Cost of Capital, Optimal Capital Structure, and Value of Firm: An Empirical Study of Indian Companies Raj § 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 evel 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: eptember 1996 Introduction During the last SO 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, ther imelevancy 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 @ 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), Baton (1974,1975), and Scott (1976, 1977), supported the M-M model, but only under the conditions of risk free debt and costless bankruptcy. When bankruptey 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 (197). He argued that bankruptey and ageney 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 mte, 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 costs are taken into account, an optimal capital structure exists, irrespective of availability of non-debt tax shields. Masulis (1980,1983), Brennen and Schwartz (1978), and Jensen and Meckling (1976) also advocated the existence of an optimal capital structure in an imperfect market, while using different mechanisms. Besides, a lot more work has been done on this problem till now, but a formal model, showing the mechanism for determining an optimal capital structure in an imperfect market, is yet to be developed. On the basis of the major work done by M-M and 30 ‘others so far in this area, we can identify three clearly defined models: Model 1 When the capital market is perfect EBIT EBIT VL=vu= an Ka Kev where VL and VU = Valucoftheleveragedand unleveraged firm respectively EBIT = _Earmings before interest and taxes Ka = Average cost of capital KeU = _ Cost of equity of unleveraged firm ‘What themodel showsis that the value ofa firm and itscost of capitalare independent ofits capital structure. In other words, there does not exist an optimal capital structure. Model 2 When only corporate taxes are taken into account VL = WU+T.D where T Corporate tax rate D = — Amountof debt Thismodelassumesbankruptcy, agency, and related costs to be too insignificant to affect the value of a firm. Thus, the value of a firm increases with increase in debt level. But it fails to answer the question, “how long will the value increase with increase in debt level?” Model 3 When bankruptcy, agency, and other related costs are taken into account Frsent va] [RRA] RST) [Pre otexpected | finvaive | [invsiue | | vatocos vievuerp-[bantpey [ftom ecm twer| | agency cae jewea [feat eee ost of te This model clearly advocates the existence of an optimal capital structure which is a trade-off between tax advantages and disadvantages of leverage. However, M-M (1966) again proved their inrelevancy hypothescis. But the study by Davenport (1971) supported the tradi- tional view. In the Indian context, one comes actoss two works, one by Sharma and Rao (1969) and the other by Pandey (1992). The former tested the M—M model using cross- sectional analysis for engineering companies, wherein the value of a firm was found to be independent of its capital structure after allowing for tax advantage. But the results could not be generalized as the sample was Vikalpahomogeneous. The other work by Pandey (1992) obser- ved that the M-M theory is not fully valid under Indian conditions. He concluded that, initially, cost of capital and value of a firm are independent of the capital structure changes, but they rise after a certain level. All these studies have helped understand the dynamics of this crucial issue better but have not been able to come to a definite conclusion as to how firms determine their ‘optimal capital structure. So, the present study was planned to make another attempt to resolve this contentious issue. It may be pointed out that the study hhas not included the effect of factors like agency and bankruptcy costs, as they are difficult to measure in the Indian scenario. Objectives The basic objective of this study is to find out if there exists an optimal capital structure either at the micro and/or at the macro level in Indian private sector companies. If yes, how do companies determine it? In addition, it has the following sub-objectives: Does change in capital structure affect the cost of capital? Are capital structure and dividend policy correlated? Does capital structure of companies differ significantly? Research Methodology For conducting the study, a sample of 26 widely held Indian private sector companies from top 300 large scale companies was taken. The sample is not homogeneous as the companies are taken from 15 different industry groups. It is a multi-period study covering the period from 1981-82 to 1990-91. The study used both primary and secondary data, The main source of secondary