The current issue and full text archive of this journal is available at

www.emeraldinsight.com/1469-1930.htm

JIC
8,1 Intellectual capital and financial
returns of companies
Hong Pew Tan, David Plowman and Phil Hancock
76 The Graduate School of Management, University of Western Australia,
Crawley, Australia

Abstract
Purpose – The purpose of the paper is to investigate the association between the intellectual capital
(IC) of firms and their financial performance.
Design/methodology/approach – The paper uses the Pulic framework, has an Asian focus, and
draws on data from 150 publicly listed companies on the Singapore Exchange. It is an empirical study
using partial least squares (PLS) for the data analysis. The paper tests four elements of IC and
company performance.
Findings – The findings show that: IC and company performance are positively related; IC is
correlated to future company performance; the rate of growth of a company’s IC is positively related to
the company’s performance; and the contribution of IC to company performance differs by industry.
Research limitations/implications – The data sample is restricted to 150 companies listed on the
Singapore Exchange between the years 2000 and 2002.
Practical implications – IC is an area of interest to numerous parties, such as shareholders,
institutional investors, scholars, policymakers and managers. The findings help to embolden modern
day managers to better harness and manage IC.
Originality/value – The study of IC has undergone a number of stages, from early conscious
awareness efforts to classification of IC, and to the search for appropriate measures of IC. This paper
builds on the current research on IC and provides empirical evidence on the relevance of IC (as
measured by the Pulic model) to the financial performance of companies.
Keywords Intellectual capital, Company performance, Regression analysis
Paper type Research paper

Journal of Intellectual Capital
Vol. 8 No. 1, 2007
pp. 76-95
q Emerald Group Publishing Limited
1469-1930
DOI 10.1108/14691930710715079
1. Introduction
The rise of the “new economy”, one principally driven by information and knowledge,
has led to an increased interest in intellectual capital (IC) (Stewart, 1997; Thurow, 1999;
Petty and Guthrie, 2000; Bontis, 2001). An area that has captured the interest of a
number of scholars and practitioners is the utility of IC as an instrument for
determining enterprise value (Stewart, 1997; Edvinsson and Malone, 1997; Sveiby,
1998; Thurow, 1999; Lev and Feng, 2001; Guthrie, 2001). This has been a vexed issue,
with some writers suggesting that established management and reporting systems are
increasingly losing their relevance because they are unable to provide executives with
information that is essential for managing knowledge-based processes and intangible
resources (Bornemann and Leitner, 2002).
Historically, the distinction between intangible assets and IC has been, at best,
vague with intangibles, including IC, being referred to as “goodwill” (Accounting
Principles Board, 1970; Accounting Standards Board, 1997; International Accounting
Standards Committee, 1998). This can be traced back to the early 1980s when the
general notion of intangible value, often labelled as goodwill, began to surface in
accounting and business practices (International Federation of Accountants, 1998).
However, traditional accounting practice does not provide for the identification and IC and financial
measurement of these intangibles in organisations, especially knowledge-based returns
organisations (Guthrie et al., 1999; International Federation of Accountants, 1998;
Society of Management Accountants of Canada, 1998). New intangibles such as staff
competencies, customer relationships, simulation models, computer and administrative
systems receive no recognition in the traditional financial and management reporting
models (Stewart, 1997, pp. 56-9). Interestingly, even traditional intangibles such as 77
brand equity, patents, and goodwill are rarely reported in the financial statements
(International Federation of Accountants, 1998; International Accounting Standards
Committee, 1998). In fact, International Accounting Standard IAS 38, Intangible Assets,
prohibits the recognition of internally generated brands, mastheads, publishing titles,
and customer lists. (International Accounting Standards Board, 2004)
The objective of this paper is to empirically examine the relationship between an
extant measure of IC – the Value Added Intellectual Coefficiente developed by Pulic
(1998) – and traditional measures of corporate performance:
(1) return on equity (ROE);
(2) earning per share (EPS); and
(3) annual share returns (ASR).

The paper also explores the different contributions made by IC to companies operating
in different industries. Data for the analysis are drawn from a sample of 150 publicly
listed companies from the Singapore Exchange for 2000 and 2002. The first date is
associated with the introduction of new reporting requirements in Singapore,
requirements that included a number of measures necessary for the Pulic model that
informs the paper’s methodology. The last date coincides with data availability at the
time of writing. Singapore is chosen because of that country’s stated objective of
developing that country as an important centre in the knowledge economy.
The paper is broken into seven substantive parts. Following this introductory section,
the next section examines various classifications and measures that have been used for
IC. This is followed by an exposition of the Pulic model that forms the basis of this paper.
Section 4 details the research questions examined in this paper and the various measures
used. Section 5 describes the methodology used and the results of the data analysis are
detailed in Section 6. The final section is by way of summary and conclusion.

2. Classifying and measuring IC

There has been an epidemic of research into IC in recent years that has transformed
both its focus and scope. The research has also led to a number of frameworks for
classifying and measuring the concept. The classificatory models that have been
developed include Petrash’s (1996) Value Platform model. This classifies IC as the sum
of human capital, organisational capital and customer capital. Edvinsson and Malone
(1997) developed the Skandia Value Scheme, which classifies IC into structural capital
and human capital. Haanes and Lowendahl (1997) classify the IC of a company into
competence and relational resources. The Lowendahl’s (1997) model refines the above
model and divides the competence and relational categories into two subgroups:
(1) individual; and
(2) collective.
JIC Stewart (1997) classifies IC into three basic forms:
8,1 (1) human capital;
(2) structural capital; and
(3) customer capital.

The Danish Confederation of Trade Unions (1999) classifies IC as people, system and
78 market. The European Commission (MERITUM, 2001) classifies IC as human capital,
structural capital and relationship capital. Leliaert et al. (2003) developed the 4-Leaf
model, which classifies IC into human, customer, structural capital and strategic
alliance capital. This non-exhaustive list serves to show the similarities and differences
in the classificatory systems that have been developed.
IC measuring methods can be grouped broadly under two categories:
(1) those that do not use a monetary valuation of IC; and
(2) those that put a monetary value on IC.

