eThesis - School of Business electronic theses

PURPOSE OF THE STUDY

The banking market is hypothesized of having a tendency to ration and constrain lending to firms, while the public debt market (bond market) is hypothesized to afford borrowers with added supply of debt capital thereby having the effect of relaxing credit constraints. This study examines the degree to which bond market access of firms is correlated with the leverage of firms to uncover whether bond market access has a significant effect on leverage. Two-stage least squares regressions employing instrumental variables are employed to robustly uncover the exogenous effect of bond market access on leverage. The research questions ask whether bank dependent firms are credit constrained and whether access to the bond market relaxes possible credit constraints? Also, important theories and determinants of firm leverage are reviewed and examined in regressions providing for an understanding of what determines firm leverage.

DATA

The sample of this study includes all of the Finnish and Swedish publicly listed companies in the OMXH and OMXS exchanges excluding companies included in the GICS groups of 4010-4040 (Financials). Yearly observation from years 2000-2009 are observed and a maximum sample size of 4690 firm-year observations is obtained. The Thomson ONE Banker service and the Datastream bond issuances database are used as data sources.

RESULTS

Firms predicted to have bond market access are found to have significantly higher financial leverage compared to firms without predicted bond market access even after controlling for leverage demand factors, industry, time period and for country. This difference ranges from 3.41 to 8.85 percentage points in market leverage which translates to 22.4% and 58.6% more relative debt when measuring from average leverage. Without the control for endogeneity, bond market access seems to significantly increase the debt of firms. However, controlling for endogeneity of bond market access makes the positive correlation statistically insignificant albeit remaining positive. Evidently, the degree of credit constraints in the bank lending markets of Finland and Sweden and the degree to which bond market access relaxes credit constraints are insufficient to be observable in my empirical tests.

The leverage demand factors in the regression models are found to be highly explanatory factors of leverage. At best, 37.8% of all variation in firm leverage is explained in regressions. Significant findings for the factors of tangibility of assets, profitability and growth opportunities are found to support the trade-off- and pecking order theories of capital structure.