An Empirical Test of The Dual-Path Impact of mental Accounting Classification on Investment Behavior Bias

Abstract

This study focuses on Chinese investors and empirically examines the dual-path impact of "funding source" and "funding use" classifications in mental accounting on investment behavior biases, with mental accounting theory as the core. Based on a questionnaire survey of 250 Chinese residents with investment experience, methods such as linear regression, ordinal Logistic regression, mediation effect, regression, and moderation effect tests were employed. The results show that different funding sources have significant differences in their impact on asset allocation risk preferences. Funding use classifications affect biases such as the disposition effect through risk perception and evaluation frequency. Risk perception plays a partial mediating role between funding use and risk mismatch, and high-income groups are more affected by mental accounting labels. This study constructs an integrated framework of "fund classification - mental accounting - decision bias," which has innovations are: first, it is the first to integrate fund sources and uses into a unified analytical model, revealing their interaction effects; second, it deepens the micro-mechanism of mental accounting's impact on investment decisions through the "cognition-behavior" dual-layer mediation path; third, it incorporates Chinese market features such as high savings rates and rigid expenditure expectations.