The latter includes not only methods that attempt to estimate dollar values of IC, but
also those that derive the monetary values through the use of financial ratios. A
selective list of key measures is shown below.
The key non-dollar valuation of IC models are:
.
the Balance Scorecard, developed by Kaplan and Norton (1992);
.
Brooking’s (1996) Technology Broker method;
.
the Edvinssion and Malone (1997) Skandia IC Report method;
.
the IC-Index developed by Roos et al. (1997);
.
Sveiby’s (1997) Intangible Asset Monitor approach;
.
the Heuristic Frame developed by Joia (2000);
. Vanderkaay’s (2000) Vital Sign Scorecard; and
. the Ernst & Young Model (Barsky and Marchant, 2000).

The key dollar valuation of IC models are:

.
the EVA and MVA model (Bontis et al., 1999);
.
the Market-to-Book Value model (various authors);
.
Tobin’s q method (Luthy, 1998);
.
Pulic’s VAICTM Model (1998, 2000);
.
calculated intangible value (Dzinkowski, 2000); and
.
the Knowledge Capital Earnings model (Lev and Feng, 2001).

Other methods from accounting bodies and practitioners are:

.
Human Resource Costing & Accounting (Johanson and Grojer, 1998);
.
Accounting for the Future (Nash, 1998);
.
Total Value Creation (McLean, 1999); and
.
The Value Explorere and Weightless Weights (Andriessen and Tissen, 2000;
Andriessen, 2001).
The measuring techniques for IC are still evolving and researchers are now attempting IC and financial
to apply the concept to competitive advantage. The classificatory and measurement
model of interest to this paper is the Pulic model. This is outlined in the next section.
returns

3. The Pulic model

Pulic (1998, 2000) developed the “Value Added Intellectual Coefficient” (VAICe) to
measure the IC of companies. He is concerned with two other important aspects of 79
valuation and value creation yet unsolved by other methods:
(1) Market-based IC value cannot be calculated for companies that are not listed on
the stock market. Such companies need an alternative way to determine their
market-based IC value.
(2) There is no adequate system of monitoring the efficiency of current business
activities performed by employees, or whether their potential is directed
towards value creation or value destruction.
The VAICe method is designed to provide information about the value creation
efficiency of tangible and intangible assets within a company. The model starts with a
company’s ability to create value added (VA). VA is the difference between sales (OUT)
and inputs (IN) and is represented by the following equation:
VA ¼ OUT 2 IN:
Outputs (OUT) represent the revenue and comprise all the products and services sold
on the market. Inputs (IN) contain all the expenses incurred in earning the revenue
except manpower costs. It is important to note that in this model labour expenses are
not included in IN. Due to its active role in the value creating process, intellectual
potential (represented by labour expenses) is not counted as a cost. Thus, a key aspect
in Pulic’s method is to treat labour as a value creating entity. The result is that VA
expresses the new created wealth of a period. VA is influenced by the efficiency of
Human Capital (HC) and Structural Capital (SC).
The second relation of VA, one employing physical capital (CA), is called the ‘value
added capital coefficient’ (VACA). This is an indicator for the VA created by one unit of
physical capital:
VACA ¼ VA=CA:
Pulic assumes that if a unit of CA generates greater returns in one company than another,
then the first company is better at utilisation of its CA. Thus, better utilisation of CA is
part of the IC of companies. When compared over a group of companies, VACA becomes
an indicator of the intellectual abilities of the company to better harness physical capital.
The next relation is VA and HC. The ‘human capital coefficient’ (VAHU) shows how
much VA is created by a dollar spent on employees. The relation between VA and HC
indicates the ability of HC to create value in a company. Consistent with views of other
leading IC authors (Edvinsson, 1997; Sveiby, 1998), Pulic (1998) argues that total salary
and wage costs are an indicator of a firm’s HC. Pulic posits that since the market
determines salaries as a result of performance, it is only logical that the success of HC
should be expressed by the same criteria. Thus, the relation between VA and HC
indicates the ability of HC to create value in a company:
VAHU ¼ VA=HC:
JIC Similarly, when VAHU is compared over a group of companies, VAHU becomes an
indicator of the quality of the human resources of the company and their abilities to
8,1 generate VA for every dollar spent on HC.
The third relation is “structural capital coefficient” (STVA), which shows the
contribution of structural capital (SC) in value creation. In Pulic’s model, SC is VA
minus HC. The lesser the contribution of HC in value creation the greater is the
80 contribution of SC. According to Pulic (2000), this has been verified by empirical
research that shows in traditional industrial sectors. In heavy industry and mining for
example, VA is only slightly bigger than HC with an insignificant SC component. On
the other hand, in the pharmaceutical industry and software sectors, an entirely
different situation is observed. HC creates only 25-40 percent of the entire VA and the
major contribution is due to SC. Therefore, the third relation between VA and
employed SC is calculated in a different way because HC and SC are in reverse
proportion as far as value creation is concerned. STVA measures the amount of SC
needed to generate a dollar of VA and is an indication of how successful SC is in value
creation. Unlike VACA and VAHU, VA is in the denominator for STVA. Thus, the
third relation between VA and SC is calculated as:

STVA ¼ SC=VA:

The final ratio is the calculation of the intellectual ability of a company. It is the sum of
the previously mentioned coefficients. This results in a new and unique indicator – the
VAICe:

VAICe ¼ VACA þ VAHU þ STVA:

Pulic’s method has the attraction of ease of data acquisition and enables further
analysis to be conducted on other data sources. Data needed to derive the various ratios
are standard financial numbers that are normally available from audited financial
reports of companies. Alternative IC measures are limited in that they involve unique
financial and non-financial indicators and are usually customised to fit the profile of
individual firms (Roos et al., 1997). Some of these indicators, especially non-financial
ones, are not readily available or may not be recorded by other firms. Consequently, the
ability to apply alternative IC measures consistently across large and diversified
samples for comparative analysis is diminished (Firer and Williams, 2003). Also, to
enhance the external validity of a study, data sources must be available for a sufficient
sample size and tests can be replicated to other data sources.

