LEVERAGING SARIMA MODELS FOR PREDICTIVE ANALYSIS OF TOBACCO SECTOR STOCK PRICES IN INDIA

Background: The Indian tobacco industry, with its regulatory complexity and market volatility, plays a major role in the national economy. Accurate forecasting of tobacco stock prices is critical for investors, policymakers, and analysts to make informed decisions in such a dynamic sector. Methodology: This study employs the Seasonal Autoregressive Integrated Moving Average (SARIMA) model to forecast the monthly stock prices of VST Industries Ltd., a leading tobacco company in India. Historical adjusted closing prices from January 2010 to December 2022 were collected from multiple financial sources. The dataset was tested for stationarity using the Augmented Dickey–Fuller test, followed by seasonal and non-seasonal differencing. Model parameters were identified through ACF/PACF plots, and the optimal SARIMA model was selected using AIC, BIC, and diagnostic checks. Results: Among the tested configurations, SARIMA(1,0,0)(0,1,1)12_{12}12 provided the best fit, effectively capturing both seasonal and trend components of stock prices. The model was validated using a training–testing split, with performance measured by RMSE, MAE, and MAPE. Forecasting accuracy was satisfactory (Test RMSE = 0.30; MAPE = 13.26%), confirming the robustness of the selected model. Conclusion: The study demonstrates that SARIMA is an effective and reliable tool for forecasting stock price movements in the tobacco sector, offering practical implications for investment planning and risk management. This research fills a gap in sector-specific forecasting literature in India and suggests integrating exogenous economic factors or machine learning models for improved predictive performance in future studies.