Essays on Capital Structure

Research output: ThesisDoctoral Thesis (compilation)


Chapter 2 analyses the relationship between capital structure and industry classification of Swedish firms. This study also analyses the determinants of the capital structure at the industry level. Since the assumption required for a parametric analysis of variances are not satisfied, a nonparametric method is applied to test for cross-industry differences in capital structure. The results show that leverage varies significantly across industries. However, the multiple comparison analysis reveals that the significance of the result is mainly due to the construction and real estate industries. Analysis of some determinants of the capital structure at the industry level, with a descriptive method, verifies that the variation of debt ratios has almost the same pattern as the variation of the collateral value of the firms’ assets.

Chapter 3 presents an analysis of the relationship between a firm’s characteristics and its choice of capital structure. A large number of hypotheses of different classes of theories are tested. To reduce deficiencies resulting from using a single indicator for attributes which cannot be directly observed, I apply the LISREL method of structural equation modeling. Furthermore, the Extreme Bound Analysis is applied to test for fragility of the estimated coefficients to changes in the number of explanatory variables of the model. The results show that growth, size, collateral value of the assets, and managers’ shareholding positively affect firm leverage, while the effect of profitability on leverage is negative.

Chapter 4 studies determinants of the firm leverage during the economic downturn in Sweden 1989-1992. I use a panel data framework to capture both cross-sectional and intertemporal relationships between firm leverage and its determinants. In the first part, I use a descriptive analysis to study variations in mean leverage ratios during the period 1988 to 1992. In the second part, a regression analysis is applied to find variables which explain variations in firm leverage during the recession. The model allows for firm and time effects as well as time varying coefficients. The consistency and efficiency of the estimators regarding problems such as self selection bias and heteroscedasticity are taken into account. Among other things, I find that dividend payments and stock returns affect firms’ leverage negatively.

Chapter 5 studies the relationships between firm performance and leverage during the economic recession in Sweden and tests the hypothesis that highly leveraged firms lose market shares to their less leveraged rivals in an industry downturn. I apply both parametric and semiparametric regression methods to analyze the relationships between firm performance and leverage. I find that the highly leveraged firms in distressed industries face relatively lower sales growth and stock returns, while seemingly able to retain a relatively higher growth in profitability. The finding suggests that the decline in sales of the highly leveraged firms might be a result of managers’ preferences to decrease the activity of product lines with low profitability.
Original languageEnglish
Awarding Institution
  • Department of Economics
  • [unknown], [unknown], Supervisor, External person
Award date1997 Apr 8
Publication statusPublished - 1997

Bibliographical note

Defence details

Date: 1997-04-08
Time: 13:00
Place: Holger Crafoords Ekonomicentrum 1, EC135

External reviewer(s)

Name: Eckbo, Espen
Title: [unknown]
Affiliation: [unknown]


Subject classification (UKÄ)

  • Economics

Free keywords

  • economic theory
  • econometrics
  • Economics
  • Extreme Bound Analysis
  • Selection bias
  • Lisrel
  • Semiparametric
  • Nonparametric
  • Financial distress
  • Panel data analysis
  • Leverage
  • Capital structure
  • economic systems
  • economic policy
  • Nationalekonomi
  • ekonometri
  • ekonomisk teori
  • ekonomiska system
  • ekonomisk politik


Dive into the research topics of 'Essays on Capital Structure'. Together they form a unique fingerprint.

Cite this