Abstract:
The purpose of this study is to find whether the Static trade-off theory or the Pecking order
theory provides a better explanation of the capital structure patterns of listed manufacturing
firms in Sri Lanka. Data were collected from randomly selected 20 firms for 6 consecutive
years from 2011 to 2016. In order to test the tradeoff theory, tangibility, profitability, non-debt
tax shields, volatility and firm size were considered as response variables, and debt ratio was
used to measure the capital structure. The simple pecking order model originally developed
by Shyam - Sunder and Myers (1999) and modified by Frank and Goyal (2003) was applied to
test the pecking order theory in this study. To assess the application of pecking order theory,
financial deficit was considered as an explanatory variable and change in long term debt was
considered as dependent variable. Hypotheses derived from the static trade-off and pecking
order were tested using an OLS regression model. Results revealed that there is a significant
positive relationship of tangibility and non-debt tax shields with the capital structure. Further,
the results revealed that there is no significant relationship of profitability, volatility and
firm size with the capital structure. Therefore, it can be concluded that trade-off theory is
supported by tangibility and non-debt tax shields. The study finds that the beta coefficient
value of pecking order model is not close to 1. Therefore, validation of pecking order theory
is weak in selected Sri Lankan companies. This indicates the pecking order theory is not
applicable to capital structure of manufacturing firms in Sri Lanka and static trade-off is more
applicable for deciding capital structure in this respect. The results of this study are helpful
to policy makers in defining policies and procedures to ensure reliable and easy accessibility
to different sources of finance for Sri Lankan listed manufacturing companies. It suggested to
extend this study to other sectors of Colombo Stock Exchange based on a large sample
