STOCK VOLATILITY ANALYSIS USING A STATISTICAL MODEL ON NIFTY50

Mikkilineni Nithin Sai , Bhargavi Rayal Pasupuleti , Gajula Kartikeya Chandra ganesh , Are Ajay , Dr. Jalaluddin Khan

ime series analysis, GARCH model, Volatility modeling, Autoregressive conditional heteroscedasticity, Financial forecasting, Risk management, Asset pricing, Stock returns, Conditional variance, ARCH effect

The project's goal is to evaluate different algorithms for detecting stock volatility and assess their effectiveness using machine learning methods. Understanding the intrinsic value of a stock in a particular industry is essential due to the volatility in the stock market. The machine learning algorithm uses previously available traded data to determine its efficiency in predicting stock volatility. Time series data with heteroscedasticity can be modelled using the statistical model GARCH, which is frequently used in finance and econometrics. The ultimate goal is to use these algorithms to aid in stock market decision-making. Because stock price time series are the nature of stationary and nonlinear systems involves the ability to anticipate future price trends is very difficult. Statistical algorithm strategies like the GARCH method, which is used to rely on lengthier data collection and overall replace patterns of stock units to learn the long-term dependence on stock unit prices. In order to demonstrate the efficacy and benefits of deep learning methods, this thesis conducts empirical research using 5000 units information extracted from the NIFTY50 index and introduces models for evaluation and comparison of the ARIMA, GARCH, and other research approaches.

Volume 8 Issue 11, November-2023

Pages : 282 - 288