4. Research questions and measures

Using Pulic’s model, this paper addresses four research propositions:
H1. There is a positive correlation between a company’s IC and its performance.
H2. The higher the value of a company’s IC, the higher the company’s future
performance.
H3. There is a positive correlation between the rate-of-growth of a company’s IC
and the company’s future performance.
H4. The contribution of IC to a company’s future performance will differ by industry.
The statistical validation test for H1 is contemporaneous, namely company’s IC and financial
performance correlated with same year data of IC. Contemporaneous correlation
indicates the relevance of the information to investors (Lev and Feng, 2001). However,
returns
if the information is already priced, its value will be minimal to investors. To test
whether IC can be used to gain “abnormal returns” one must use a multi-period
predictive test (Lev and Feng, 2001). H2 is formed to test for the predictive capability of
IC. If IC is a major driver of corporate value, then logically the growth rate of IC should 81
also correlate with the increase in future performance. H3 will be tested to validate this
prediction. Although IC is seen to be crucial to the success of companies, other assets
and capabilities will also contribute to the profitability and market value of companies.
Hence, companies from different industries will have a different range of assets and
capabilities to operate their businesses and compete effectively. Some will rely more on
IC, while others will depend more on their financial or physical assets for their success.
H4 is formulated to test whether the contribution of IC differs for companies from
different industries.
The Pulic model determines that measures for the IC of a company will be VACA,
VAHU and STVA. The rate of growth of IC (ROGIC) is taken to be the year-on-year
growth rate of VACA, VAHU and STVA of the company.
Thus, for the purpose of this research, three financial ratios are selected as proxy
measures for a company’s performance. These cover return on investments, earnings
and also the shares’ performance on the stock market:
(1) Return on equity (ROE) measures how much profit a company can generate for
each dollar of shareholders’ equity. ROE is a profitability ratio relating profits
to investments. This ratio provides an indication on the earning power of
shareholders’ book investment and is frequently used when comparing two or
more firms in an industry (Van Horne, 1989, p. 129). ROE is also chosen instead
of rate of return on assets (ROA) because a company’s assets are used in
deriving VACA. Thus, to minimise possible multicollinearity, ROE is selected.
The formula to obtain the ROE is:
Profit to shareholders
ROE ¼ :
Total shareholder’s funds
(2) Earnings per share (EPS) is a commonly used measure by analysts in the
evaluation of companies in the financial market. It is also a requirement for
companies listed on the Singapore Stock Exchange (SGX) to state EPS in
companies’ annual reports. It gives a measure of profitability that incorporates
the result of operating, investing, and financing decisions (Stickney and Weil,
1997, p. 288). Also, EPS is a compulsory disclosure item in the quarterly and
annual reports for all companies listed on the SGX (Singapore Stock Exchange,
2003). It is a commonly cited item in most analysts’ reports and
recommendations in Singapore. The formula to obtain the EPS is:
Profit to shareholders
EPS ¼ :
Weighted average number of shares

(3) Annual stock return (ASR) – measures the changes in stock price inclusive of
dividends and adjusted for any stock splits. The total return from owning stock
JIC arises from two sources: dividends and other cash distributions, and capital
8,1 gains (Siegel, 2002). Thus, the formula to obtain the ASR is:

½Share price ðyear x þ 1Þ 2 Share price ðyear xÞ
þ Dividends

ASR ¼ :
Share price ðyear xÞ
82 Figure 1 shows the conceptual model for this study.

5. Methodology
The data is gathered from 150 companies publicly listed on the SGX. At the start of year
2000, there were a total of 327 companies listed on the main-board of the SGX. Table I
summarises the companies list on the main-board of the SGX at the start of 2000.
Not all companies are useable for this study, for a variety for reasons. The
performance of a company is dependent on internal factors, but is also affected by

Figure 1.
Conceptual model for the
research

Total number of Number of foreign Singapore Available

Sector companies companies companies companies

1 Multi-industry 19 1 18 13
2 Manufacturing 110 25 85 49
3 Commerce 43 8 35 23
4 Tpt/Stor/Comm 24 4 20 12
5 Finance 47 12 35 10
6 Construction 23 2 21 10
7 Properties 27 3 24 12
8 Hotels/restaurants 17 2 15 9
9 Services 14 2 12 12
10 Others 3 2 1 0
Table I. Total 327 61 266 150
Companies List on SGX
in 2000 Source: Singapore Stock Exchange (2000)
external factors that may be beyond the company’s control. To isolate the effect of IC and financial
external factors, companies to be analysed will be Singapore companies, listed on the returns
main-board of the SGX and which generate most of their revenue from the local
market. There were a total of 61 foreign companies listed on the SGX at the start of
year 2000. These were eliminated from the sample.
Also, over the three-year period from 2000 to 2002, several companies were delisted,
merged or acquired. Some companies which incurred huge losses and whose balance 83
sheets degenerated into negative net worth were eliminated. A few companies were
also suspended from trading while others did not submit their annual reports for at
least one of the three years to the SGX. Further, some companies did not register any
trading of their shares for a whole year, and so it is impossible to determine their ASR
for that year. Given these limitations and constraints, all other remaining companies
were selected which yielded a sample of 150 companies for this study.
For this research, 450 annual reports from the 150 selected companies between 2000
and 2002 were gathered from the SGX. It is argued that 150 companies is a reasonable
representation of the 327 companies listed on the SGX starting from the year 2000.
As some of the sectors are relatively small, the companies were grouped into four
related industry groups to increase the sample size of each group. The groupings have
merely brought companies in the SGX industries together. The number of companies
from each sector is shown in Table II.
A two-fold approach was adopted to the analysis of the data. Initially this involved
multiple regression using the following equation:

Y i ¼ b0 þ b1 VACA þ b2 VAHU þ b3 STVA þ m;

where Yi is the dependent variable (the dependent variables ROE, EPS and ASR were
tested sequentially using the regression model), and the independent variables are
VACA, VAHU and STVA were derived from information available in the companies’
annual report for the years 2000 and 2001.
The results from using multiple regression were inconclusive. Of the 21 multiple
regression tests conducted, only nine produced statistically significant results. The
results were statistically significant for some years but were not for others. Thus,
multiple regression was not considered adequate for this study and further analysis
was undertaken using PLS. In this, the companies’ performance was treated as a latent
variable with ROE, EPS and ASR as indicators. The model treats both IC and a
companies’ performance as latent variables each with three indicators. Multiple
regression is not able to provide this type of analysis.

