Cost of Capital, Optimal Capital Structure, and Value of Firm: An
Empirical Study of Indian Companies
Raj S Dhankar and A jit S Boora
Academicians and practitioners alike have
found it difficult to resolve the issue of
optimal capital structure in the perfect
capital markets of the West as well as in the
imperfect capital markets, as in India. This
paper examines whether there exists an
optimal capital structure in Indian
companies, both at the micro and the macro
level and whether financing decisions affect
the value of a firm.
Raj S Dhankar is an Associate Professor,
Faculty of Management Studies, University
of Delhi and Ajit S Boora is an Associate
Professor, SBM Institute of Management,
Rohtak.
Vol. 21, No. 3, July-September 1996
Introduction
During the last 50 years or so, the role of financial
management has undergone a tremendous change. The
ownership structure, size of business firms, security
markets, financial system and instruments have greatly
changed. As a result, the role of a finance manager has
become far more important than merely a fund raiser.
The finance manager is expected to maximize the
economic welfare of the owners, which is represented by
the market value of the firm. To achieve this objective,
one has to take a number of decisions, the most important
being the investment, financing and dividend decisions.
Do changes in capital structure affect the value of a
firm? This question has been puzzling the minds of both
the finance managers and academicians for the last 40
years, especially since the publication of the path breaking
articles by Franco Modigliani and Merton Miller. In a
perfect capital market, their irrelevancy model is perfectly
valid and is supported by all. But, in case of an imperfect
market, the views differ greatly and, as a result, till date,
no universally accepted model has been developed on
this crucial issue. In India, no significant work has been
done in this regard. This paper makes a humble attempt
to empirically test whether there exists an optimal capital
structure in Indian companies, both at the micro and the
macro level. The paper also tries to examine whether the
financing decisions affect the value of a firm.
The paper is divided into four sections. Section 1
deals with the review of literature and objectives of the
study; section 2 explains the research methodology;
section 3 shows the empirical results and section 4 gives
the conclusions.
Review of Literature
Sound financing decisions of a firm basically should lead
to an optimal capital structure. Capital structure repres-
ents the proportion in which various long-term capital
components are employed. Over the years, these decis-
ions have been recognized as the most important
decisions that a firm has to take. This is because of the
fact that capital structure affects the cost of capital, net
profit, earning per share, dividend payout ratio and
liquidity position of the firm. These variables coupled
with a number of other factors determine the value of a
29
firm. So, capital structure is a very important determi-
nant of the value of a firm.
Franco Modigliani and Merton Miller (hereafter
called M -M) were the first to present a formal model on
valuation of capital structure. In their seminal papers
(1958,1963), they showed that under the assumptions of
perfect capital markets, equivalent risk class, no taxes,
100 per cent dividend-payout ratio and constant cost of
debt, the value of a firm is independent of its capital
structure. When corporate taxes are taken into account,
the value of a firm increases linearly with debt-equity
(D/E) ratio because of interest payments being tax
exempted. M-M'S work has been at the centrestage of
the financial research till date. Their models have been
criticized, supported, and extended over the last 35 years.
David Durand (1963) criticized the model on the
ground that the assumptions used by M-M are unrealistic.
Solomon (1963) argued that the cost of debt does not
always remain constant. When the leverage level exceeds
the accepted level, the probability of default in interest
payments increases thus raising the cost of debt. Stiglitz
(1969,1974) proved the validity of the M-M model under
relaxed assumptions whereas Smith (1972), Krause and
Litzenberger (1973), Baron (1974,1975), and Scott (1976,
1977), supported the M-M model, but only under the
conditions of risk free debt and costless bankruptcy.
When bankruptcy has positive costs, there exists an
optimal capital structure which is a trade-off between
tax advantage of debt and bankruptcy costs.
This trade-off theory was challengedby Miller (1977).
He argued that bankruptcy and agency costs are too
small to offset the tax advantage of debt. But when
personal taxes are taken into account, this advantage is
completely offset by the disadvantage of personal tax
rnte. Thus, in equilibrium, the value of a firm is
independent of its capital structure, even when the
market is imperfect. But Miller's model was rejected by
DeAngelo and Masulis (1980). They argued that even if
bankruptcy, agency and related costs are ignored,
introduction of non-debt tax shields is enough for a firm
to have an optimal capital structure. And even if these This model clearly advocates the existence of an
costs are taken into account, an optimal capital structure optimal capital structure which is a trade-off between
exists, irrespective of availability of non-debt tax shields. tax advantages and disadvantages of leverage. However,
Masulis (1980,1983), Brennen and Schwartz (1978), and M-M (1966) again proved their irrelevancy hypotheseis.
