MODELING FINANCIAL PERFORMANCE OF MANUFACTURING FIRMS ON CAPITAL STRUCTURE. DOES THE PANEL DATA MODEL USED MATTER? EVIDENCE FROM THE NAIROBI SECURITIES EXCHANGE
MODELING FINANCIAL PERFORMANCE OF MANUFACTURING FIRMS ON CAPITAL STRUCTURE. DOES THE PANEL DATA MODEL USED MATTER? EVIDENCE FROM THE NAIROBI SECURITIES EXCHANGE
Akali James Agembe - Department of Accounting & Finance, Kisii University, Kenya
Prof. Christopher Ngacho - Associate Professor in Operations Management, Kisii University, Kenya
Dr. Wafula Joshua Chesoli - Senior Lecturer in Accounting & Finance, Kisii University, Kenya
ABSTRACT
Manufacturing firms contribute a significant proportion to Kenya’s Gross Domestic Product (GDP), putting them at the center of her development. To optimize production, most manufacturing stakeholders have taken cognizance of the capability of econometric panel data models to maximize output on minimal input. Yet, a gap exists on whether the panel data used in modeling financial performance on capital structure matters. Therefore, this research used the financial performance-capital structure nexus in listed manufacturing firms trading at the Nairobi Securities Exchange (NSE) to establish whether the panel data model in use mattered. Using 85 observations drawn from 14 firms, and covering the period 2016 to 2022 inclusive, the study compared the pooled Ordinary Least Squares (OLS), the random effects (RE), and the fixed effects (FE) models. The study revealed that parameters such as model restrictiveness, estimation consistency and efficiency, and temporal variations dictate the model to be used, confirming that the panel data model indeed matters. Retrospectively, the pooled OLS model suits situations without unobservable entity-specific effects, the RE model suits situations where differences across firms do not correlate with the predictors, and the FE model is preferred when some time-invariant characteristics such as company culture are omitted. The significance of this finding to manufacturers is that robust decision making regarding leveraging financial performance of manufacturing firms on capital structure is a function of careful consideration of available panel data models as defined by existing parameters. Future studies can strengthen this finding by including dynamic panel data models.