Code Number of companies in each industry group Number

1 Manufacturing related (manufacturing) 49

2 Trading related (commerce and multi-industries) 36
3 Services related (services, finance, 34 Table II.
transport/storage/communications) Number of companies
4 Property related (properties, construction, hotels) 31 from each industry type
JIC PLS is a method of carrying out structural equation modelling (SEM), which, for the
8,1 present purposes, was considered more appropriate than other SEM techniques. The
small sample size, the potential lack of normally distributed variables, and the use of
formative rather than reflective indicators, all made PLS a more suitable choice of
technique than, say, maximum likelihood SEM (Anderson and Gerbing, 1988; Marsh
et al., 1988; Chin and Gopal, 1995; Chin, 1997; Cassel et al., 2000). In addition, PLS is well
84 suited for exploratory studies such as this. In studies where theories are yet to be
tightly defined, or when it is still uncertain what variables should be included in a
model or the relationships between variables, then miss-specified models will yield
inferior estimates of explained variance. PLS is less sensitive to miss-specification than
maximum likelihood SEM. Missing variables and other mis-specifications only slightly
affect the estimates made by PLS (Cassel et al., 2000).
Figure 2 shows the conceptual model using PLS. IC and companies’ performance
are latent variables and they have three indicators each. The indicators of IC are
VACA, VAHU and STVA. For companies’ performance, the indicators are ROE,
ASR and EPS.

6. Results of data analysis

6.1 One-way between groups ANOVA
Firstly, all the companies are ranked from highest to lowest values of VAICTM (i.e.
VACA þ VAHU þ STVA). They are then split into three different groups:
.
Group 1 – top 50 companies with higher VAICTM;
. Group 2 – 50 companies with middle value of VAICTM; and
.
Group 3 – bottom 50 companies with lower values of VAICTM.

A one-way between groups ANOVA was conducted on the three groups of companies
to determine whether the financial performance of companies with higher IC was
statistically different from those with lower IC for each year 2000 to 2002. A significant
result would indicate the population means were statistically not equal and companies
with higher IC had different performance measures than those of lower IC. If the
ANOVA tests does not yield any significant results for all the variables, then the data
set will have to be discarded.
6.1.1 Results for 2000. Table III shows the ANOVA results for data for the year
2000. It will be seen that the F-ratio is significant for ROE00 and ASR00 but is not
significant for EPS00 at the 0.05 level. Thus, ROE00 and ASR00 are different for

Figure 2.
Conceptual model for the
research using PLS
 IC and financial
2000 Sum of squares df Mean square F Significance
returns
ROE00 Between groups 0.905 2 0.732 38.225 0.000 *
Within groups 2.185 147 0.019
Total 3.090 149
ASR00 Between groups 3.159 2 1.580 6.106 0.003 *
Within groups 38.023 147 0.259 85
Total 41.182 149
EPS00 Between groups 0.268 2 0.021 0.195 0.823
Within groups 23.397 147 0.110 Table III.
Total 23.666 149 ANOVA result for 2000

companies with higher IC as compared to those with lower IC. For EPS00, it does not
appear to be statistically different for companies with higher or lower IC in year 2000.
6.1.2 Results for 2001. Table IV shows the ANOVA results for data from year 2001.
The table suggests that the F-ratio is significant for all the dependent variables at the
0.05 level. Thus, ASR01, EPS01 and ROE01 are different for companies with higher IC
as compared to those with lower IC in year 2001.
6.1.3 Results for 2002. Table V shows the ANOVA results for data from year 2002
and suggests that the F-ratio is significant for all the variables at the 0.05 level. Thus,
all the variables are different for companies with higher IC as compared to those with
lower IC in year 2002.
The results of the one-way between groups ANOVA show sufficient statistical
difference between companies with higher and lower IC for the dependent variables.
Thus, it is argued these variables are suitable for use in further testing.

6.2 Correlation analysis of independent variables

It is important to test whether the independent variables derived from Pulic’s method
are sufficiently independent of each other or whether there is multicollinearity between
these variables. If the independent variables are highly correlated with each other, then
they are no longer independent of each other.
Table VI shows the Pearson pair wise correlation matrix for all the independent
variables.

2001 Sum of squares df Mean square F Significance

ROE01 Between groups 16.847 2 8.424 5.088 0.007 *

Within groups 243.387 147 1.656
Total 260.234 149
ASR01 Between groups 1.720 2 0.860 10.488 0.000 *
Within groups 12.054 147 0.082
Total 13.775 149
EPS01 Between groups 3.635 2 1.817 28.750 0.000 *
Within groups 9.292 147 0.063 Table IV.
Total 12.927 149 ANOVA result for 2001
JIC
2002 Sum of squares df Mean square F Significance
8,1
ROE02 Between groups 10.631 2 5.316 10.568 0.000 *
Within groups 73.942 147 0.503
Total 84.573 149
ASR02 Between groups 2.880 2.000 1.440 11.154 0.000 *
86 Within groups 18.979 147.000 0.129
Total 21.859 149.000
EPS02 Between groups 2.445 2.000 1.223 32.080 0.000 *
Table V. Within groups 5.603 147 0.038
ANOVA result for 2002 Total 8.048 149

Tabachnick and Fidell (1996) suggest that harmful multicollinearity exists when the
bivariate correlation between independent variables is high, that is, 0.9 or above. It can
be observed from Table IV that the correlations between the independent variables are
not high. They range from a low of 0.11 to a high of 0.31 for the independent variables
of the same year. Thus, mutlicollinearity amongst the data set is minimal and does not
invalidate their use. It also indicates that Pulic’s IC measures (VACA, VAHU and
STVA) are sufficiently independent of each other.