data was the Bombay Stock Exchange Directory. Primary data were collected through a mailed questionnaire. In al, 100 questionnaires were mailed but the response rate was about 26 per cent. Hence, a sample of 26 companies was taken, Most of the respondents were above the level of Finance Manager. The responses were ranked on the basis of relative weightage given to various factors by the respondents, Measure of Variables The variables used in the study are as follows: 1) Capital structure (C): C= Sum of the book values of various components of capital structure 2) Market value of equity (E): E = Number of shares outstanding at the end of an accounting year X Average market price per share where, ‘Number of shares = Total equity capital/Face value per share Average price = Mean value of monthly high and low price/share during the accounting year. 3) Value of firm (V): v Market value of equity + Book value of preference share and debt’ 4) Cost of eapital (Ka): KpP+KdD+KeE Ka P+D+E where, Kp = Cost of preference capital” Kd = Cost of debt capital? Ke = Cost of equity capital* P,D = Book value of preference and debt capital respectively E Market value of common equity capital wherein, Ke ~ RF+B (Rm-R#) where, RF = Risk free rate’ B Beta cocfficient® Cov (Ri,Rm) Var (Rm) where, Rm = Return on market portfolio Rj = Retum on common stock Var (Rm) = Variance of return on market portfolio 1. Preference share and debt are taken at book values because there is no significant yearly fluctuations in the prices of these sources of capital. 2, Preference dividend/price of preference share. 3, Interest (1 - T)/Total debt, 4, Calculated by using Capital Asset Pricing Model Frequency of variables is on an annual basis. 5. Risk free rate is taken as 10 per cent, 6, It represents the systematic risk associated with "a security. 3si) Leverage (L): L=Debt to total capital ratio, ie. 6) Dividend payout ratio (D/P): Equity dividend paid L=D/(D+E)=DIC bP Net profit - Preference dividend In the absence of a well developed model on capital structure problem, Karl Pearson's bivariate correlation coefficient was used to find the relationship among the variables, The value of r for the pooled data represents a correlation between average change in D/C ratio (change in capital structure) on the one hand, and total market value of firms, average Ka and average D/P ratio, respectively, on the other (Tables 1,4,5). Considering the nature and objectives of the study, we thought it proper to use the bivariate correlation technique. As a ‘matter of fact, most of the earlier studies in this arca have used a similar technique. The correlation results were tested by using student's t-test; ftest has been used to test the significance of difference in inter-company capi- tal structure, Empirical Results Optimal Capital Structure : There is no definite relationship between change in the capital structure and the value of a firm (Table 1), Out of 26, only 6 companies show statistically significant relationship (3 showing negative and 3 showing positive values). Further, none of the sample companies was found to be highly levera- ged. Except for ICICI, the maximum value of DiC ratio for a company was 0.714. For ICICI, it was 0.916 which is normal for a financial institution. These insignificant and inconsistent results at the micro level can be attributed to the fact that capital structure isnot the only determinant ‘of market price of a company's share and its value and that there are other factors as well which do affect their values. Incidentally, this was confirmed during the course of our discussions with company executives (Tables 2 and 3). ‘Table 1: Coefficient of Bivariate Correlation (r) bet= ‘ween Change in Capital Structure and Valu e of the Firm Nilme o/ the Company — D/C Ratio 1 Amrit Banaspati Co, 0536 2 Apollo Tyres 0.765 3 Asea Brown Boveri 0,580 4 Atlas Cycle Industries 0.582 5 Ballarpur Industries 0358 6 Bata India oars > BlueStar 0.446 8 0.668 0.091 9 0477-9015 10 Godfiey Philips 0488 0.226- 11 Goodyear Grasim 0283 0682" 12 Industries, 0431 0.746 * 13 HEGLd. 