The Determinants of the Capital Structure of Listed Companies in Bangladesh: An Assessment of Total Factor Productivity

Abstract

This study explores the key factors influencing capital structure (CS), with a focus on the impact of total factor productivity (TFP) as the primary indicator of firm productivity in explaining capital structure choices. Despite extensive research on CS decisions since Modigliani and Miller's foundational work in 1958, no definitive theory has emerged to guide optimal financial policy. This research seeks to further examine the relationship between TFP and various forms of debt, specifically total debt (TD), short-term debt (STD), and long-term debt (LTD). The comprehensive analysis investigates how a firm's total factor productivity (TFP), firm-specific characteristics- financial constraints, and the cost of debt affect different debt structures in the manufacturing firms of Bangladesh. The main variable, total factor productivity (TFP), measures the overall efficiency of resource utilization in production, capturing how effectively inputs like labor and capital are combined to yield output. TFP illustrates the portion of output not explained by input quantities, reflecting the effectiveness of input usage, technological advancements, and managerial prowess. It showcases the output-to-input ratio, revealing the efficiency of production. TFP captures the impact of technological progress, often resulting in heightened productivity, and reflects managerial efficiency in organizing production processes. It is influenced by resource allocation, emphasizing the importance of directing resources to their most productive uses. i TFP growth is a key driver of long-term economic growth, enabling higher output without a proportional increase in inputs, thus improving living standards over time. The variations in TFP values can reflect differences in productivity and performance across regions, firms, and industries. Policymakers often use TFP as a guide for economic policies that promote innovation and create a supportive business environment, contributing to overall business development. This research estimates TFP using the Solow Residual method, which, in the context of the Solow Growth Model, provides insights into efficiency and technological progress. By identifying the relationship between TFP and CS, this study seeks to understand the broader implications of technological efficiency on a firm's debt structure. Additionally, it considers firm-specific characteristics such as size, age, tangibility, liquidity, volatility, and non-debt tax shields, along with two key firm heterogeneity factors: financial constraints and the cost of debt. These factors may affect a firm's access to capital. By examining various factors—including TFP, financial constraints, firms’ internal characteristics, and leverage costs—the study offers a detailed analysis of the determinants of CS. In doing so, it provides a fresh perspective on these dynamics within the context of Bangladesh. This study employed the SA index to assess the extent of financial constraints affecting firm behavior within the sample. The SA index serves as an evaluative indicator for financial constraints, categorizing them into two levels based on the quantiles of the index. The variable 'fchigh' is a dummy variable that takes the value of 1 if the SA index is above the 50th percentile and 0 otherwise. ii Additionally, the research introduced the cost of debt as another firm heterogeneity factor in the regression model. The cost of debt was measured using the interest rate. The variable 'Cost' represents leverage cost, which was categorized into two levels based on the quantiles of the institutional development index. The dummy variable 'Cost high' is assigned a value of 1 if the cost of leverage is above the 50th percentile and 0 otherwise. To address endogeneity and firm-specific differences, this research used the two-step system Generalized Method of Moments (GMM), as recommended by Arellano and Bond (1991). This method helps mitigate simultaneity issues, such as omitted variable bias and reverse causality, providing more accurate results compared to Ordinary Least Squares (OLS) and fixed-effects models. The Hansen test confirmed the validity of the instruments used in the GMM method, with a p-value above 0.05, ensuring that the results were unbiased and efficient by addressing simultaneity concerns. The study collected data from 155 manufacturing firms across 10 industries listed on the Dhaka Stock Exchange (DSE) from 2012 to 2022, resulting in a balanced panel dataset of 1,550 observations. Only firms with complete information for the entire period (2012-2022) were included, while those with incomplete data were excluded from the analysis. This research is a pioneering attempt to analyze and strengthen the argument regarding the relationship between total factor productivity (TFP) and capital structure (CS) choices for Bangladeshi firms. The study assesses the connection between TFP and CS using three (3) separate regression models. Each model examines three distinct debt ratios—total debt (TD), short-term debt iii (STD), and long-term debt (LTD)—as the dependent variables. The baseline regression model (1) considered nine firm-specific variables: growth, non-debt tax shield, liquidity, tangibility, volatility, firm size, firm age, return on assets, and the key variable TFP, analyzed for the dependent variables TD, STD, and LTD. The results of regression model (1) revealed that TFP is a significant factor influencing the CS decisions of listed manufacturing firms in Bangladesh. Econometric