Determinants of Capital Structure with Reference to Select Indian Companies: A Panel Data Regression Analysis

Introduction

Capital structure is the combination of various long-term resources such as equity share capital, preference share capital, retained earnings, debentures and long-term loans. It represents how a company uses its long-term funds in operations and growth by utilising various capital resources. The assessment of present and future capital requirement is essential for smooth functioning of a firm as well as for deciding its capital mix. Optimal capital structure is that blend of long-term debt and shareholders’ fund, which maximises firm value and minimises its cost of capital.

In the corporate world, capital structure has been an important topic of diverse theoretical and empirical works in past decades, and inconsistencies of outcomes of the majority of past studies have shown that this topic has not been so far put to rest. To add, the conclusion of Modigliani and Miller’s study (1958) gave birth to the investigation about the relevance or irrelevance of capital structure in deciding a firm’s value.

From earlier decades, finance researchers are involved in hypothesising extensively on the identified relevant determinants of capital structure. Some of them have viewed that capital structure is a function of corporate (Modigliani & Miller, 1963) and personal taxes only, and other researchers have established that financing decisions reveal equating the tax-shield benefits of higher leverage with bankruptcy cost (Trade-Off Theory by Kraus & Litzenberger, 1973; Kim, 1978; Myers, 1984). In the context of lesser bankruptcy costs, some researchers postulate that higher leverage can still have significant advantages in managing over-investment (i.e., free cash problems) (Agency Cost Theory by Jensen & Meckling, 1976). Yet others remark that financing decisions are mainly related to ‘Signalling’ effects of such choices (Signalling Theory by Ross, 1977). Extending this signalling effects, some researchers have proposed that, actually, capital structure is cumulative outcome of successive distinct financing choices, where managers adopt pecking order, that is, retained earnings first, then debt and, lastly, equity, if needed (Pecking Order Theory by Myers & Majluf, 1984). Even if these theories explain, at best, a few significant parts of corporate financial policies, some of them accurately argue that the capital structure policy remains still unresolved.

Based on capital structure theories, a considerable number of empirical researches are carried out on the determinants of capital structure across various countries. Rajan and Zingales (1995) investigated firms from group of seven (G-7) countries (the USA, Japan, Germany, France, Italy, the UK and Canada) from 1987 to 1991 and realised that correlation found between determinants of capital structure and leverage in past cross-sectional research works in the USA was the same as correlation of determinants of capital structure with leverage for other countries also under study.

Wald (1999) made a cross-country comparison of five developed economies, namely France, Japan, Germany, the UK and the USA to analyse the determinants of capital structure in an international scenario. The explanatory variables, like moral hazard, non-debt tax shields, research and development and profitability have their predicted influences on leverage and risk, growth, size and inventories have varied influences across various nations under study. This evidence suggests that institutional factors play an important role in capital structure decision, and agency and monitoring costs existing in each nation generate varied results.

Booth, Aivazian, Demirguc-Kunt, and Maksimovic (2001) assessed firms from 10 developing economies (Brazil, South Korea, Turkey, Zimbabwe, Mexico, India, Malaysia, Pakistan, Jordan and Thailand). The study observes that factors significant in determining the capital structure of the USA and European firms are also significant in the cases of firms operating in these 10 emerging economies, regardless of large disparities in institutional variables in these emerging economies. Further, firms with higher profitability use less debt, irrespective of the measure of debt ratio applied for. In addition, study illustrates that higher the tangibility, higher is the long-term debt ratio and lower is the total debt ratio.

Ozkan (2001) had evaluated the factors influencing capital structure of 390 UK firms, spanning a study period from 1984 to 1996. The study discloses that leverage (ratio of total debt to total assets) is significantly negatively dependent on profitability, liquidity, growth opportunities and non-debt tax shields. However, little evidence in support of significant positive relationship between size and leverage is observed.

Pandey (2001) analyses the factors determining the capital structure of 106 Malaysian companies from 1984 to 1999. The study found that growth and size had significant positive influence on leverage, whereas profitability had significant negative influence.

Antoniou, Guney, and Paudyal (2002) assessed the determinants of capital structure of UK firms, French firms and German firms for the study periods from 1969 to 2000, 1983 to 2000 and 1987 to 2000, respectively. The profitability and investment opportunities have significant negative influence on leverage in France and the UK, while size has significant positive relation with leverage for all three countries. The influence of tangibility on leverage is, however, diverse (significant positive in case of Germany, but insignificant in France and negative in case of the UK).