6.3 Results of hypotheses testing

The first hypothesis posited a positive correlation between a company’s IC and its
performance. In this case, IC was tested against the companies’ performance of the
same year. Table V shows the PLS results for the various tests.
The data in Table VII shows that the T-statistics of all the paths between IC and
companies’ performance for year 2000 to 2002 are greater than 1.96. This means that
the loading is significant at the p , 0:05 level and indicates a significant relationship
between IC and companies’ performance for all three years. The R 2 is 0.514 for year
2000, 0.219 for year 2001 and 0.368 for year 2002. These are reasonably good numbers
for an exploratory study. They indicate that the model is able to explain more than 51
percent of the variance in the dependent variable for year 2000, 22 percent for year 2001
and 37 percent in year 2002, respectively. In an exploratory model, an R 2 of 0.10 can be
considered satisfactory and worth reporting (Bellman, 2003). The results appear to
support H1, namely that there is positive correlation between the IC of a company and
its performance.
The second hypothesis posited a positive correlation between the value of a
company’s IC and its future performance. In this case, IC of year x is tested against
companies’ performance of year x þ 1. This is to test whether IC is a predictor of future
companies’ performance.
Data in Table VIII shows that the T-statistics of all the paths are greater than 1.96,
which means that the loading is significant at the p , 0:05 level and indicates a
significant relationship between IC and companies’ performance for all three tests. R 2
is 0.087 for year 2000 versus 2001, 0.094 for year 2000 versus 2002, and 0.122 for year
2001 versus 2002. The R 2 values for the three tests are not very large. This indicates
that the model is able to explain about 9 percent of the variance in the dependent
variable for year 2000 versus 2001, over 9 percent for year 2000 versus 2002 and more
 Correlations VACA02 VAHU02 STVA 02 VACA01 VAHU01 STVA 01 VACA00 VAHU00 STVA 00

VACA02 Pearson 1.00 0.31 0.20 0.76 0.08 0.07 0.56 20.04 0.14
Significance (one-tailed) 0.00 * 0.01 * 0.00 * 0.16 0.20 0.00 * 0.30 0.05
VAHU02 Pearson 0.31 1.00 0.20 0.12 0.53 0.07 0.17 0.62 0.18
Significance (one-tailed)) 0.00 * 0.01 * 0.07 0.00 * 0.21 0.02 * 0.00 * 0.01 *
STVA02 Pearson 0.20 0.20 1.00 0.14 0.12 0.11 0.07 0.08 0.10
Significance (one-tailed) 0.01 * 0.01 * 0.04 * 0.07 0.09 0.21 0.18 0.12
VACA01 Pearson 0.76 0.12 0.14 1.00 0.17 0.21 0.63 20.03 0.12
Significance (one-tailed) 0.00 * 0.07 0.04 * 0.02 * 0.01 * 0.00 * 0.36 0.07
VAHU01 Pearson 0.08 0.53 0.12 0.17 1.00 0.16 0.04 0.81 0.16
Significance (one-tailed) 0.16 0.00 * 0.07 0.02 * 0.02 * 0.33 0.00 * 0.03 *
STVA01 Pearson 0.07 0.07 0.11 0.21 0.16 1.00 0.05 0.04 0.01
Significance (one-tailed) 0.20 0.21 0.09 0.01 * 0.02 * 0.28 0.33 0.44
VACA00 Pearson 0.56 0.17 0.07 0.63 0.04 0.05 1.00 0.11 0.29
Significance (one-tailed) 0.00 * 0.02 * 0.21 0.00 * 0.33 0.28 0.09 0.00 *
VAHU00 Pearson 20.04 0.62 0.08 20.03 0.81 0.04 0.11 1.00 0.24
Significance (one-tailed) 0.30 0.00 * 0.18 0.36 0.00 * 0.33 0.09 0.00 *
STVA00 Pearson 0.14 0.18 0.10 0.12 0.16 0.01 0.29 0.24 1.00
Significance (one-tailed) 0.05 * 0.01 * 0.12 0.07 0.03 * 0.44 0.00 * 0.00 *
Note: *Correlation is significant at the 0.05 level or better (one-tailed)
IC and financial
returns

variables
Correlation matrix for the
87

Table VI.
JIC
Model summary IC 2000 versus 2000 performance
8,1
R2 0.514
Regression weights of path 0.717
T-statistic of path 10.9801

88 IC 2001 versus 2001 Performance

R2 0.219
Regression weights of path 0.468
T-statistic of path 5.9495

IC 2002 versus 2002 performance

R2 0.368
Table VII. Regression weights of path 0.607
PLS results of H1 T-statistic of path 7.1073

Model summary 2000 versus 2001

2
R 0.087
Regression weights of path 0.295
T-statistic of path 3.5594

2000 versus 2002

R2 0.094
Regression weights of path 0.306
T-statistic of path 5.2060

2001 versus 2002

R2 0.122
Table VIII. Regression weights of path 0.349
PLS results of H2 T-statistic of path 6.2780