analysis showed that TFP plays a substantial role, indirectly affecting the ratios of total debt (TD) and long-term debt (LTD), but it does not exhibit a significant link with short-term debt (STD). In addition to TFP, the study incorporated two firm heterogeneity factors—financial constraints and the cost of debt—into two additional regression models (2) and (3) to more comprehensively analyze and explain the relationship between TFP and CS. The model incorporating financial constraints used the SA index to measure a firm's financial difficulties, representing a novel approach. Firms were then categorized into high and low financial constraint groups. The analysis of regression model (2) also includes the original nine variables, along with the financial constraint variable (fchigh) and the interaction between TFP and fchigh. The results of regression model (2) indicated that independently financial constraints are not significantly correlated with short-term debt (STD) and long-term debt (LTD) measures within the companies. However, the study found that firms facing higher financial constraints exhibit a stronger relationship with total debt (TD) compared to those with lower financial constraints. This highlights the importance of financial constraints as a significant factor for manufacturing firms, suggesting that firms with financial constraints are more iv sensitive in their decisions regarding total debt (TD) only. Furthermore, the interaction between TFP and high-level financial constraints had no impact on any of the three leverage measures (TD, STD, LTD). Third model included two (2) more variable cost of debt and the interaction of TFP and cost of debt (COSTHIGH) along with the original (9) variable of model (1). In regression model (3), the analysis demonstrates that a firm's cost of debt has a significant and positive impact on both total debt (TD) and long-term debt (LTD), showing a positive correlation. High productivity firms signal their ability to access diverse financing options and effectively manage funding through retained earnings. This indicates that manufacturing firms, even when faced with higher debt costs, are inclined to secure loans, as the higher cost serves as a signal of their efficiency and ability to secure both TD and LTD. Furthermore, the research reveals a significant negative interaction effect between the cost of debt and total factor productivity (TFP) concerning TD and short-term debt (STD). However, this joint variable exerts a positive impact on LTD. This underscores the sensitivity of capital structure (CS) in Bangladeshi manufacturing firms to the combined influence of the cost of debt and TFP. High-productivity firms typically prioritize internal financing for TD and STD, aligning with the pecking order theory. In contrast, for LTD, these firms tend to pursue loans at higher costs to capitalize on superior investment opportunities, supporting the trade-off theory. The empirical findings suggest that firms with higher TFP usually have better investment opportunities and are more willing to offer higher interest rates to lenders. This is consistent v with the idea that more productive firms are better positioned to generate higher returns, which allows them to cover the costs associated with debt. Therefore, TFP has a stronger impact on CS for firms facing higher leverage costs. The relationship between TFP and CS is particularly pronounced in scenarios where leverage costs are high, emphasizing the role of leverage cost as a key factor affecting the link between TFP and leverage in manufacturing firms. Higher leverage costs increase the sensitivity of TFP to CS. The study observes that TFP is indirectly associated with both TD and LTD in Bangladeshi firms. Firms with high productivity tend to prioritize internal financing, favoring retained earnings over external debt. This preference suggests that Bangladeshi companies are inclined to favor equity over debt, which aligns with the Pecking Order Theory. Firms with higher productivity and profitability are more likely to opt for equity financing before issuing debt. Thus, TFP, measured by the efficient use of input factors, plays a crucial role in shaping capital structure (CS) decisions. The study also recommends prioritizing technological advancements to boost productivity, which would encourage greater reliance on internal financing. Additionally, factors such as profitability, asset tangibility, and liquidity have an inverse effect on the debt structure, whereas firm age and size positively influence debt decisions. Moreover, the institutional and political environment can shape the relationship between productivity and financing decisions, highlighting the need for future research to explore these dynamics further. It also highlighted that profitability, tangibility, and liquidity are the significant determinants influencing the theories of CS; however, these factors exhibit an inverse relationship with vi STD and LTD, dependable with the pecking order theory. Additionally, the variables of firm age and firm size show a direct relationship with the debt ratio of firms. This research aims to offer valuable insights into the financial decision-making processes of firms, emphasizing the importance of optimal debt management and its influence on productivity and technological progress. The findings are expected to contribute to the academic literature on financial management and provide practical implications for policymakers, investors, and corporate managers. The significant contributions of this study enrich contemporary research on the capital structure of firms in Bangladesh. vii