A Critical Review of Predictive Techniques for Stock Price Volatility in the Indian Market: Limitations and the Need for Sentiment Analysis Integration

This theme offers a critical examination of predictive techniques for stock price volatility, particularly focusing on the Indian stock market, which has experienced significant growth and an increase in investment activities. The fluctuations in the stock price have consistently posed a challenge in forecasting the stock's volatility. Traditional econometric models like ARCH and GARCH have gained popularity in forecasting volatility; however, they face limitations in dynamic market conditions and external shocks. With the recent rise of machine learning techniques such as Artificial Neural Networks (ANNs), Support Vector Machines (SVMs), and Random Forests (RF), predictive accuracy has significantly improved by effectively capturing complex non-linear relationships. Furthermore, sentiment analysis has evolved through social media and financial news, serving as a method to gauge market sentiment and its influence on volatility. This review highlights the significance of integrating sentiment analysis with traditional and machine learning models to enhance the precision of predictions. The combination of quantitative financial data and sentiment obtained from qualitative sources will allow predictors to take into account the various psychological and emotional factors that affect stock prices. The paper further explores the limitations and challenges of current methodologies and suggests a direction for future research that should integrate real-time sentiment analysis, hybrid models, and alternative data sources into volatility forecasting.