Behavioral biases in the cryptocurrency market: a study on the impact of investor sentiment on price anomalies

Abstract

By combining behavioral finance theory and big data analysis technology, this study explores the mechanism of the impact of investor sentiment on cryptocurrency market price anomalies. Based on the fusion database of traditional exchange historical market data and social media sentiment data, the research team constructed multidimensional sentiment indicators to quantify the emotional fluctuations of market participants. The research design adopts the strict data cleaning process, feature engineering processing, and the hybrid modeling method combining the traditional statistical model and the machine learning algorithm. The empirical results show that extreme optimism or pessimism is significantly associated with abnormal price events, and the predictive ability of the composite sentiment index is better than that of the single volatility index. This research reveals the transmission pathways of cognitive biases such as overconfidence and the anchoring effect in the cryptocurrency market, confirming the significant influence of irrational psychological factors on the price formation of digital assets. The findings not only deepen our understanding of the interaction mechanism between investor psychology and market behavior, but also provide an innovative analytical framework for risk management and quantitative investment strategies in the cryptocurrency field.