Meta-Analysis on the Role of Control Variables in the Results of Studies Related to the Effect of Information Asymmetry on the Cost of Equity Capital

Objective: Empirical evidence regarding the impact of information asymmetry on the cost of equity capital reveals a persistent conflict within the literature. Some studies have identified a positive relationship, while others report it as negative or insignificant. A significant factor contributing to this divergence is the role of control variables. Although there exists a wide range of potential control variables, the inclusion of certain variables may distort the results, whereas others can enhance the accuracy of the research findings. This meta-analysis aims to investigate which control variables, based on those repeatedly employed in prior studies, significantly influence the regression coefficient in comparison to their exclusion.

Methods: This meta-analysis synthesizes the findings from 35 studies published between 1986 and 2022, encompassing a total of 198 tests, to illuminate the intricate relationship between information asymmetry and the cost of equity capital. The analysis was conducted in several stages: Information asymmetry was established as the independent variable, while the cost of equity capital served as the dependent variable. Control variables included the book-to-market ratio, market beta, company size, type of ownership, yield fluctuations, equity, leverage, and growth. Comprehensive searches were conducted in prominent databases such as Science Direct, Emerald, Google Scholar, SSRN, and ResearchGate, resulting in the extraction of 188 articles based on pertinent keywords. The articles were screened for alignment with the research hypotheses, focusing on studies that provided effect size information and employed correlation methods. Consequently, 35 studies published between 1986 and 2022, incorporating 198 effect sizes, were selected for meta-analysis. Relevant general information and details about effect sizes were meticulously extracted from the selected articles. For each sample, the effect size, denoted as r, was calculated. Effect sizes were aggregated using weighted means, accounting for sampling error. Finally, the homogeneity of effect sizes was assessed.

Results: The empirical examination of the relationship between information asymmetry and the cost of equity capital has revealed notable divergences. To investigate this issue, a meta-analysis approach was employed, highlighting the varying effects of control variables on this relationship. The findings indicate considerable variation in effect sizes across the studies analyzed. Consequently, a random effects model was utilized to calculate the cumulative effect size. Control variables play a crucial role in shaping the relationship between information asymmetry and the cost of equity capital. Specifically, the inclusion of control variables such as the book-to-market ratio, market beta, firm size, and ownership type tends to diminish the strength of this relationship. In contrast, factors such as stock return volatility, leverage, and firm growth do not significantly alter this relationship. A robustness analysis further corroborates these preliminary findings.

Conclusion: Empirical researchers frequently encounter challenges related to the selection of appropriate control variables when regressing the cost of equity capital on information asymmetry. In this context, a critical question arises: which variables should be considered as control variables to enhance the validity of experimental results? This study identifies four control variables that significantly influence the targeted relationship: the book-to-market ratio, market beta, company size, and ownership type. The inclusion of these variables in the regression equation markedly alters the outcome of the analysis.