Jensen and Meckling (1976) also advocated the existence But the study by Davenport (1971) supported the tradi-
of an optimal capital structure in an imperfect market, tional view.
while using different mechanisms. Besides, a lot more In the Indian context, one comes across two works,
work has been done on this problem till now, but a one by Sharma and Rao (1969) and the other by Pandey
formal model, showing the mechanism for determining (1992). The former tested the M—M model using cross-
an optimal capital structure in an imperfect market, is sectional analysis for engineering companies, wherein
yet to be developed. the value of a firm was found to be independent of its
On the basis of the major work done by M-M and capital structure after allowing for tax advantage. But
the results could not be generalized as the sample was
30
Vikalpa
homogeneous. The other work by Pandey (1992) obser- Number of shares = Total equity capital/Face
ved that the M-M theory is not fully valid under Indian value per share
conditions. He concluded that, initially, cost of capital
and value of a firm are independent of the capital Average price = Mean value of monthly high
structure changes, but they rise after a certain level. All and low price/share during
these studies have helped understand the dynamics of the accounting year.
this crucial issue better but have not been able to come to 3) Value of firm (V):
a definite conclusion as to how firms determine their
optimal capital structure. So, the present study was V = Market value of equity + Book value of
planned to make another attempt to resolve this preference share and debt1.
contentious issue. It may be pointed out that the study 4) Cost of capital (Ka):
has not included the effect of factors like agency and
bankruptcy costs, as they are difficult to measure in the Kp.P + Kd.D + Ke.E
Indian scenario. Ka =
Objectives P+D+E
The basic objective of this study is to find out if there where,
exists an optimal capital structure either at the micro Kp = Cost of preference capital2
and/or at the macro level in Indian private sector
companies. If yes, how do companies determine it? In Kd = Cost of debt capital3
addition, it has the following sub-objectives: Does change Ke = Cost of equity capital4
in capital structure affect the cost of capital? Are capital
structure and dividend policy correlated? Does capital P, D = Book value of preference and debt capital
structure of companies differ significantly? respectively
Research Methodology E = Market value of common equity capital
For conducting the study, a sample of 26 widely held wherein, Ke = Rf + B (Rm-Rf)
Indian private sector companies from top 300 large scale where,
companies was taken. The sample is not homogeneous
as the companies are taken from 15 different industry Rf = Risk free rate5
groups. It is a multi-period study covering the period B = Beta coefficient6
from 1981-82 to 1990-91. The study used both primary
and secondary data. The main source of secondary data Cov (Rj,Rm)
was the Bombay Stock Exchange Directory. Primary data
were collected through a mailed questionnaire. In all, Var (Rm)
100 questionnaires were mailed but the response rate where,
was about 26 per cent. Hence, a sample of 26 companies
was taken. Most of the respondents were above the level Rm = Return on market portfolio
of Finance Manager. The responses were ranked on the Rj = Return on common stock
basis of relative weightage given to various factors by
the respondents. Var (Rm) = Variance of return on market portfolio
Measure of Variables
The variables used in the study are as follows: 1. Preference share and debt are taken at book values
1) Capital structure (C): because there is no significant yearly fluctuations in the
prices of these sources of capital.
C = Sum of the book values of various components of
capital structure 2. Preference dividend/price of preference share.
2) Market value of equity (E): 3. Interest (1 - T)/Total debt.
E = Number of shares outstanding at the end of an 4. Calculated by using Capital Asset Pricing Model.
accounting year X Average market price per Frequency of variables is on an annual basis.
share 5. Risk free rate is taken as 10 per cent.
where, 6. It represents the systematic risk associated with "a
security.

