Investigating exchange rate volatility in some West African Countries using traditional time series and machine learning models
Year: 2025
Type: article
Abstract: This study used both conventional time series models and machine learning techniques to examine the volatility of exchange rates in Nigeria, Ghana, Niger, Gambia, and Sierra Leone. Due to their economic importance, regional representation, and variety of exchange rate regimes, the chosen countries were the subject of the 1999–2022 study. Data were sourced from the World Bank and the International Monetary Fund (IMF). The traditional time series models employed included ARIMAX and GARCH, while machine learning techniques included Long Short-Term Memory (LSTM) and hybrid models combining traditional econometric models with advanced machine learning methods. Results from the analyses revealed that the currencies of the countries investigated were unstable for the period covered. For Nigeria, Niger, Ghana, Gambia, and Sierra Leone, the conventional time series models calculated volatility and showed statistically significant returns at the 0.05 level. This implied that current volatility was impacted by previous performance. Stability in volatility modeling was indicated by the GARCH and ARCH parameter sums being less than 1 (ranging from 0.346 to 0.716). One of the study's important conclusions was that, with the lowest Mean Squared Error (MSE), the ARIMAX-LSTM model performed better than the GARCH-type and GARCH-LSTM models. This study provided valuable insights for policymakers, investors, and financial institutions, aiding in more informed decision-making and contributing to economic stability in the region.
Language: English
Subfield: Finance
Domain: Social Sciences
Open Access status: hybrid
OpenAlex ID: W4410695510