than 12 percent for year 2001 versus 2002, respectively. Although the R 2 numbers are
not very large, nevertheless, the paths are statistically significant at p , 0:05.
The results lend modest support to H2 that the higher the value of a company’s IC,
the higher the company’s future performance.
The third hypothesis posits a positive correlation between the rate-of-growth of a
company’s IC and the company’s future performance. The rate of growth (ROG) of IC is
the year-on-year growth rate of VACA, VAHU and STVA of a company. The rate of
growth of IC from year 2000 to 2001 and year 2001 and 2002 is tested against the
companies’ performance of 2002.
Data in Table IX shows that the T-statistics of both paths are greater than 1.96,
which means that the loading is significant at the p , 0:05 level and indicates a
significant relationships between IC and companies’ performance for the two tests. R 2
is 0.149 for ROG 2001-2000, which indicates that the model is able to explain about 15
percent of the variance in the dependent variable. R 2 is 0.066 for ROG 2002-2001,
which indicates that the model is able to explain nearly 7 percent of the variance in the IC and financial
dependent variable. Although the all R 2 values are not very large, the path is returns
statistically significant.
If the higher the value of a company’s IC the higher is the company’s future
performance, then logically, the rate of growth of IC will also correlate with future
performance. The results of H3 further reaffirm the findings of H2 and help to
reinforce the contribution of IC to companies’ performance. The results further 89
reinforce proponents of IC as a competitive tool and that companies must manage and
grow their IC to remain competitive (Nonaka, 1995; Bontis, 1998; Brennan and Connell,
2000; Hurwitz et al., 2002).
The fourth hypothesis posits that the contribution of IC to a company’s future
performance will differ by industry. To test this hypothesis the data were broken into
four different industry groups. As some of the sectors were relatively small, they were
then further grouped into four related industry types to increase the sample size. The
four industry groups were:
(1) Manufacturing related – These companies come from the manufacturing sector
of the SGX and are primarily engaged in the production and manufacturing of
goods.
(2) Trading related – These companies come from the commerce and
multi-industries sectors of the SGX. Commerce sector comprises companies
that have businesses mainly in retail and trading, and includes holding
companies with subsidiaries in several industries.
(3) Services related – These companies come from the services, finance and
transport/storage/communications sectors of the SGX. These companies are
primarily engaged in the services industry.
(4) Property related – These companies come from the property, construction and
hotel sectors of the SGX. They rely heavily on property infrastructure and
physical buildings for the conduct of their businesses.

Data in Table X shows that the T-statistics of the paths are greater than 1.96, except for
the Trading sector for year 2000 versus 2001 and year 2000 versus 2002. This suggests a
significant relationship between IC and companies’ performance for all sectors except for
the Trading sector for year 2000 versus 2001 and year 2000 versus 2002. R 2 figures are
0.252, 0.171 and 0.394 for the Manufacturing sector in the three tests. For the Trading
sector, R 2 figures are 0.088, 0.327 and 0.179. On the other hand, R 2 figures are larger for
both the Services and Property sectors. R 2 figures for the Services sector are 0.278, 0.726
and 0.389 while they are 0.398, 0.323 and 0.496 for the Property sector.

Model summary ROG 2001-2000 versus 2002 ROG 2002-2001 versus 2002
2
R 0.149 0.066
Regression weights of path 0.386 0.257 Table IX.
T-statistic of path 2.0305 2.1473 PLS results of H3
JIC
Model summary – Manufacturing 2000 versus 2001 2000 versus 2002 2001 versus 2002
8,1 R2 0.252 0.171 0.394
Regression weights of path 0.502 0.413 0.628
T-statistic of path 4.5615 2.8721 5.6100

Model summary – Trading 2000 versus 2001 2000 versus 2002 2001 versus 2002
90 R2 0.088 0.327 0.179
Regression weights of path 0.296 0.571 0.423
T-statistic of path 0.7241 1.6230 3.5805

Model summary – Services 2000 versus 2001 2000 versus 2002 2001 versus 2002
R2 0.278 0.726 0.389
Regression weights of path 0.527 0.852 0.624
T-statistic of path 4.2120 3.9704 2.1345

Model summary – Property 2000 versus 2002 2000 versus 2002 2001 versus 2002
R2 0.398 0.323 0.496
Table X. Regression weights of path 0.631 0.568 0.704
PLS results of H4 T-statistic of path 2.3180 4.3676 4.0309

The central tenet of this paper is to provide evidence that acquiring and applying
knowledge will become a competitive factor leading to above average financial returns.
Companies, especially those in knowledge-intense industries, need to know the
importance of IC, and that knowledge is becoming a critical factor affecting a company’s
ability to remain competitive in the new global marketplace. Companies in the trading
group that rely mainly on buying and selling product are expected to be less dependent
on the application of knowledge and the result supports such a prediction.
The results indicate a higher contribution of IC to companies in the Services and
Property sector, less in the Manufacturing sector and even less in the Trading sector.
The results of H4 appear to support the concept proposed by Treacy and Wiersema
(1995). They argue that although IC is seen to be crucial to the success of companies,
other assets and capabilities will also contribute to the profitability and market value
of companies. Hence, companies from different industries will have a different range of
assets and capabilities to operate their businesses and compete effectively. Some will
rely more on IC, while others will depend more on their financial or physical assets for
their success.

7. Summary and conclusions

Drawing upon a sample of 150 publicly listed companies on the Singapore Exchange,
this paper has examined the relationship between IC and company performance. In
doing so, it has examined four aspects of the relationship:
(1) a positive correlation between a company’s IC and its performance;
(2) a positive relationship between increased value of a company’s IC and that
company’s future performance;
(3) a positive correlation between the rate of growth of a company’s IC and that
company’s future performance; and
 (4) that the contribution of IC to company performance will differ from industry to IC and financial
industry. returns
There are a number of classifications and measures of IC. For the purposes of this
paper the Pulic model was used. In this, the Value Added Intellectual Coefficient
(VAICe) is used to measure the IC of companies. This method is designed to provide
information about the value creation efficiency of tangible and intangible assets within 91
a company during operations. In this model, the value added intellectual coefficient
(VAICe) is the sum of three other coefficients:
(1) physical capital coefficient (VACA);
(2) the human capital coefficient (VAHU); and
(3) the structural capital coefficient (STVA).

