Do Investors Emotions Contribute to Equity Market Anomalies? Addressing the Empirical Gap using Machine Learning Models

The study explores the relationship between investor emotions and stock market anomalies in the Indian financial landscape, examining both regular and irregular market occurrences. By utilizing mixed methods, the study uses an LSTM model for anomaly detection and sentiment analysis from Twitter data. It also includes regression analysis to measure the influence of public sentiment on stock prices. The results suggest that investor emotions play a significant role in market anomalies during extraordinary events like the 2008 financial crisis and the demonetization initiative in 2016, but have a lesser effect during expected, repetitive events. The study stands out for its empirical investigation of emotional finance theory to uncover the reasons behind stock market anomalies in India, an area that has not been thoroughly examined before. This research has two main implications: it questions the Efficient Market Hypothesis and provides suggestions for regulatory measures and investment strategies that consider emotional influences on market behavior.

本研究聚焦印度金融市场环境下投资者情绪与股市异象之间的关联，同时考察常规与非常规市场运行场景。研究采用混合研究方法，运用长短期记忆网络（LSTM）模型对Twitter数据开展异常检测与情感分析，并通过回归分析量化公众情绪对股票价格的影响程度。研究结果显示，在2008年金融危机、2016年印度废钞令等极端事件期间，投资者情绪对市场异象具有显著影响；而在可预期的重复性市场事件中，该影响则相对微弱。本研究的独特之处在于，针对此前尚未得到充分探究的印度股市异象成因问题，基于情感金融理论展开实证研究。本研究主要具备两方面学术意义：一是对有效市场假说（Efficient Market Hypothesis）提出质疑，二是为考虑情绪对市场行为影响的监管措施制定与投资策略优化提供参考建议。