Huang and Song (2002) examined the capital structure of more than 1,000 Chinese listed companies from 1994 to 2000. The study concluded that leverage had significant negative relation with profitability and significant positive relation with size. Tangibility had significant positive relation with long-term debt ratio. Further, companies with fast sales growth were likely to have greater leverage, whereas companies with better growth opportunities were likely to have lower leverage. In addition, leverage tended to increase with increase in earnings’ volatility.

Deesomsak, Paudyal, and Pescetto (2004) investigated the factors influencing the capital structure of firms functioning in four countries of Asia Pacific region, namely Thailand, Malaysia, Singapore and Australia. The study found significant negative association of leverage with growth opportunities, non-debt tax shield, liquidity, share price performance and significant positive linkage with firm size. Further, profitability had significant negative relation with leverage in case of Malaysian firms, whereas firm size had no significant relation with leverage for Singaporean firms. They concluded that although the firm-specific factors were vital for designing capital structure policy, country-specific factors also had a strong bearing on such choice, and their impact could not be disregarded.

Frank and Goyal (2009) investigate determinants of capital structure of publicly traded US firms for the time period 1950–2003. The study observed that tangibility, firm size (log of assets) and expected inflation had significant positive impact on leverage; on the contrary, growth (market to book ratio) and profitability had significant negative impact.

Fauzi, Basyith, and Idris (2013) have analysed the determinants of 79 New Zealand listed firms. The study demonstrates that tangibility, growth, signalling, managerial ownership and firm size have significant influence on total debt level. All explanatory factors, except firm size (validated by Pecking Order Theory), produce evidence in support of Trade-Off Theory. Further, non-debt tax shield and profitability have insignificant relation with total debt.

Under the backdrop of this discussion, the present study is a modest attempt in the quest of resolving the unsettled issues on capital structure.

Objectives of the Study

The study is undertaken:

to examine the capital structure decision of 50 sample companies across different industries taken together over a 15-year study period, in order to get an aggregated analysis of capital structure determinants. For this purpose, eight firm-specific factors, namely profitability, size, liquidity, tangibility, uniqueness, non-debt tax shield, growth and income variation are selected here to verify their relationship with leverage;

Hypotheses of the Study

In consistency with the objectives of the present work, the following statistical hypotheses are devised for testing:

There is negative relationship between profitability and leverage.

Data Base and Research Methodology

Panel Data Analysis

In our study, we have chosen 50 Bombay Stock Exchange (BSE)-listed sample companies by choosing 5 sample companies from each of the 10 industries, namely computer hardware, large heavy engineering, large electric equipment, mining/mineral, textile, pharmaceutical, large tyres, cement, fertilizer, fast-moving consumer goods over a 15-year time period, ranging from 1999–2000 to 2013–2014. Therefore, we have panel data sets for our analysis. Total number of observations for the panel data set should be 75 (5 companies × 15 years) × 10 industries, that is, a total of 750. But we have actually 742 observations in our study, ignoring rest 8 observations, because of change in financial years for a few companies {JK Lakshmi Cement Ltd (2002—year ending September to 2004—year ending March), Gillette India Ltd (2005—year ending December to 2007—year ending June), Hindustan Unilever Ltd (2007—year ending December to 2009—year ending March), ALSTOM India Ltd (2000—year ending December to 2002—year ending March), W S Industries (India) Ltd (2001—year ending December to 2003—year ending March, 2012—year ending March to 2013—year ending September)}, non-availability of data due to late incorporation of few companies (UltraTech Cement Ltd [2001] and Bombay Rayon Fashions Ltd [2001–2003]) and data missing for Himadri Chemicals & Industries Ltd for the year 2000. In our study, for the analysis of determinants of capital structure, the dependent variable is selected as leverage (ratio of total debt to total assets) of the companies and independent variables for each company are (a) Profitability, (b) Size, (c) Growth, (d) Tangibility, (e) Non-Debt Tax Shield (NDTS), (f) Liquidity, (g) Uniqueness and (h) Income Variation.

OPEN IN VIEWER

Table 1.

