ROLE OF AI MODELS IN FORECASTING INR–USD VOLATILITY

Abstract

Forecasting exchange rate volatility, especially for emerging market currencies like the INR–USD pair, has become increasingly important for policymakers, investors, and multinational firms. Traditional econometric models such as GARCH and stochastic volatility (SV) capture persistence and clustering but often fail to incorporate nonlinear patterns produced by global shocks, capital flows, algorithmic trading, and structural breaks. This study evaluates the effectiveness of Artificial Intelligence (AI) and Machine Learning (ML) models—including Long Short-Term Memory (LSTM), Bidirectional LSTM, Transformer-based time-series models, Random Forest, and Support Vector Regression—in forecasting INR–USD volatility between 2010 and 2024.