Model)