This study aims to examine how perceived control is associated with retail investor behavior in high-variance cryptocurrency markets. It develops a moderated-mediation framework in which overconfidence, operationalized as calibration error, transmits the effects of perceived control to trading frequency, portfolio concentration, risk-taking and net investment performance, while realized volatility conditions the strength of these relationships.
This study uses a theory-calibrated simulated panel of 1,000 heterogeneous investors. Standardized linear models, interaction terms and bootstrapped indirect-effect estimates are used to assess internally coherent, design-implied relationships under explicitly maintained assumptions. The analysis is intended as a transparent structural pattern check rather than as a causal estimate derived from observed brokerage or exchange data. Standardized coefficients are subsequently translated into practically interpretable indicators, including portfolio turnover, the Herfindahl–Hirschman index and drawdown exposure.
Within the simulated environment, higher perceived control is associated with more frequent trading, greater portfolio concentration, higher risk-taking and weaker net performance. Overconfidence partially mediates these relationships, reducing the magnitude of the direct coefficients once calibration error is introduced. Realized volatility strengthens both the direct associations between perceived control and investor behavior and the indirect pathways operating through overconfidence. The findings indicate that behavioral vulnerabilities linked to perceived control may become more consequential when market variance increases.
This study integrates perceived control, overconfidence and market volatility within a unified moderated-mediation framework and connects standardized behavioral relationships to platform-relevant risk metrics. It also derives design implications, including volatility-adaptive trading throttles, leverage gating, calibrate-before-trade prompts and cost-salience panels. By clearly separating simulation-based theoretical validation from empirical causal inference, this study provides a reproducible foundation and an external-validation roadmap using brokerage panels, exchange records and event-based quasi-experimental designs.