Theoretically Predicted Relationship of Independent Variables with Leverage and Measurement of Variables

Independent Variables	Expected Sign			Proxy
	Trade-Off Theory	Pecking Order Theory	Agency Cost Theory
Profitability (profit)	+ (ve)	– (ve)	+ (ve)	Ratio of earnings before interest and tax to total assets
Size (size)	+ (ve)	– (ve)	+ (ve)	Natural logarithm of net sales
Growth (GR)	–(ve)	+ (ve)	– (ve)	Percentage change in total assets
Tangibility(Tan)	+ (ve)	– (ve)	+ (ve)	Ratio of fixed assets to total assets
Non-debt tax shield (NDTS)	– (ve)	?	?	Ratio of annual depreciation expenses to total assets
Liquidity (Liq)	+ (ve)	–(ve)	–(ve)	Ratio of current assets to current liabilities
Uniqueness (Uni)	– (ve)	+ (ve)	?	Ratio of selling expenses to sales
Income variation (IV)	– (ve)	– (ve)	+ (ve)	Natural logarithm of standard deviation of earnings before interest and tax

Source: The authors.

Model Selection

In panel data regression Technique, there are prevalently three models, namely pooled regression model, fixed-effects model and random-effects model. To make a selection between the fixed-effects model and the random effects model, we have employed popular Hausman specification test. If test statistic value of Hausman specification test is significant, the fixed-effects model is a suitable model for choice; otherwise, the random-effects model should be selected. On the other hand, for the choice between the pooled regression model and the random-effects model, we have implemented the Breusch and Pagan Lagrange Multiplier test. If the test statistic value of the Breusch and Pagan Lagrange Multiplier test is found to be significant, the random-effects model is an appropriate model for selection. In case of insignificant value of test statistic, one should choose the pooled regression model. However, for contradictory results arising from these two tests, we have employed the fixed-effects model as per the convention of earlier studies and also on the basis of textbook suggestion (Greene, 2006, pp. 283–338; Maddala, 2005, pp. 573–584). Industry-wise model specification for analysing the determinants of capital structure is summarised and presented in Table 2.

OPEN IN VIEWER

Table 2.

Industry-wise Selection of Model for Panel on Determinants of Capital Structure

Industry	Observation	Hausman Specification Test	Breusch and Pagan Lagrange Multiplier Test	Selected Model
Cement	73	Failed	0.00 (1.0000)	Pooled regression model
Computer hardware	75	134.63* (0.0000)	0.00 (1.0000)	Fixed-effects model
Large heavy engineering	75	Failed	0.00 (1.0000)	Pooled regression model
Fertiliser	75	Failed	0.00 (1.0000)	Pooled regression model
Fast moving consumer goods	73	45.87* (0.0000)	0.00 (1.0000)	Fixed-effects model
Large electric equipment	73	Failed	0.00 (1.0000)	Pooled regression model
Mining/mineral	74	382.95* (0.0000)	0.00 (1.0000)	Fixed-effects model
Textile	74	Failed	0.00 (1.0000)	Pooled regression model
Large tyres	75	20.60* (0.0083)	0.00 (1.0000)	Fixed-effects model
Pharmaceutical	75	Failed	0.00 (1.0000)	Pooled regression model

Source: The authors.

Note: * implies significance at 1% level.

Limitations of the Study

Data Analysis and Findings

Comparative Analysis Among the Industries

This section presents the comparative analysis of each determinant of capital structure across the industries in a concise way for better explanation of observed estimated coefficients of determinants (from panel regression estimation) that analyse the capital structure, their consistency or inconsistency, with the expected results and acceptance of relevant capital structure theories across selected industries.

The regression analyses (Table 3) demonstrate that profitability has strong significant negative effect on leverage, which follows the expected result in almost all cases except fertilizer and large tyres industries. Therefore, it may be concluded that profitability emerges as a vital determinant in devising the capital structure of the selected industries, and this strong negative relationship, supported by the Pecking Order Theory, reveals that greater proportion of profits is likely to raise the internal fund for financing future investment projects and, therefore, less dependency on external borrowings.

OPEN IN VIEWER

Table 3.

