Title
Does the Choice of Firm-Performance Metric Matter? Bayesian Evidence on Macro Risks from the US Market
Authors
Abstract
Purpose: Organizational decision-making depends critically on selecting performance metrics that accurately reflect firms’ economic conditions. Choosing a suitable firm-performance metric and understanding its determinants are significant for managers, investors, and scholars. In the literature, researchers often employ a single firm-performance measure, the accounting-based one, and then implicitly assume that empirical findings are invariant to performance measurement. This study challenges that assumption and examines whether the choice of firm-performance metrics, including the accounting-based ROA, a hybrid Q-based ratio, and the market-based excess stock return, affects conclusions about the relationship between macro risks and firm performance via a Bayesian framework and the macro risk asset pricing model. Additionally, the relationship between macro risks and metrics is scrutinized.
Methodology: Using a sample of 432 S&P 500 firms from 2007 to 2022, this study investigates the effects of key macroeconomic risk factors, including market risk premium, prime rate risk premium, long-term government bond yield risk premium, and exchange rate movements. A Bayesian estimation framework with Gibbs sampling is employed to account for parameter uncertainty and improve robustness. Model performance is assessed using Bayesian goodness-of-fit measures and posterior statistical inference.
Findings: The findings indicate substantial differences across performance metrics. The Q-based and excess stock return metrics outperformed ROA in terms of model fit or explanatory power. The excess stock return displays the highest Bayesian R-squared, followed by the Q-based ratio. Importantly, the macro risks exert statistically and economically significant effects on firm performance, although the magnitude and direction of these effects vary across performance measures. The findings suggest that performance metric selection materially influences empirical inferences about firm performance and highlight the importance of incorporating macro risks into performance evaluation.
Originality: This study contributes to the macro-finance and firm performance literature by showing that conclusions regarding macro risks are sensitive to the choice of performance metric. By integrating accounting-based, hybrid, and market-based measures within a unified Bayesian framework and a macro risk asset pricing model, the study provides additional evidence that market-oriented and hybrid performance metrics more effectively capture firms’ exposure to macroeconomic shocks.
Practical implications: This study has important practical implications. An effective choice of performance metric can provide a better approach to managers, investors, and scholars in evaluating firm performance, the cost of equity, resource allocation, dividend policy, and investment decisions. It also highlights the effectiveness of the macro risk asset pricing model in the literature. The use of the Bayesian framework, a robust Bayes estimator, and MCMC sampling offers a robust methodology for addressing parameter uncertainty in this field of study.
Keywords
Firm performance, ROA, Tobin’s Q, excess stock return, Bayes
Classification-JEL
E43, E44, G12
Pages
261-291
How to Cite
Phuoc, L. T., & Dao, T. T. T. (2026). Does the Choice of Firm-Performance Metric Matter? Bayesian Evidence on Macro Risks from the US Market. Advances in Decision Sciences, 30(4), 261-291.