Building an Idiosyncratic Risk Model for Asset Pricing in Emerging Markets (2010–2024)

The research problem The classic models for asset pricing (CAPM) capture only systematic risk and ignore firm-specific volatility. However, in developing countries where diversification and information asymmetry are low, the idiosyncratic risks can be priced. We investigate whether (accounting-based) measures, specifically, accrual quality and anomalies, affect firm-specific volatility and the extent to which idiosyncratic risk mediates their links to asset pricing. Institutional setting The paper relies on Indonesia as a proxy for an emerging market with concentrated ownership, changing levels of disclosure, and economic shocks such as the 2020–2021 COVID-19 pandemic era. It is in the institutional environment that a judgment can be made on whether accounting quality and firm-specific volatility affect asset pricing under market inefficiencies. The hypothesis testing and model establishment (1) greater persistence of accruals, (2) non-current and abnormal levels of accruals, and (3) idiosyncratic risk all moderate or mediate the relation between the level of accruals and valuation. In order to investigate such claims, we expand the Fama–French five-factor model by including a set of accounting-based factors in building our modified idiosyncratic risk model. Adopted methodology Based on a sample of 110 manufacturing companies listed in the capital market in Indonesia Stock Exchange (2010–2024 period), this study estimates firm-specific residuals from Fama–French five-factor regression. These residuals are idiosyncratic risk, endogenously modelled dynamically with persistence (GARCH(1, 1)) and asymmetry (EGARCH). To do that, a SEM-AMOS technique is applied to test direct and indirect relationships between accrual components, idiosyncratic risk, and the valuation measures: price-book value (PBV) and price-earnings ratio (PER). Findings and implications The results evidence a decrease in idiosyncratic volatility associated with current operating accruals, while an increase is observed in the case of non-current and abnormal accruals. Financial accruals do not significantly contribute. The results suggest the existence of priced idiosyncratic risk, suggesting that in EM, investors incorporate firm-specific volatility into return expectations. Volatility series expose the structural change present during COVID-19, though asymmetry remains as a pervasive characteristic. The results contribute to the literature of asset pricing by integrating firms’ accrual quality and firm-specific volatility in a multifactor model. On a practical level, the study underscores the need to enhance earnings visibility to reduce value risk and keep an effective market in developing countries. JEL Classification: G12, G15, M41, C58