A0310
Title: Sector-specific interrelationship between capital structure and sales growth: A Bayesian machine learning approach
Authors: Kousik Guhathakurta - Indian Institute of Management Indore (India) [presenting]
Sujay Mukhoti - Indian Institute of Management Indore (India)
Abstract: The aim is to investigate the sector-specific dynamics between capital structure and firm growth using a Bayesian machine learning framework. Traditional models often impose rigid parametric structures and struggle with endogeneity and model uncertainty. In contrast, the approach flexibly accommodates structural breaks, nonlinearities, and heterogeneous effects across industries by allowing the model space itself to evolve with the data. Findings reveal stark inter-sectoral differences in the leverage-growth nexus. In certain industries, higher financial leverage is positively associated with sustained sales growth, likely due to greater scale economies and access to external financing. Conversely, firms in some other sectors exhibit optimal growth under moderate leverage levels, beyond which excessive debt constrains operational flexibility and performance. These patterns highlight the importance of aligning financial strategies with sectoral characteristics such as capital intensity and competitive dynamics. Results provide a novel perspective for corporate managers, investors, and policymakers, emphasizing the need for industry-contingent capital structure decisions. The Bayesian machine learning framework offers a powerful and generalizable tool for understanding complex financial relationships in diverse economic contexts.