The method has the attraction of ease of data acquisition and the development of ratios
from standard financial data available in company annual reports.
The analysis was undertaken using partial least squares (PLS), since considerations
of sample size, the use of formative rather than reflective indicators, the distribution of
variables, and the exploratory nature of the study all suggested PLS as the most
appropriate methodology. The analysis reveals general support for each of the
propositions. Thus, the results of the study support the notion that companies that
actively nurture and increase their IC are likely to experience superior performance.
The study supports those theories and practices that place value on IC. The paper
would also suggest that the Pulic model provides companies with a simple approach to
measuring their IC.
This research is not without its limitations. The selected companies are
confined to Singaporean companies listed on the SGX. Different countries have
different accounting practices and stock exchanges around the world have different
disclosure and other listing requirements. As the Pulic model uses data from the
published financial statements any differences in national accounting rules may
influence results in other countries. However, the primary use of reported financial
results limits the impact of cultural and social differences on the results. Moreover,
the move to adopt International Financial Reporting Standards in many regions,
including Singapore, Australia and the European Union, will mitigate differences
in accounting rules. This research is limited to Singaporean companies and the
SGX using the accounting rules applicable during the research period. It is also
limited to publicly listed companies as financial information for privately held
companies is not readily available. The shares of private companies are not traded
freely and are not subjected to market forces. Hence, their market values are not
easily or reliably determined. Also, the selected companies were analysed over a
three-year period between 2000 and 2002. Data from earlier years cannot be used
because there was no requirement to disclose manpower costs prior to the year
2000. Manpower costs are needed in Pulic’s model for the calculation of VAHU. It
is also limited to three years of data as information for financial year 2003 was
not available at the completion of this study due to the late publication of results
by many of the selected companies.
In summary, the results of this study provide additional empirical evidence on the
contribution of IC to companies’ current and future performance. Thus, how well
JIC companies do in acquiring and applying knowledge will become a competitive factor.
8,1 Managers, especially those in knowledge-intense industries, need to know the
importance of IC and that knowledge is a critical factor affecting a company’s ability to
remain competitive in the new global marketplace.

