Hybrid GARCH-LSTM Forecasting for Foreign Exchange Risk

This study proposes a hybrid forecasting model that integrates the Generalized Autoregressive Conditional Heteroskedasticity (GARCH) model with a Long Short-Term Memory (LSTM) neural network to estimate Value at Risk (VaR) in the Rwandan foreign exchange market. The model is designed to capture both...

Full description

Saved in:
Bibliographic Details
Main Authors: Elysee Nsengiyumva, Joseph K. Mung’atu, Charles Ruranga
Format: Article
Language:English
Published: MDPI AG 2025-06-01
Series:FinTech
Subjects:
Online Access:https://www.mdpi.com/2674-1032/4/2/22
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849472210242109440
author Elysee Nsengiyumva
Joseph K. Mung’atu
Charles Ruranga
author_facet Elysee Nsengiyumva
Joseph K. Mung’atu
Charles Ruranga
author_sort Elysee Nsengiyumva
collection DOAJ
description This study proposes a hybrid forecasting model that integrates the Generalized Autoregressive Conditional Heteroskedasticity (GARCH) model with a Long Short-Term Memory (LSTM) neural network to estimate Value at Risk (VaR) in the Rwandan foreign exchange market. The model is designed to capture both volatility clustering and temporal dependencies in daily exchange rate returns. Using daily data on USD, EUR, and GBP from 2012 to 2024, we evaluate the model’s performance relative to standalone GARCH(1,1) and LSTM models. Empirical results show that the hybrid model improves VaR estimation accuracy by up to 10%, especially during periods of elevated market volatility. These improvements are validated through MSE, MAE, and backtesting statistics. The enhanced accuracy is particularly relevant in emerging markets, where exchange rate dynamics are highly nonlinear and sensitive to external shocks. The proposed approach offers practical insights for financial institutions and regulators seeking to improve market risk assessment in emerging economies.
format Article
id doaj-art-96b652ed5d914cf7a0afd03af82e2f06
institution Kabale University
issn 2674-1032
language English
publishDate 2025-06-01
publisher MDPI AG
record_format Article
series FinTech
spelling doaj-art-96b652ed5d914cf7a0afd03af82e2f062025-08-20T03:24:36ZengMDPI AGFinTech2674-10322025-06-01422210.3390/fintech4020022Hybrid GARCH-LSTM Forecasting for Foreign Exchange RiskElysee Nsengiyumva0Joseph K. Mung’atu1Charles Ruranga2African Centre of Excellence in Data Science, University of Rwanda, Kigali P.O. Box 428, RwandaDepartment of Statistics and Actuarial Science, Jomo Kenyatta University of Agriculture and Technology, Nairobi P.O. Box 62000-00200, KenyaAfrican Centre of Excellence in Data Science, University of Rwanda, Kigali P.O. Box 428, RwandaThis study proposes a hybrid forecasting model that integrates the Generalized Autoregressive Conditional Heteroskedasticity (GARCH) model with a Long Short-Term Memory (LSTM) neural network to estimate Value at Risk (VaR) in the Rwandan foreign exchange market. The model is designed to capture both volatility clustering and temporal dependencies in daily exchange rate returns. Using daily data on USD, EUR, and GBP from 2012 to 2024, we evaluate the model’s performance relative to standalone GARCH(1,1) and LSTM models. Empirical results show that the hybrid model improves VaR estimation accuracy by up to 10%, especially during periods of elevated market volatility. These improvements are validated through MSE, MAE, and backtesting statistics. The enhanced accuracy is particularly relevant in emerging markets, where exchange rate dynamics are highly nonlinear and sensitive to external shocks. The proposed approach offers practical insights for financial institutions and regulators seeking to improve market risk assessment in emerging economies.https://www.mdpi.com/2674-1032/4/2/22hybrid GARCH-LSTM modeldeep learning in financeforeign exchange riskvalue at risk (VaR)emerging market forecasting
spellingShingle Elysee Nsengiyumva
Joseph K. Mung’atu
Charles Ruranga
Hybrid GARCH-LSTM Forecasting for Foreign Exchange Risk
FinTech
hybrid GARCH-LSTM model
deep learning in finance
foreign exchange risk
value at risk (VaR)
emerging market forecasting
title Hybrid GARCH-LSTM Forecasting for Foreign Exchange Risk
title_full Hybrid GARCH-LSTM Forecasting for Foreign Exchange Risk
title_fullStr Hybrid GARCH-LSTM Forecasting for Foreign Exchange Risk
title_full_unstemmed Hybrid GARCH-LSTM Forecasting for Foreign Exchange Risk
title_short Hybrid GARCH-LSTM Forecasting for Foreign Exchange Risk
title_sort hybrid garch lstm forecasting for foreign exchange risk
topic hybrid GARCH-LSTM model
deep learning in finance
foreign exchange risk
value at risk (VaR)
emerging market forecasting
url https://www.mdpi.com/2674-1032/4/2/22
work_keys_str_mv AT elyseensengiyumva hybridgarchlstmforecastingforforeignexchangerisk
AT josephkmungatu hybridgarchlstmforecastingforforeignexchangerisk
AT charlesruranga hybridgarchlstmforecastingforforeignexchangerisk