Capital Structure and Performance at Nairobi Stock Exchange (NSE)

Abstract

In Kenya, despite the importance of corporate financial decisions there is deficiency of studies on capital structure that managers would rely on in making financial decision. Most studies on capital structure often assume that causation is from performance to capital structure; but there is evidence that the causation may also be reverse, hence the possibility of bias. The research question is: Does capital structure and performance affect each other? Design/methodology/approach: The sample is from 56 companies listed at NSE during 1996 to 2003. Different accounting and market measures of capital structure and performance are calculated and used to determine the causality between two latent variables, capital structure and performance. This study employs panel data to enhance the number of observations. Since the interest in this study is to compute the (simultaneous) relationship between five measures of performance with five measures of capital structure, canonical correlation is used. Findings: The results show that leverage has effect on performance only to the extent that long term debt to equity ratio andtotal debt to total assets ratio are used as measures of leverage, the correlations are too low. The other finding is that performance measurements has some weak predictive power for capital structure variables only to the extent that EBITto total assets, return on total assets, return on equity are used as performance indicators. Research limitations/implications: The low correlations could be due to inclusion of financial institutions in the sample and the results might be different when they are excluded. It could also be due to the quality and reliability of information contained in financial statements. The data may also indicate that managers may not be using information from appropriate indicators to make financial decisions. Originality/Value of paper/Further Research: This study point out how financing decisions or lack of it is at NSE. Such findings are useful to manager and researchers in designing financing strategies. The next logical steps would be to find out from managers how they actually make financing decisions and to evaluate the models they employ; use the capital structure and performance variables identified in this study as describing the relationship in building the relationship; and exclude financial from the sample to measure information loss, i.e. determine the change in canonical correlation.