Exchange Rate Prediction Under Data Volatility: A Comparison Of Deep Learning Techniques
Open Access
- 6 January 2026
- journal article
- Published by Izmir Akademi Dernegi in Journal of Metaverse
- No. 6,p. 100-115
Abstract
This Accurate prediction of exchange rates is of critical importance for economic policy related decision making. This study employs a deep learning based framework to model exchange rate dynamics by explicitly accounting for the inherently nonlinear, and regime sensitive nature of financial time series. Unlike traditional artificial neural network approaches that overlook temporal dependencies, recurrent neural network architectures are capable of directly modeling long term time dependencies. Accordingly, this study applies Long Short Term Memory (LSTM), Bidirectional LSTM (Bi-LSTM), and Gated Recurrent Unit (GRU) models to predict the USD/TRY and EUR/TRY exchange rate series. The empirical analysis is conducted using data from the Turkish economy across three distinct subperiods, representing a relatively stable regime and two high-volatility regimes. Hyperparameters are independently optimized for each model and regime with grid search. Predictive performance is evaluated using Mean Absolute Error, Root Mean Squared Error, and Mean Absolute Percentage Error, with Diebold Mariano tests confirming statistical significance and formal benchmark comparisons against ARIMA and random walk models. Results indicate that no single architecture dominates across all conditions: GRU consistently outperforms competitors during the Crisis period for both exchange rates, while LSTM and BiLSTM perform best during COVID19 for USD/TRY, and GRU maintains its advantage for EUR/TRY. Diebold Mariano tests confirm that 89% of pairwise performance differences are statistically significant. Robustness analysis using lag-3 specifications supports that regime-dependent patterns are not artifacts of input structure. By incorporating volatility regimes, statistical testing, and econometric benchmarks, this study demonstrates the practical importance of regime-aware model selection for exchange rate forecasting in emerging market economies.Keywords
All Related Versions
Funding Information
- This research did not receive any outside funding or support. The authors report no involvement in the research by the sponsor that could have influenced the outcome of this work.
This publication has 72 references indexed in Scilit:
- LSTM–GARCH Hybrid Model for the Prediction of Volatility in Cryptocurrency PortfoliosComputational Economics, 2023
- Dynamic forecasting for nonstationary high‐frequency financial data with jumps based on series decomposition and reconstructionJournal of Forecasting, 2022
- Deep learning-based exchange rate prediction during the COVID-19 pandemicAnnals of Operations Research, 2021
- Foreign Currency Exchange Rate Prediction Using Bidirectional Long Short Term Memory2021
- Multi-Factor RFG-LSTM Algorithm for Stock Sequence PredictingComputational Economics, 2020
- Forex exchange rate forecasting using deep recurrent neural networksDigital Finance, 2020
- An ensemble of LSTM neural networks for high‐frequency stock market classificationJournal of Forecasting, 2019
- Improving GARCH volatility forecasts with regime-switching GARCHEmpirical Economics, 2002
- Forecasting Austrian IPOs: An application of linear and neural network error‐correction modelsJournal of Forecasting, 1996
- Forecasting exchange rates using feedforward and recurrent neural networksJournal of Applied Econometrics, 1995