Behavioral overtrading, bot filtering and retail performance in centralized cryptocurrency markets: transaction-level evidence from anonymized exchange data
Wen-Bao Lin, Jia-Ying LyuPurpose
This study investigates how investor overconfidence shapes trading behavior and realized performance in cryptocurrency markets. Specifically, it examines whether overconfident trading leads to systematic return penalties, and whether these effects intensify under high-volatility conditions and for less experienced investors.
Design/methodology/approach
Using 77 million transaction records from 405,000 investors across four major cryptocurrency exchanges over 2014–2024, this paper proxies overconfidence by trading frequency and portfolio concentration (HHI). To strengthen causal interpretation, it exploits quasi-exogenous variation in trading frictions – exchange outages and fee-tier resets – as instruments for trading intensity, supplemented by extensive robustness checks and alternative concentration measures distinguishing benchmark (BTC/ETH) holdings from altcoin-dominant portfolios.
Findings
Both overconfidence proxies are robustly associated with significantly lower realized and risk-adjusted returns, indicating economically meaningful welfare losses for retail investors. The penalty is strongly state-dependent, becoming substantially larger during high-volatility regimes and among novice traders. Instrumental-variable results confirm that intensified trading causally contributes to performance deterioration rather than merely reflecting unobserved investor heterogeneity.
Originality/value
This study contributes transaction-level evidence to behavioral finance by quantifying volatility-sensitive overconfidence costs in an emerging digital-asset setting. It also provides managerial and regulatory implications by highlighting trading frictions as a market-design margin: exchange fee policies, operational resilience and consumer-facing risk disclosures can serve as practical levers to mitigate avoidable retail losses and improve market quality.