Vol. 21, No. 3, July-September 1996 31

;i) Leverage (L):
L = Debt to total capital ratio, i.e. L = D/ (D+E) = D/C 8 DCM 0.668 0.091
6) Dividend payout ratio (D/P): 9 Escorts 0.477 -0.015
Equity dividend paid 10 Godfrey Philips 0.488 0.226 -
D/P = 11 Goodyear Grasim 0.283 0.682 **
Net profit - Preference dividend 12 Industries 0.431 0.746 *
13 HEG Ltd. 0.410 0.838 *
In the absence of a well developed model on capital
structure problem, Karl Pearson's bivariate correlation 14 ICI Ltd 0.494 -0.137
coefficient was used to find the relationship among the 15 ICICI 0.916 -0.467
variables. The value of r for the pooled data represents .16 Kelvinator 0.534 -0.206
a correlation between average change in D/C ratio 17 Kinetic Engg. 0.438 0.079 -
(change in capital structure) on the one hand, and total 18 Kirloskar Brothers 0.684 0.375
market value of firms, average Ka and average D/P 19 Mukand Ltd 0.563 0.131
ratio, respectively, on the other (Tables 1,4,5). Considering
20 Nestle 0.384 0.459
the nature and objectives of the study, we thought it
proper to use the bivariate correlation technique. As a 21 Ranbaxy Laboratories 0.688 -0.239
matter of fact, most of the earlier studies in this area have 22 Rathi Alloys Ltd SPIC 0.714 0.280 -
used a similar technique. The correlation results were 23 0.569 0.298
tested by using student's t-test; f-test has been used to 24 SRFLtd 0.437 0.273
test the significance of difference in inter-company capi- 25 TELCO 0.583 -0.626**
tal structure. 26 TISCO 0.512 -0.796 *
Empirical Results Pooled data 0.666# 0.706**
Optimal Capital Structure : There is no definite * and ** represent the significance of results at 1 per cent
relationship between change in the capital structure and and 5 per cent level respectively.
the value of a firm (Table 1). Out of 26, only 6 companies
show statistically significant relationship (3 showing # 0.666 is the weighted average D/C ratio of all the
negative and 3 showing positive values). Further, none companies for ten years.
of the sample companies was found to be highly levera- Table 2 : Determinants of Share Prices
ged. Except for ICICI, the maximum value of D/C ratio
for a company was 0.714. For ICICI, it was 0.916 which Rank Determinant
is normal for a financial institution. These insignificant
and inconsistent results at the micro level can be attributed Management of the company
to the fact that capital structure is not the only determinant
of market price of a company's share and its value and Dividend policy
that there are other factors as well which do affect their Role of bulls and bears Capital structure
values. Incidentally, this was confirmed during the
course of our discussions with company executives Government policies Takeover bid by others
(Tables 2 and 3). Cost of capital
Table 1: Coefficient of Bivariate Correlation (r) bet- Table 3 : Other Factors which Affect the Value of a
ween Change in Capital Structure and Valu e Firm
of the Firm
Rank Factor
Nflme o/ the Company D/C Ratio r
1 Operating results
1 Amrit Banaspati Co. 0.536 0.348
2 Business risk
Apollo Tyres 0.765 -0.146
3 Asea Brown Boveri 0.580 -0.307 2 Economic conditions
4 Atlas Cycle Industries 0.582 0.014 Promoters
5 Ballarpur Industries 0.358 0.049 3 Tax rates and structures
6 Bata India 0.474 0.903 * Political conditions
Blue Star 0.446 -0.052 4
7
Vikalpa
32 5

6
 Since a majority of the factors are non-measurable as since the cost of capital is measured using historical data,
they are qualitative in nature, it is not possible to segregate the weighted average" cost of capital is bound to go down
their effects. Therefore, an exact relationship between with increasing debt, other things being equal. But the
capital structure and value of a firm cannot be establi- relationship is not statistically significant. It means that,
shed. This conclusion is further strengthened by the in general, changes in capital structure are not acco-
highly volatile behaviour of the stock markets. One ob- mpanied by proportionate changes in cost of capital.
serves that 22 companies showed coefficient of variance The statistically insignificant values of r'can be explai-
(C V) greater than 50 per cent. The average C V for all the ned by the following reasons:
companies was 62.11 per cent which is quite high. Fur-
ther, in general, variations in values were not found to * Indian companies have no specific model or
be significantly associated with the financial performan- mechanism to compute the specific costs of capital,
ce of the companies. So, the market price of the stock is particularly the cost of equity capital and the average
not the true index of a company's performance. The va- cost of capital. This became clear, when ,in response
riations in stock prices of a company may also represent to the question, "how do you compute the cost of
the effects of qualitative factors. These factors can capital?", none of the companies suggested a defi
undervalue or overvalue these prices at the micro level nite/specific mechanism.
