Neural networks and ARMA-GARCH models for foreign exchange risk measurement and assessment

Market turnover levels and liquidity changes across various territories significantly influence currency prices, leading to continuous fluctuations. Consequently, traders and investors constantly seek strategies to mitigate exchange rate risks. This study aimed to measure and assess foreign exchange risk utilizing Neural Networks and ARMA-GARCH models. Data on five leading currencies, covering the period from 6 January 2016 to 28 June 2024 were sourced from the National Bank of Rwanda. Specifically, the study employed the long-short-term memory (LSTM) model, a type of recurrent neural network, to evaluate the riskiness of asset currencies. The estimated volatilities were compared with those derived from traditional ARCH-GARCH models. Notably, the LSTM model yielded lower root mean square error values compared to the ARMA-GARCH models, demonstrating superior accuracy in forecasting currency volatilities. The findings indicate that EGP and KES are riskier than USD, EUR, and GBP. This research explores advanced methods for measuring and assessing foreign exchange risk using Neural Networks, specifically Long Short-Term Memory (LSTM), and ARMA-GARCH models. By focusing on five significant currencies traded in the Rwandan foreign exchange market, the study demonstrates the superiority of the LSTM model over traditional statistical models, offering a more accurate and reliable approach to predicting currency volatilities. These findings provide valuable insights for financial institutions, investors, and policymakers, equipping them with robust tools for risk management in currency trading and enhancing decision-making capabilities. The model's success in accurately forecasting exchange rate fluctuations also highlights the potential for integrating machine learning into finance, contributing to improved stability and foresight in volatile markets.