0410 0.838 * 14 ICLLtd 0494 0.137 1s ICICI 0916 0.467 16 Kelvinator 0534 -0.206 17 Kinetic Engg, 0438 0,079- 18 _ Kirloskar Brothers 0.684 0375 19 Mukand Ltd 0563 0.131 20 Nestle 0384 0459 21 Ranbaxy Laboratories 0.688 -0.239 22 ~— Rathi AlloysLtdSPIC 0.714 0.280- B 0569 0298 24 SRFLtd 0437 0273 25 TELCO 0583 0.626" 26 TISCO 0512 -0.796* Pooled data 066670708" and ** represent the significance of results at | per cent and 5 per cent level respectively. # 0.666 is the weighted average D/C ratio of all the companies for ten years. ‘Table 2 : Determinants of Share Prices Rank Determinant Management of the company Dividend policy Role of bulls and bears Capital structure Government policies Takeover bid by others Cost of capital ‘Table 3 : Other Factors which Affect the Value of a Firm Rank Factor 1 Operating results Business risk 2 Economic conditions Promoters 3 Tax rates and structures Political conditions 4 Vikalpa 5Since a majority of the factors are non-measurable as they are qualitative in nature, itis not possible to segregate their effects. Therefore, an exact relationship between capital structure and value of a firm cannot be establi- shed, This conclusion is further strengthened by the highly volatile behaviour of the stock markets. One ob- serves that 22 companies showed coefficient of variance (CV) greater than 50 per cent. The average C V for all the companies was 62.11 per cent which is quite high. Fur- ther, in general, variations in values were not found to be significantly associated with the financial performan- ce of the companies. So, the market price of the stock is not the true index of a company’s performance. The va- riations in stock prices of a company may also represent the effects of qualitative factors. These factors can undervalue or overvalue these prices at the micro level, as a result of which they may not be the true! indices of a company's performance, However, at the macro level, the relationship bet- ween change in capital structure and value ofa firm was found to be highly positive and statistically significant (r= 0.706) (Table 1), These results are totally different from those obtained at the micto level. The reason is quite clear, because when we take the aggregate figures, the positive and negative effects of the external and qualitative factors on individual shares neutralize one another and we get a closely approximate true value of the share prices. That is why we get a highly positive cortelation between the two at the macto level, The above results clearly show that in imperfect ‘market, at the macro level, an optimal capital structure definitely does not exist. It is not possible to determine it exactly, because of the difficulty faced in the measu ment of qualitative factors and other problems. These results fully support the views expressed by Brigham and Gopanski (1985) when they observed, "Unfortuna- tcly, it is almost impossible to test the leverage effects empirically, because (1) future earnings are impossible to measure and (2) most real world leverage changes are accompanied by asset changes which may be changing the firm's risk class. Thus, empirical tests have not pro- duced conclusive results. However, the evidence does generally support the contention that some benefits from leverage do exist, atleast ifthe firm does not exceed reasonable limits of debt. Capital Structure and Cost of Capital : In gencral, change in capital structure and cost of capital were found to be negatively related, as 81 per cent of the companies showed negative relationship (Table 4), thus, supporting the theory that cost of capital decreases with increase in debt level because cost of debt is less than that of equity and interest payments are tax exempted. Furthermore, Vol. 21, No, 3, July-September 1996 since the cost of capital is measured using historical data, the weighted average" cost of capital is bound to go down with increasing debt, other things being equal, But the relationship is not statistically significant. It means that, in general, changes in capital structure are not acco mpanied by proportionate changes in cost of capital. The statistically insignificant values of r'can be explai- ned by the following reasons: * Indian companies have no specific model or mechanism (o compute the specific costs of capital, particularly the cost of equity capital and the average cost of capital. This became clear, when ,in response to the question, "how do you compute the cost of capital?", none of the companies suggested a defi nite/specifie mechanism. * Cost of capital is not the only determinant of the capital structure, though it is one of the most important determinants. Since the effects of other determinants cannot be segregated, so, an exact relationship between the two could not be established, * Cost of debt in India is quite high as compared to that in the developed countries. Table 4: Coefficient of Bivariate Correlation (r) between Change in Capital Structure and Cost of Capital Name of the Company 7 Amrit Banaspati Co, “0217 Apollo Tyres Asea ois T 2 3 Brown Boveri 4 Atlas Cycle Industries 5 Ballarpur Industries 6 7 8 9 Bata India Blue Star DCM Escorts 10 Godfrey Philips 1 Goodyear Grasim 12. Industries 13° HEGLtd 14 IciLtd 15 ICICl 16 Kelvinator 17 Kinetic Engg. 18. Kirloskar Brothers 19 Mukand Ltd, 20 Nestle 21 Ranbaxy Laboratories 22 Rathi Alloys