as a result of which they may not be the true1 indices of a * Cost of capital is not the only determinant of the
company's performance. capital structure, though it is one of the most
However, at the macro level, the relationship bet- important determinants. Since the effects of other
ween change in capital structure and value of a firm was determinants cannot be segregated, so, an exact
found to be highly positive and statistically significant relationship between the two could not be
(r = 0.706) (Table 1). These results are totally different established.
from those obtained at the micro level. The reason is * Cost of debt in India is quite high as compared to
quite clear, because when we take the aggregate figures, that in the developed countries.
the positive and negative effects of the external and
qualitative factors on individual shares neutralize one Table 4: Coefficient of Bivariate Correlation (r)
another and we get a closely approximate true value of between Change in Capital Structure and
the share prices. That is why we get a highly positive Cost of Capital
correlation between the two at the macro level.
Name of the Company r
The above results clearly show that in imperfect
market, at the macro level, an optimal capital structure 1 Amrit Banaspati Co. -0.217
definitely does not exist. It is not possible to determine it 2 Apollo Tyres Asea 0.621**
exactly, because of the difficulty faced in the measure- 3 Brown Boveri 0.095
ment of qualitative factors and other problems. These 4 Atlas Cycle Industries -0.550**
results fully support the views expressed by Brigham 5 Ballarpur Industries -0.108
and Gopanski (1985) when they observed, "Unfortuna- 6 Bata India -0.325
tely, it is almost impossible to test the leverage effects 7 Blue Star -0.314
empirically, because (1) future earnings are impossible 8 DCM 0.069
to measure and (2) most real world leverage changes are 9 Escorts -0.166
accompanied by asset changes which may be changing 10 Godfrey Philips -0.100
the firm's risk class. Thus, empirical tests have not pro- 11 Goodyear Grasim -0.131
duced conclusive results. However, the evidence does 12 Industries -0.156
generally support the contention that some benefits 13 HEGLtd -0.177
from leverage do exist, at least if the firm does not exceed 14 ICI Ltd -0.239
reasonable limits of debt." 15 ICICI -0.148
16 Kelvinator -0.543
Capital Structure and Cost of Capital : In general, 17 Kinetic Engg. 0.043
change in capital structure and cost of capital were found 18 Kirloskar Brothers -0.227
to be negatively related, as 81 per cent of the companies 19 Mukand Ltd. 0.375
showed negative relationship (Table 4), thus, supporting 20 Nestle -0.436
the theory that cost of capital decreases with increase in -0.256
21 Ranbaxy Laboratories
debt level because cost of debt is less than that of equity -0.117
22 Rathi Alloys Ltd.
and interest payments are tax exempted. Furthermore, -0.411
23 SPIC

Vol. 21, No. 3, July-September 1996 33

24 SRF Ltd. -0.344 Table 5 : Coefficient of Bivariate Correlation (r) bet-
25 TELCO -0.176 ween Change in Capital Structure and
26 TISCO -0.216 Dividend Payout Ratio
Pooled data -0.245 Name of the Company r
* and ** represent the significance of results at 1 per cent 1 Amrit Banaspati Co. 0.081
and 5 per cent level, respectively. 2 Apollo Tyres 0.756*
Capital Structure and Dividend Policy : At the micro 3 Asea Brown Boveri -0.110
level, no definite and consistent relationship exists 4 Atlas Cycle Industries 0.074
between change in capital and dividend policy (Table 5),
as 14 companies showed positive values of r whereas 12 5 Ballarpur Industries 0.057
companies showed negative values. The relationship 6 Bata India -0.089
was statistically significant in case of 3 companies only. 7 Blue Star 0.366
These inconsistent results could be attributed to two 8 DCM -0.512
main reasons: 9 Escorts -0.039
* The Indian companies do not apply any dividend 10 Godfrey Philips -0.469
model or theory while deciding the D/P ratio. 11 Goodyear 0.141
When asked, all the companies replied in negative. 12 Grasim Industries -0.428
* Dividend policy and dividend payout ratio are 13 HEG Ltd. -0.632**
influenced by a number of factors (Tables 6 and 7), 14 ICI Ltd. -0.179
and most of these factors are not measurable as they 15 ICICI 0.702 *
are qualitative in nature. Hence, their effect cannot
16 Kelvinator 0.389
be segregated. As a result, it was not possible to
establish an exact and definite relationship between 17 Kinetic Engg. 0.093
