Comparative Analysis of Emotional Biases in Investment Decisions

Abstract

Background: Emotional Biases have influence in investment decision of SACCO members. These biases are overconfidence, loss aversion, and herd behavior. We seek to establish empirically to what extent each emotional bias contributes to investment decisions and identify which emotional bias has an overarching influence.

Objective: To identify influence of each emotional bias and establish which among them has overarching influence.

Method: Perception based data was collected among 284 Imarisha SACCO members in Kericho County, Kenya between 28th to 31st July, 2025. The data collected was self-filled questionnaire using KoboCollect. The data was checked for completeness, entered in excel, cleaned, coded and analyzed. First the data was tested for normality before Principal Component Analysis (PCA) to compare the strength of the three dimensions; PC1, PC2, PC3. Results were obtained and interpreted.

Results: The principal component analysis revealed that herd behavior (PC1) emerged as the most significant emotional bias, loading strongly at 75.2% and explaining the largest share of variance at 58.26%. Overconfidence bias (PC2) also featured prominently, with a positive loading of 70.1%, and together with herd behavior, the two accounted for 81.26% of the total variance, underscoring their combined influence on investment decisions. Conversely, loss aversion bias (PC3) loaded negatively (-76.4%), indicating that it primarily acts as a constraint, discouraging SACCO members from making investment decisions. Overall, herd behavior exerts a disproportionate influence compared to other biases, while loss aversion uniquely inhibits investment participation.