Relationship Between Profitability and Leverage

Industry	Estimated Coefficient	Standard Error	Null Hypothesis Accepted or Rejected	Consistent or Inconsistent to Expectation	Acceptance of Relevant Theory
Cement	–0.7356*	0.1571	Rejected	Consistent	Pecking Order Theory
Computer hardware	–0.5073*	0.1797	Rejected	Consistent	Pecking Order Theory
Large heavy engineering	–0.3607*	0.0888	Rejected	Consistent	Pecking Order Theory
Fertilizer	–0.2989	0.2389	Accepted	–	–
FMCG	–0.1037***	0.0605	Rejected	Consistent	Pecking Order Theory
Large electric equipment	–0.5044*	0.1389	Rejected	Consistent	Pecking Order Theory
Mining/minerals	–0.3358*	0.0926	Rejected	Consistent	Pecking Order Theory
Textile	–0.5622*	0.1889	Rejected	Consistent	Pecking Order Theory
Large tyres	–0.1588	0.1824	Accepted	–	–
Pharmaceutical	–0.6426*	0.1804	Rejected	Consistent	Pecking Order Theory

Source: The authors.

Note: * and *** Imply significant at 1% and 10% levels, respectively.

The estimated regression results presented in Table 4 disclose that size is a least significant factor in determining the capital structure of the selected industries. The negative and statistically significant association between size and leverage is found for the computer hardware industry, which is contrary to the expected impact but in conformity with the Pecking Order Theory, and this observed negative association implies that sample companies in these industries prefer internal financing over external financing. On the other hand, a positive and statistically significant linkage between size and leverage is exhibited by the large heavy engineering, large tyres and pharmaceutical industries, which are consistent with expected outcome and support the combined propositions of the Trade-Off Theory and the Agency Cost Theory. This indicates that the sample companies belonging to these industries have greater access to capital market and can obtain debt at economical rate, are capable of bearing risk and, consequently, are less prone to the probability of bankruptcy risk.

From Table 5, it is evident that growth is a fundamental influencing factor of leverage, indicated by majority of selected industries excluding cement, large electric equipment, textile and large tyres industries. More specifically, the positive and statistically significant connectivity between growth and leverage is observed with respect to computer hardware, large heavy engineering, fertiliser, FMCG, mining/minerals, and pharmaceutical industries, which is opposing the expected impact, although validated by the Pecking Order Theory. This implies that internal capital is retained for financing new investment proposals and thus, asset growth is financed by debt capital.

OPEN IN VIEWER

Table 4.

Relationship Between Size and Leverage

Industry	Estimated Coefficient	Standard Error	Null Hypothesis Accepted or Rejected	Consistent or Inconsistent to Expectation	Acceptance of Relevant Theory
Cement	0.0338	0.0371	Accepted	–	–
Computer hardware	–0.0803**	0.0399	Rejected	Inconsistent	Pecking Order Theory
Large heavy engineering	0.1169*	0.0232	Rejected	Consistent	Trade-Off Theory/Agency Cost Theory
Fertilizer	0.0541	0.0477	Accepted	–	–
FMCG	–0. 0333	0.0445	Accepted	–	–
Large electric equipment	0.0591	0.0524	Accepted	–	–
Mining/minerals	0.0649	0.0398	Accepted	–	–
Textile	–0.0018	0.0245	Accepted	–	–
Large tyres	0.1847*	0.0405	Rejected	Consistent	Trade-Off Theory/Agency Cost Theory
Pharmaceutical	0.1508*	0.04999	Rejected	Consistent	Trade-Off Theory/Agency Cost Theory

Source: The authors.

Note: * and ** Imply significance at 1% and 5% levels, respectively.

The estimated regression results display that tangibility has no clear-cut influence on leverage as shown (in Table 6) by the statistically significant coefficients of 4 industries (i.e., in 40% cases) and insignificant for 6 (i.e., in 60% cases) out of 10 industries. The positive and statistically significant relation between tangibility and leverage is consistent with our anticipation, which is confirmed by the combined predictions of the Trade-Off Theory and the Agency Cost Theory; this represents that high tangible assets provide higher collateralised value to the debt holders to safeguard their interest, so they have greater access to the debt market to obtain fund at reduced rate and face low possibility of bankruptcy risk.

OPEN IN VIEWER

Table 5.

Relationship Between Growth and Leverage

Industry	Estimated Coefficient	Standard Error	Null Hypothesis Accepted or Rejected	Consistent or Inconsistent to Expectation	Acceptance of Relevant Theory
Cement	–0.00000285	0.00000257	Accepted	–	–
Computer hardware	0.00194**	0.000819	Rejected	Inconsistent	Pecking Order Theory
Large heavy engineering	0.0013*	0.0004	Rejected	Inconsistent	Pecking Order Theory
Fertiliser	0.0027*	0.0009	Rejected	Inconsistent	Pecking Order Theory
FMCG	0. 0009*	0.0003	Rejected	Inconsistent	Pecking Order Theory
Large electric equipment	–0.0008	0.0007	Accepted	–	–
Mining/minerals	0.0008***	0.0004	Rejected	Inconsistent	Pecking Order Theory
Textile	–0.00016	0.00018	Accepted	–	–
Large tyres	0.00048	0.00058	Accepted	–	–
Pharmaceutical	0.0026*	0.0007	Rejected	Inconsistent	Pecking Order Theory

Source: The authors.

