Since the pioneering study of Modigliani and Miller (MM) in 1958, the capital structure has been one of the most researchable issues in the field of Financial Management, yet little evidence is available where the emerging markets are con- cerned. Among the emerging markets in the south Asian region, the Sri Lankan capital market has enjoyed a long history of more than a hundred years and com- prehensive studies on the market have only grown with the economic reforms of recent decades. These reforms enable the market to attract local and foreign in- vestments and revitalize the stock exchange. This study empirically investigates the corporate capital structure of the listed firms in the Colombo Stock Exchange (CSE), its determinants and the industry variations.
According to the argument made by MM in their seminal paper, in a frictionless world the valuation of the firm is independent from its capital structure under three propositions.1 Further, their assumptions were based on a perfectly functioning capital market where there are no transaction or bankruptcy costs, no taxation and the productive activity of the firm is independent of it's method of financing. In such a market, external funds are supposed to be a perfect substitute for internal financing. However, recent researchers have added market imperfections such as bankruptcy costs (Kim, 1978), availability of tax shields (DeAngelo and Masulis, 1980) and agency costs (Jensen and Meckling, 1976) to their models, and have emphasized that an optimal capital structure may exist in the real world. Bradley, Jarrel and Kim (1984) argued that agency cost and bankruptcy cost could be treated as partial determinant of lever- age. These imperfections cause the costs of alternative financing sources to vary and upset and realize that a firm's capital structure, but also the entire industry.
Singh and Hamid (1992) and Singh (1995), in their researches on corporatecapital structure in developing countries, concluded that firms located in develop- ing or transitional economies rely heavily on equity than levered capital in financ- ing growth relative to the firms in industrial or developed economies. Several reasons explain why one might expect different financing patterns between the firms in developed economies and their counterparts in emerging markets. Singh and Hamid (1992) show that the capital markets are less developed in developing countries, and the range of financial instruments available are relatively narrow with many constraints on financing decisions than in the developed countries. On the other hand, most of the firms in emerging markets are recently privatized, state owned entities. This heritage consequently affects the corporate objectives and strategies that may not go in line with the private sector motives seen in industrial countries.
Prasad et al. (2001) identify corporate strategy as one of the most signif- icant factors in determining capital structure in developing context. Cobham and Subramaniam (1998) argue that lack of rigorous accounting standards and audit controls in developing countries may create relatively higher information asym- metry among stakeholders than in major developed countries.
Sri Lanka has been liberalizing its economy since 1977, by gradually moving away from a highly state-controlled economy to a private sector-led econ- omy. The transformation process is dealing with revolutionizing its economic structure and mechanism, and opening up the economy with a view to create a market controlled economy. The rapid emergence of active capital market af- ter economic liberalisation has led to serious market reforms. Policies, pro- grams and regulations were adopted, which in turn created a dramatic increase in non-conventional thinking and more flexible governance. This creates the need for keeping up with the financial developments taking place outside the region, particularly in bond and securities markets. Although a significant change affected the Sri Lankan capital market with the successive waves of privatisation, it remained small due to a number of reasons. However, having many hindrances, the market managed to position as the second most promis- ing market in the world. Like firms located in other transitional economies where the external sources of finance are narrow and restricted, Sri Lankan firms are also experiencing lack of accessible sources to meet their capital requirements to finance their po- tential investment opportunities. The prevailing macroeconomic policies im- pose a number of constraints on the development of capital market, discouraging use of levered finance in the firm's capital structure. Although not as evident in the emerging market context, the recent trend towards globalization and inter- nationalization of financing sources made the studies more significant, partic- ularly in transitional economies.
Any study on corporate capital structure naturally pays attention to the fac-
tors influencing the firm's decision on financial leverage. Thus, this paper is de- signed to address the usage of leverage finance relative to equity finance in the capital structure of Sri Lankan-listed firms. More specifically, to investigate the significance of leverage in firm's capital structure, what determines leverage and how it differs across different industry sectors with diverse characteristics.
The contribution of this study is twofold. First, this study may fill the gap of
empirical evidence in the context of an emerging market in Asia, provided that most empirical studies on capital structure decisions are done in developed markets. Second, the findings add to empirical literature some evidences on how the capital structure choice varies across difference industries by examin- ing the firms located in one of the highly transitional economies in Asia.
The pioneering work of Modigliani and Miller (1958) illustrates that the valuation of a company will be independent from its financial structure under certain key assumptions. Internal and external funds may be regarded as perfect substitutes in a world where capital markets function perfectly, where there are no transactions or bankruptcy costs, no distortionary taxation, and the productive activity of the firm is independent of its methods of financing. Once these fundamental assumptions are relaxed, however, capital structure may become relevant. Additionally, firms may find that there are restrictions to their access to external financing, and the costs of alternative forms of external finance may differ. Under such market imperfections, firms will attempt to select levels of debt and equity in order to reach an optimal capital structure.
This study attempts to extend our knowledge of capital structure and its determinants in listed UK companies. In their study of capital structure in the G-7 economies, Rajan and Zingales (1995) find gearing in the UK to be positively related to tangibility (the proportion of fixed to total assets) and the size of the company (logsales), but negatively related to the level of profitability and the market-to-book ratio. In this study we build on the UK component of the Rajan and Zingales study, by testing the sensitivity of the determinants of capital structure to various gearing measures and their sub-elements. We find that Rajan and Zingales' results are highly dependent upon the precise definition of gearing being examined. Having found evidence of significant definitional dependence, we attempt to gain a fuller understanding of our results by further sub-dividing the debt element of our gearing measures, in order to test the relation of each of the elements to our explanatory variables. We find that the results of this analysis differ significantly depending upon whether we consider short or long-term debt elements. Consequently, our results highlight the sensitivity of the analysis to which form of debt is being considered. Given the predominance of short-term debt forms in corporate financial structure, we suggest that analyses based solely upon long-term forms of debt provide limited insight into the mechanisms which operate in the financial and corporate sectors.
Empirical analysis of capital structure is fraught with difficulty, and as argued by Harris and Raviv (1991), "The interpretation of results must be tempered by an awareness of the difficulties involved in measuring both leverage and the explanatory variables of interest". In this paper we focus on the difficulties in defining gearing. We note, however, that many of the potential explanatory attributes are often, at best, imperfectly reflected by the variables observed in corporate accounts data. Hence Titman and Wessels (1988) note that the required explanatory variables may frequently be imperfect proxies for the desired corporate attributes, so inducing 'errors-in-variables' problems to regression analysis. Moreover, complex attributes frequently are not reflected by a single, unique, explanatory variable, nor do variables reflect a single attribute. As a result, the conclusions which may be derived from empirical analysis of corporate financial structure have the potential to become 'definition-dependent', as they may rely upon the researchers' choice of imperfect proxy variables.
This paper examines the issue of corporate financial structure and its determinants from three distinct perspectives. We utilise corporate accounts data for the United Kingdom in an attempt to replicate the analysis of Rajan and Zingales, so permitting us to test their conclusions with a larger sample of firms. We conduct our analysis on data for 1991, the same year as Rajan and Zingales. By using data from the same year, we ensure direct comparability with their work and limit the possibility that any differences in the results may be due to variations in the level of gearing over time, rather than indicating definitional dependence1. Secondly, given the preceding discussion, we implement the same analysis with alternative definitions of the dependent variable, gearing, in order to examine the robustness or otherwise of the results of Rajan and Zingales. As noted above, we then seek to explore this finding of definitional-dependence further, and find that the determinants of gearing vary significantly depending on the nature of the debt sub-element being analysed.