Lid. 2B SPIC24 SRF Lid. 0344 25 TELCO 0.176 26 TIsco 0216 Pooled data “0.245 and ** represent the significance of results at | per cent and 5 per cent level, respectively. Capital Structure and Dividend Policy : At the micro level, no definite and consistent relationship exists between change in capital and dividend policy (Table 5), as 14 companies showed positive values of r whereas 12 companies showed negative values. The relationship ‘was statistically significant in case of 3 companies only. These inconsistent results could be attributed to two * The Indian companies do not apply any dividend model or theory while deciding the D/P ratio. When asked, all the companies replied in negative, * Dividend policy and dividend payout ratio are influenced by a number of factors (Tables 6 and 7), and most of these factors are not measurable as they are qualitative in nature. Hence, their effect cannot be segregated. As a result, it was not possible to ‘establish an exact and definite relationship between the two. Furthermore, it is also evident from these tables that the single most important factor is the ‘opinion of the directors, and the shareholders have very little say in dividend policy matters. Also, the Indian companies have no specific criteria for deciding the retention ratio, Generally, itis quite high. This fact, too, was confirmed in the study, as the average retention ratio turned out to be 69.04 per cent, Though the relationship between capital structure and dividend payout ratio is not statistically significant, it docs tell that as degree of leverage increases, dividend payout ratio also increases moderately. This is because of the fact that EPS increases with leverage as long as the company is solvent. Differences in Inter-company Capital Structure : The differences in the capital structure of companies, whether 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 depends on a number of factors whose magnitude varies from company to company. This was confirmed by executives when they were asked, "What are the determi- nants of your existing capital structure?" (Table 8). * One way ANOVA test was applied on the mean values of DIC ratio of all the 26 companies. Detailed results available on request. 34 Table 5 : Coefficient of Bivariate Correlation (r) bet- ween Change in Capital Structure and Dividend Payout Ratio Name of the Company r T Amrit Banaspati (0.081 2 Apollo Tyres 0.756" 3 Asea Brown Boveri 0.110 4 Atlas Cyele Industries 0.074 5 Ballarpur Industries 0.037 6 Bata India 0,089 7 Blue Star 0366 8 DCM 0512 9 Escorts 0.039 10 Godfrey Philips 0,469 11 Goodyear oat 12 Grasim Industries 0.428 13° HEGLd. 0.632" 14 ICIL. 0.179 1s ICIcr 0.702 * 16 Kelvinator 0389 17 Kinetic Engg. 0.093 18 Kirloskar Brothers os 19 Mukand Ltd, o775* 20 Nestle -0337 21 Ranbaxy Laboratories 0511 22 Rathi Alloys Ltd, out 23° SPIC 0441 24 SRF Lid. 0270 25 TELCO 0.480 26 TISCO 0,090 Pooled data 0368 * and ** represent the significance of results at | percent and 5 per cent level, respectively. Table 6 : Determinants of Div jend Policy Rank Deteminant Board of Director's Opinion financial needs of the company Growth of the company New security issues Liquidity Restriction in loan agreement Desires o sharcholders Legal restrictions Investment opportunity VikalpaTable 7: Factors Affecting DIP Ratio Decision Rank Factor 1 Operating results 2 Business risk eonomie conditions 4 Promoters 5 Tax rates structures 6 Political conditions ‘Table 8 : Determinants of. 1 Structure Rank Factor 1 Cost o capital 2 Dividend potic 3 Market conditions 4 Eaming stabilit Nature of industry Government rules Size of the company Restrictions by fis, 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 companics 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 DiC 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 truc 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 't’ value of 0.706 for a weighted average DiC 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, $3. Jensen, M and Meckling, W (1976). "Theory of the Firm: nagerial Behaviour, Ageney 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, ). "The Impact of Capital Structure on Journal of Finance, 38, pp 107-25. Miller, MH (1977). "Debt and Taxes," Journal of Finance, 32, pp 261-73. Modigliani, F and Miller, MH (1958). "The Cost of Capital, Corporation Finance and the Theory of 36 Investment," American Economic Review, 48, pp 261-97. Modigliani, F and Miller, MH (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 Cast 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 Jowmal, 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. StiglitzIE (1969). "A Re-cxamination 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