Note: *, ** and *** Imply significance at 1%, 5% and 10% levels, respectively.

OPEN IN VIEWER

Table 6.

Relationship Between Tangibility and Leverage

Industry	Estimated Coefficient	Standard Error	Null Hypothesis Accepted or Rejected	Consistent or Inconsistent to Expectation	Acceptance of Relevant Theory
Cement	0.1289	0.1152	Accepted	–	–
Computer hardware	0.2898	0.1953	Accepted	–	–
Large heavy engineering	0.3934*	0.1251	Rejected	Consistent	Trade-Off Theory/Agency Cost Theory
Fertiliser	0.5377*	0.1337	Rejected	Consistent	Trade-Off Theory/Agency Cost Theory
FMCG	0. 1545	0.0981	Accepted	–	–
Large electric equipment	0.4113*	0.1499	Rejected	Consistent	Trade-Off Theory/Agency Cost Theory
Mining/minerals	–0.0268	0.0972	Accepted	–	–
Textile	0.3322*	0.1020	Rejected	Consistent	Trade-Off Theory/Agency Cost Theory
Large tyres	0.0618	0.1279	Accepted	–	–
Pharmaceutical	0.1705	0.1497	Accepted	–	–

Source: The authors.

Note: * Imply significance at 1% level.

In case of most of the industries as shown by the estimated regression results in Table 7, the non-debt tax shield is a less important factor affecting capital structure. Out of 10 selected industries, the statistically significantly inverse linkage is reported for mining/minerals industry, which is along the lines of our predicted outcome and also in conformity with the Trade-Off Theory. This signifies that non-debt tax shield like depreciation lessens the need of debt capital as depreciation charged on fixed assets facilitates to decline in the cash outflow by way of savings in tax payment, although it is not used as a substitute for debt capital for supplying fund as per demand. On the other hand, statistically significantly positive relation between non-debt tax shield and leverage is established by computer hardware, which is contrary to the predictions of the Trade-Off Theory, and it denotes that non-debt tax shield cannot be applied as a substitute for debt financing.

OPEN IN VIEWER

Table 7.

Relationship Between NDTS and Leverage

Industry	Estimated Coefficient	Standard Error	Null Hypothesis Accepted or Rejected	Consistent or Inconsistent to Expectation	Acceptance of Relevant Theory
Cement	0.10704	0.16894	Accepted	–	–
Computer hardware	0.5275**	0.2329	Rejected	Inconsistent	!!
Large heavy engineering	0.0385	0.1726	Accepted	–	–
Fertiliser	–0.0625	0.1344	Accepted	–	–
FMCG	–0. 0674	0.1476	Accepted	–	–
Large electric equipment	–0.1005	0.1370	Accepted	–	–
Mining/minerals	–0.4519*	0.1612	Rejected	Consistent	Trade-Off Theory
Textile	–0.1353	0.2268	Accepted	–	–
Large tyres	0.0217	0.2071	Accepted	–	–
Pharmaceutical	0.7358	0.5843	Accepted	–	–

Source: The authors.

Note: !! Indicates that no established theory supports this significant relation. * and ** Imply significance at 1% and 10% levels, respectively.

The estimated coefficients, presented in Table 8, reveal that liquidity has no clear impact on leverage as we obtain a statistically significant result in the cases of 5 industries (i.e., 50% cases) out of total 10 selected industries. We observe that in case of the cement industry, the relationship is negative and statistically significant, which is approved by the combined suggestions of the Pecking Order Theory and the Agency Cost Theory, and it is also consistent with the predicted result; it explicates that current assets may be used to fund potential investments. Whereas for industries like large heavy engineering, FMCG, large electric equipment and textile, the connectivity between liquidity and leverage is found positive and statistically significant, which is inconsistent with our expectation, though supported by the Trade-Off Theory and indicates better potentiality to meet short-term obligations and also shows ability to settle principal amount of debt timely.

OPEN IN VIEWER

Table 8.

