Why do companies voluntarily disclose? A structural equations modelling approach

Abstract

Many papers have attempted to identify why managers voluntarily disclose information about their companies. This is commonly analysed with regression methods, but these often do not account for endogeneity among common explanatory variables. This study aims to resolve the endogeneity. Data is gathered on various characteristics of 1,436 UK-listed companies. The primary models of interest are made using Structural Equation Modelling (SEM). This technique allows for correlations and causal relationships between explanatory variables, potentially solving the endogeneity problems common to regression techniques. Regression models are made as a basis for comparison to both existing literature and the SEM. Analysis of SEMs indicates that Signalling Theory is the most consistently supported explanation of disclosure across all data types used. In addition, several of the best fitting models are those for this theory. As Signalling Theory is the best explanation of disclosure tested, managers are primarily providing voluntary disclosure of information in order to demonstrate the company’s recent financial performance (and by extension, the managers’ capabilities) in order to attract more investment at a lower required rate of return. Results also indicate some power to Legitimacy Theory, suggesting that disclosure is to some extent provided in order to improve the company’s image.