the two. Furthermore, it is also evident from these 18 Kirloskar Brothers 0.415
tables that the single most important factor is the 19 Mukand Ltd. 0.775*
opinion of the directors, and the shareholders have 20 Nestle -0.337
very little say in dividend policy matters. Also, the 21 Ranbaxy Laboratories -0.511
Indian companies have no specific criteria for
22 Rathi Alloys Ltd. 0.211
deciding the retention ratio. Generally, it is quite
high. This fact, too, was confirmed in the study, as 23 SPIC -0.441
the average retention ratio turned out to be 69.04 24 SRF Ltd. 0.270
per cent. 25 TELCO -0.480
Though the relationship between capital structure 26 TISCO 0.090
and dividend payout ratio is not statistically significant, Pooled data 0.368
it does tell that as degree of leverage increases, dividend
payout ratio also increases moderately. This is because * and ** represent the significance of results at 1 percent
of the fact that EPS increases with leverage as long as the and 5 per cent level, respectively.
company is solvent.
Table 6 : Determinants of Dividend Policy
Differences in Inter-company Capital Structure : The
differences in the capital structure of companies, whether
Rank Deteminant
belonging to the same group or not was statistically
significant as f value was 6.174 at 1 per cent significant
level. This is because of the fact that capital structure 1 Board of Director’s Opinion
depends on a number of factors whose magnitude varies 2 financial needs of the company
from company to company. This was confirmed by 3 Growth of the company
executives when they were asked, "What are the determi- 4 New security issues
nants of your existing capital structure?" (Table 8). 5 Liquidity
6 Restriction in loan agreement
* One way ANOVA test was applied on the mean values 7 Desires o shareholders
of D/C ratio of all the 26 companies. Detailed results 8 Legal restrictions
available on request. 9 Investment opportunity

34 Vikalpa
Table 7: Factors Affecting DIP Ratio Decision
Rank Factor
1 Operating results
2 Business risk
3 Economic conditions
4 Promoters
5 Tax rates structures
6 Political conditions
Table 8 : Determinants of Capital Structure
Rank Factor
1 Cost o capital
2 Dividend policy
3 Market conditions
4 Earning stability
5 Nature of industry
6 Government rules
7 Size of the company
8 Restrictions by fis
9 Industry norm
10 Management decisions and policies
Leverage : Except for ICICI, average values of debt to
total capital ratio (D/ C) of the sample companies ranged
from 0.2831 to 0.7139. So, none of the companies was
found tobe excessively leveraged, as over the years, a D/
E ratio of 2:1 (i.e. D/C ratio of 0.67) was the norm fixed
by the government and its agencies. For ICICI, it was
0.916 which is also not high because the norm fixed by
the World Bank and the RBI for financial institutions is
0.923. The mean value of D/C ratio for all the companies
in the sample was 0.666 (Table 1).
Conclusions
No significant relationship was found between change
in capital structure and the value of a firm, at the micro
level. This is because of the fact that the value of a firm
is affected by a multiplicity of factors and capital structure
is just one of them. Many of these factors like the reputa-
tion of promoters, management of the company, econo-
mic and political conditions, role of bulls and bears,
government policies, etc., are not measurable as they are
qualitative in nature. Because of this problem, their
Vol. 21, No. 3, July-September 1996
effect cannot be segregated, and hence, an exact relation-
ship between change in capital structure and value of a
firm could not be established. However, at the macro
level, the relationship was statistically significant at 5 per
cent level of significance (r = 0.706). The above factors
may result in undervaluation or overvaluation of shares
at the micro level but when we take the aggregate, their
positive and negative effects neutralize one another. So,
the market value at the macro level acts as the true index
of financial performance of all the companies. The
results clearly advocate the existence of an optimal
capital structure at the macro level but in the absence
of a model on capital structure, it is not possible to
determine its exact range. However, the 'r' value of
0.706 for a weighted average D/C ratio of 0.666 is high
and statistically significant. What it implies is that a
higher level of debt in the capital structure of these firms
will not affect their values adversely. As a matter of fact,
the additional debt will help increase their values.