Relationship Between Liquidity and Leverage

Industry	Estimated Coefficient	Standard Error	Null Hypothesis Accepted or Rejected	Consistent or Inconsistent to Expectation	Acceptance of Relevant Theory
Cement	–0.134**	0.0057	Rejected	Consistent	Pecking Order Theory/Agency Cost Theory
Computer hardware	–0.0033	0.0093	Accepted	–	–
Large heavy engineering	0.1518*	0.0274	Rejected	Inconsistent	Trade-Off Theory
Fertiliser	0.0278	0.0177	Accepted	–	–
FMCG	0.0863*	0.0249	Rejected	Inconsistent	Trade-Off Theory
Large electric equipment	0.2618*	0.0549	Rejected	Inconsistent	Trade-Off Theory
Mining/minerals	0.0101	0.0069	Accepted	–	–
Textile	0.0072***	0.0039	Rejected	Inconsistent	Trade-Off Theory
Large tyres	0.0268	0.0215	Accepted	–	–
Pharmaceutical	0.0082	0.0140	Accepted	–	–

Source: The authors.

Note: *, ** and *** Imply significance at 1%, 5% and 10% levels, respectively.

As far as the relationship between uniqueness and leverage is concerned, it is evident from Table 9 that on an average, uniqueness appears as an unimportant determinant of capital structure. The statistically significant negative relationship is disclosed by industries like large electric equipment and textile, which is in accordance with probable result and corresponding to the propositions of the Trade-Off Theory. Contrary to this, the statistically significant positive relationship between uniqueness and leverage is identified in the cases of large tyres and pharmaceutical industries, which is justified by the Pecking Order Theory and indicates that the sample companies in these two industries have unique or specialised products; therefore, they have to make more selling expenses in proportion to sales, which initiates greater need of finance through debt issue.

OPEN IN VIEWER

Table 9.

Relationship Between Uniqueness and Leverage

Industry	Estimated Coefficient	Standard Error	Null Hypothesis Accepted or Rejected	Consistent or Inconsistent to Expectation	Acceptance of Relevant Theory
Cement	0.1651	0.4912	Accepted	–	–
Computer hardware	–0.3380	0.7229	Accepted	–	–
Large heavy engineering	0.0255	0.4517	Accepted	–	–
Fertiliser	–0.4882	0.8026	Accepted	–	–
FMCG	–0.1462	0.2549	Accepted	–	–
Large electric equipment	–2.5698**	1.0478	Rejected	Consistent	Trade-Off Theory
Mining/minerals	–0.0706	0.2961	Accepted	–	–
Textile	–1.5738*	0.2768	Rejected	Consistent	Trade-Off Theory
Large tyres	1.1390***	0.6278	Rejected	Inconsistent	Pecking Order Theory
Pharmaceutical	0.5893**	0.2814	Rejected	Inconsistent	Pecking Order Theory

Source: The authors.

Note: *, ** and *** Imply significance at 1%, 5% and 10% levels, respectively.

OPEN IN VIEWER

Table 10.

Relationship Between Income Variation and Leverage

Industry	Estimated Coefficient	Standard Error	Null Hypothesis Accepted or Rejected	Consistent or Inconsistent to Expectation	Acceptance of Relevant Theory
Cement	–0.0461	0.0398	Accepted	–	–
Computer hardware	0.0373	0.0326	Accepted	–	–
Large heavy engineering	–0.1744*	0.02203	Rejected	Consistent	Trade-Off Theory/Pecking Order Theory
Fertiliser	–0.0464	0.0447	Accepted	–	–
FMCG	–0. 0061	0.0413	Accepted	–	–
Large electric equipment	–0.1133**	0.0499	Rejected	Consistent	Trade-Off Theory/Pecking Order Theory
Mining/minerals	–0.0828*	0.0294	Rejected	Consistent	Trade-Off Theory/Pecking Order Theory
Textile	0.0036	0.0198	Accepted	–	–
Large tyres	–0.1474*	0.0355	Rejected	Consistent	Trade-Off Theory/Pecking Order Theory
Pharmaceutical	–0.15505*	0.0482	Rejected	Consistent	Trade-Off Theory/Pecking Order Theory

Source: The authors.

Note: * and ** Imply significance at 1% and 5% levels, respectively.