92
References
Accounting Principles Board (1970), Intangible Assets, APB Opinion 17, American Institute of
Certified Public Accountants, New York, NY.
Accounting Standards Board (1997), Goodwill and Intangible Assets, FRS 10, Accounting
Standards Board, London.
Anderson, J.C. and Gerbing, D.W. (1988), “Structural equation modelling in practice: a review and
recommended two-step approach”, Psychological Bulletin, Vol. 103 No. 3, pp. 411-23.
Andriessen, D. (2001), “Weightless wealth: four modifications to standard IC theory”, Journal of
Intellectual Capital, Vol. 2 No. 3, pp. 204-14.
Andriessen, D. and Tissen, R. (2000), Weightless Wealth: Find Your Real Value in a Future of
Intangible Assets, Financial Times Prentice-Hall, London.
Barsky, N.P. and Marchant, G. (2000), “The most valuable resource – measuring and managing
intellectual capital”, Strategic Finance, Vol. 81 No. 8, pp. 58-62.
Bellman, S. (2003), “How to use PLS”, Graduate School of Management, University of Western
Australia, Crawley, July.
Bontis, N. (1998), “Intellectual capital: an exploratory study that develops measures and models”,
Management Decision, Vol. 36 No. 2, p. 63.
Bontis, N. (2001), “Assessing knowledge assets: a review of the models used to measure
intellectual capital”, International Journal of Technology Management, Vol. 3 No. 1,
pp. 41-60.
Bontis, N., Dragonetti, N.C., Jacobsen, K. and Roos, G. (1999), “The knowledge toolbox: a review
of the tools available to measure and manage intangible resources”, European
Management Journal, Vol. 17, pp. 391-401.
Bornemann, M. and Leitner, K.H. (2002), “Measuring and reporting intellectual capital: the case of
a research technology organisation”, Singapore Management Review, Vol. 24 No. 3,
pp. 7-19.
Brennan, N. and Connell, B. (2000), “Intellectual capital: current issues and policy implications”,
Journal of Intellectual Capital, Vol. 1 No. 3, pp. 206-40.
Brooking, A. (1996), Intellectual Capital, Thomas Business Press, London.
Cassel, C.M., Hackl, P. and Westlund, A.H. (2000), “On measurement of intangible assets: a study
of robustness of partial least squares”, Total Quality Management, Vol. 11 No. 7,
pp. S897-S907.
Chin, W.W. (1997), “Overview of the PLS method”, University of Houston, Houston, TX,
available at: http://disc-nt.cba.uh.edu/chin/PLSINTRO.htm
Chin, W.W. and Gopal, A. (1995), “Adoption intention in GSS: relative importance of beliefs”,
Data Base, Vol. 26 Nos 2/3, pp. 42-63.
Danish Confederation of Trade Unions (1999), “Your knowledge – can you book it?”, paper
presented at the International Symposium Measuring and Reporting Intellectual Capital:
Experiences, Issues, and Prospects, OECD, Amsterdam, June.
Dzinkowski, R. (2000), “The measurement and management of intellectual capital: an introduction”, IC and financial
Management Accounting, Vol. 72 No. 2, pp. 32-6.
returns
Edvinsson, L. and Malone, M. (1997), Intellectual Capital: Realizing Your Company’s True Value
by Finding Its Hidden Brainpower, HarperCollins, New York, NY.
Firer, S. and Williams, S.M. (2003), “Intellectual capital and traditional measures of corporate
performance”, Journal of Intellectual Capital, Vol. 4 No. 3, p. 348.
Guthrie, J. (2001), “The management, measurement and the reporting of intellectual capital”, 93
Journal of Intellectual Capital, Vol. 2 No. 1, p. 27.
Guthrie, J., Petty, R., Ferrier, F. and Well, R. (1999), “There is no accounting for intellectual capital
in Australia: review of annual reporting practices and the internal measurement of
intangibles within Australian organisations”, paper presented at the International
Symposium Measuring and Reporting Intellectual Capital: Experiences, Issues and
Prospects, OECD, Amsterdam, June.
Haanes, K. and Lowendahl, B. (1997), “The unit of activity: towards an alternative to the theories
of the firm”, in Thomas, H. (Ed.), Strategy, Structure and Style, Wiley, Copenhagen.
Hurwitz, J., Lines, S., Montgomery, B. and Schmidt, J. (2002), “The linkage between management
practices, intangibles performance and stock returns”, Journal of Intellectual Capital, Vol. 3
No. 1, pp. 51-61.
International Accounting Standards Board (2004), Intangible Assets, IAS 38, International
Accounting Standards Board, London.
International Accounting Standards Committee (1998), Intangible Assets, IASC 38, London.
International Federation of Accountants (1998), The Measurement and Management of
Intellectual Capital, International Federation of Accountants, New York, NY.
Johanson, U. and Grojer, J.E. (1998), “Current development in human resource costing and
accounting: reality present, researchers absent?”, Accounting, Auditing & Accountability
Journal, Vol. 11 No. 4, pp. 495-504.
Joia, L.A. (2000), “Measuring intangible corporate assets linking business strategy with
intellectual capital”, Journal of Intellectual Capital, Vol. 1 No. 1, pp. 68-84.
Kaplan, R.S. and Norton, D.P. (1992), “The balanced scorecard – measures that drive
performance”, Harvard Business Review, Vol. 70 No. 1, pp. 71-9.
Leliaert, P.J.C., Candries, W. and Tilmans, R. (2003), “Identifying and managing IC: a new
classification”, Journal of Intellectual Capital, Vol. 4 No. 2, p. 202.
Lev, B. and Feng, G. (2001), “Intangible assets: measurement, drivers, usefulness”, New York
University, New York, NY.
Lowendahl, B. (1997), Strategic Management of Professional Service Firms, Handelshojskolens
Forlag, Copenhagen.
Luthy, D.H. (1998), “Intellectual capital and its measurement”, Utah State University, Logan, UT,
available at: www3.bus.osaka-cu.ac.jp/apira98/archives/htmls/25.htm.
McLean, R. (1999), “Accounting for the creation of value”, available at: http://cpri.matrixlinks.ca/
tvc/Presentations/TVCPresent/index.htm
Marsh, H.W., Balla, J.R. and McDonald, R.P. (1988), “Goodness-of-fit indexes in confirmatory
factor analysis: the effect of sample size”, Psychological Bulletin, Vol. 103 No. 3,
pp. 391-410.
MERITUM (2001), “Measuring Intangibles to Understand and Improve Innovation
Management”, Final Report, European Community, available at: www.kunne.no/meritum
JIC Nash, H.H. (1998), “Accounting for the future”, Prospective Accounting Initiative, available at:
http://home.sprintmail.com/ , humphreynash/indexback.htm
8,1
Nonaka, I. (1995), The Knowledge Creating Company, Harvard Business Review on Knowledge
Management, Harvard Business School Press, Boston, MA, pp. 21-45.
Petrash, G. (1996), “Dow’s journey to a knowledge value management culture”, European
Management Journal, Vol. 14 No. 4, pp. 365-73.
94 Petty, P. and Guthrie, J. (2000), “Intellectual capital literature review: measurement, reporting and
management”, Journal of Intellectual Capital, Vol. 1 No. 2, pp. 155-75.
Pulic, A. (1998), “Measuring the performance of intellectual potential in knowledge economy”,
available at: www.measuring-ip.at/OPapers/Pulic/Vaictxt/vaictxt.html
Pulic, A. (2000), “VAICe – an accounting tool for IC management”, available at: www.
measuring-ip.at/Papers/ham99txt.htm
Roos, J., Roos, G., Dragonetti, N.C. and Edvinsson, L. (1997), Intellectual Capital: Navigating the
New Business Landscape, Macmillan Press, London.
Siegel, J.J. (2002), “Stock prices”, The Concise Encyclopedia of Economics, Liberty Fund, Inc.,
available at: www.econlib.org/library/Enc/StockPrices.html
Singapore Stock Exchange (2000), “Market capitalisation as at month end”, Pulses, January, p. 80.
Singapore Stock Exchange (2003), SGX Securities Trading Listing Manual, Appendix 7.2,
Financial Statements and Dividend Announcement, para. 6, Singapore Stock Exchange,
Singapore.
Society of Management Accountants of Canada (1998), The Management of Intellectual Capital:
The Issues and the Practice, Issue 16, Society of Management Accountants of Canada,
Hamilton.
Stewart, T.A. (1997), Intellectual Capital: The Wealth of New Organisations, Nicholas Brealey
Publishing, London.
Stickney, C.P. and Weil, R.L. (1997), Financial Accounting: An Introduction to Concepts, Methods,
and Rules, Dryden Press, Hinsdale, IL.
Sveiby, K.E. (1997), The New Organisational Wealth: Managing and Measuring Knowledge
Based Assets, Berrett-Koehler, San Francisco, CA.
Sveiby, K.E. (1998), “Measuring intangibles and intellectual capital – an emerging first
standard”, PhD thesis, Australian Centre for Strategic Management, Queensland
University of Technology, Brisbane.
Tabachnick, B.G. and Fidell, L.S. (1996), Using Multivariate Statistics, HarperCollins Publishers,
New York, NY.
Thurow, L.C. (1999), Building Wealth, HarperCollins, New York, NY.
Treacy, M. and Wiersema, F. (1995), The Discipline of Market Leaders, Addison-Wesley,
Reading, MA.
Van Horne, J.C. (1989), Fundamentals of Financial Management, Prentice-Hall International,
Englewood Cliffs, NJ.
VanderKaay, S. (2000), “Measuring the vital signs of intellectual capital”, CMA Management,
Vol. 74 No. 4, pp. 18-21.

About the authors

Hong Pew Tan is a graduate of UWA’s Doctor of Business Administration program. He has
many years working experience in senior management positions and is currently the Chief
Operating Officer of Nera Telecommunications Ltd, a company publicly listed on Singapore IC and financial
Exchange. Hong Pew completed a MSc from NUS, and a BSc(Hons) from the UNSW.
David Plowman has published widely in areas of human resource management and labour returns
relations. He is Director of the GSM’s doctoral programs.
Phil Hancock is Associate Dean of Teaching and Learning in the Faculty of Economics and
Commerce at UWA. He has many years experience as a researcher and has published widely in
the accounting discipline. He teaches in the Graduate School of Management and is an active
member of both professional accounting bodies in Australia. Phil Hancock is the corresponding 95
author and can be contacted at: Phil.Hancock@uwa.edu.au

To purchase reprints of this article please e-mail: reprints@emeraldinsight.com

Or visit our web site for further details: www.emeraldinsight.com/reprints