Companies were found to differ significantly in
capital structure irrespective of whether they belong to
the same industry group or different groups. This is
because of the fact that the magnitude of the effect of
determinants of capital structure vary from company to
company. In general, change in capital structure and
cost of capital were found to be negatively related, but
the results were not statistically significant. These results
suggest that though cost of capital decreases when leve-
rage increases, this decrease is very moderate and not
proportional to debt level. Probably, it is for this very
reason that most of the companies are not high leveraged.
The relationship between change in capital structure
and dividend policy was not found definite and statisti-
cally significant. Further, it was also found that Indian
companies do not employ a specific model for computing
the cost of capital and have no scientific model for
determining their target capital structure. Thus, it could
be concluded that like perfect capital markets of the
west, in India, too, wherein the capital markets are
imperfect, companies have no definite way of deter-
mining their optimal capital structure.
References
Baron, D P (1974). "Default Risk, Home-made Leverage
and M-M Theorem," American Economic Review,
64, pp 176-82.
Baron, D P (1975). "Firm Valuation, Corporate Taxes and
Default Risk," Journal of Finance, 30, pp 1251-64.
Brennen, M J and Schwartz, E S (1978). "Corporate
Income Taxes, Valuation and the Problem of
Capital Structure," Journal of Business, pp 103-15.
Brigham, Eugene F and Gopanski, Louis C (1985).
35
 Intermediate Financial Management. New York :
Dryden Press, p 204.
Davenport, M (1971). "Leverage and the Cost of Capital:
Some Tests Using British Data/' Economica, pp
136-62.
DeAngelo, Harry and Masulis, M S (1980). "Optimal
Capital Structure under Corporate and Personal
Taxation," Journal of Financial Economics, 8, pp
3-29.
Durand, David (1963). "The Cost of Capital in an
Imperfect Market: A Reply to M-M," American
Economic Review, 53.
Jensen, M and Meckling, W (1976). "Theory of the Firm:
Managerial Behaviour, Agency Costs and Owner-
ship Structure," Journal of Financial Economics, 3,
pp 305-60.
Kraus, A and Litzenberger, R H (1973). "A State Prefe-
rence Model of Optimal Financial Leverage,"
Journal of Finance, 28, pp 911-22.
Masulis, M S (1980). "The Effect of Capital Structure
Changes on Security Prices: A Study of Exchange
Offers," Journal of Financial Economics, 8, pp
139-78.
Masulis, M S (1983). "The Impact of Capital Structure on
Firm Value," Journal of Finance, 38, pp 107-25.
Miller, M H (1977). "Debt and Taxes," Journal of Finance,
32, pp 261-73.
Modigliani, F and Miller, M H (1958). "The Cost of
Capital, Corporation Finance and the Theory of
36
Investment," American Economic Review, 48, pp
261-97.
Modigliani, F and Miller, M H (1963). "Corporate Income
Taxes and the Cost of Capital: A Correction,"
American Economic Review, 53, pp 433-43.
Modigliani, F and Miller, M H (1966). "Estimates of Cost
of Capital to Electric Utility Industry 1954-1957,"
American Economic Review, 56, pp 333-91.
Pandey, IM (1992). Capital Structure and Cost of Capital.
Vikas Publishing House.
Scott Jr., J H (1976). "A Theory of Optimal Capital
Structure," Bell Journal of Economics, Spring, pp
33-54.
Scott Jr., J H (1977). "Bankruptcy, Secured Debt and
Optimal Capital Structure," Journal of Finance,
32, pp 261-73.
Sharma, R and Rao, H (1969). "Leverage and the Value
of the Firm," Finance Journal, 24.
Smith, V L (1972). "Default Risk, Scale and Home-made
Leverage Theorem," American Economic Review,
62.
Solomon, Ezra (1963)." Leverage and the Cost of Capital,"
Journal of Finance, 18, pp 273-79.
StiglitzJE (1969). "A Re-examination of M-MTheorem,"
American Economic Revieiv, 59, pp 784-93.
Stiglitz, J E (1974). "On the Irrelevance of Corporate
Financial Policy," American Economic Review,
62, pp 851-66.
Vikalpa