The estimated regression coefficients presented in Table 10 interpret that income variation is evolved clearly not as an influencing determinant of capital structure as statistically significant, but negative impact is found in the cases of large heavy engineering, large electric equipment, large tyres, mining/minerals and pharmaceutical industries. The statistically significant inverse linkage between income variation and leverage is in compliance with the anticipated outcome and also rationalised by both the Trade-Off Theory and the Pecking Order Theory, which illustrate the probable existence of low probability of bankruptcy risk, leading to greater access to debt market for raising loan at a reasonable rate.

Recommendations

Conclusion

Profitability emerges as a significant determinant in devising the capital structure of all selected industries, except large tyres and fertiliser industries, and this strong negative relationship is supported by the Pecking Order Theory, revealing that greater proportion of profits is likely to raise the internal fund for financing future investment projects and, therefore, less dependency on external borrowings. Majority of past empirical works, such as Titman and Wessels (1988), Rajan and Zingales (1995), Booth et al. (2001), Pandey (2001), Ozkan (2001), Bevan and Danbolt (2002), Huang and Song (2002), Frank and Goyal (2003, 2009), and Shah and Khan (2007), also observe negative relationship between profitability and leverage.

The positive and statistically significant linkage between size and leverage is exhibited by large heavy engineering, large tyres and pharmaceutical industries, which are supported by the combined propositions of the Trade-Off Theory and the Agency Cost Theory as well as consistent with earlier studies, such as Pandey (2001), Huang and Song (2002), Bevan and Danbolt (2002), Antoniou et al. (2002) and Frank and Goyal (2009). Similarly, Rajan and Zingales (1995) in case of the USA, Japan, the UK, Canada and Deesomsak et al. (2004) in case of Thailand, Malaysia and Australia observed similar positive relation. But negative and statistically significant association between size and leverage is also interestingly found for computer hardware industry, which is in conformity with the Pecking Order Theory.

The positive and statistically significant connectivity between growth and leverage is observed with respect to computer hardware, large heavy engineering, fertilizer, FMCG, mining/minerals and pharmaceutical industries, which is inconsistent with expected outcome but validated by the Pecking Order Theory. This implies that internal capital is retained for financing new investment proposals, and thus asset growth is financed by debt capital.

The positive and statistically significant relation between tangibility and leverage is shown by large heavy engineering, fertiliser,, large electric equipment and textile industries, which is confirmed by the combined predictions of the Trade-Off Theory and the Agency Cost Theory and also corroborates the findings of the studies done by Rajan and Zingales (1995), Bevan and Danbolt (2002), Shah and Khan (2007), and Frank and Goyal (2009).

The inverse association between non-debt tax shield and leverage in case of mining/minerals industry is validated by the Trade-Off Theory, and it suggests that tax shields benefit of depreciation may lead to diminish the need of debt financing through savings of cash outflow. Empirical evidences in support of this theoretical prediction are established also by Ozkan (2001), Huang and Song (2002) and Deesomsak et al. (2004).

The inverse relation between liquidity and leverage is observed for the cement industry, which is consistent with both the Pecking Order Theory and Agency Cost Theory. The empirical works in support of this relation are Ozkan (2001) and Deesomsak et al. (2004). But for industries like large heavy engineering, FMCG, large electric equipment and textile, the connectivity between liquidity and leverage is interestingly found to be positively significant, which supports the Trade-Off Theory, indicating better potentiality to meet short-term obligations.

The statistically significant negative relationship between uniqueness and leverage is disclosed by industries like large electric equipment and textile, which is in conformity with the propositions of the Trade-Off Theory and confirmed by the study of Titman and Wessels (1988). Contrary to this, the statistically significant positive relationship is identified in case of large tyres and pharmaceutical industries, which is justified by the Pecking Order Theory, indicating that the sample companies in these two industries have unique or specialised products; therefore, they have to make more selling expenses in proportion to sales, which initiates greater need of finance through debt issue.

The statistically significant inverse linkage between income variation and leverage is found in the cases of large heavy engineering, large electric equipment, large tyres, mining/minerals and pharmaceutical industries, which is consistent with the Trade-Off Theory and the Pecking Order Theory and also established by the study of Drobetz and Fix (2003).

The present study eventually concludes that the Pecking Order Theory and the Trade-Off Theory mostly describe the observed relationship of independent factors with capital structure of the selected Indian industries. Agency Cost Theory of capital structure remains valid, though in lesser magnitude for explaining observed empirical results on factors of capital structure. Though most of the findings confirm the earlier findings, some interesting observations are also derived here in the Indian context, which are attempted to explain using established theories.