From research: Forecasting nominal exchange rates using DMA framework

Does the dynamic model averaging (DMA) approach enhance the accuracy of the exchange rate forecasting, particularly in the medium and long run?  In a paper published in the journal Heliyon Martin Časta, on behalf of VŠE, introduced dynamic model averaging framework for exchange rate forecasting. Using data of nine currency pairs (AUD/USD, CAD/USD, CHF/USD, EUR/USD, GBP/USD, NOK/USD, NZD/USD, SEK/USD, and JPY/USD) from the past two decades, the study finds strong medium and long-term predictability for nine major currency pairs, outperforming the random walk benchmark. Short term forecasts, however, remain less reliable.

Long-term forecasts (two to three years) are particularly strong, even with simplified models. While adding more variables improves accuracy, the effects vary. The study also supports the idea of mean reversion in exchange rates, while also suggesting some fundamental value convergence over time. Short term predictions remain challenging.

Read more about this novelty approach, factors contributing to those results and opportunities for further research at:

https://www.sciencedirect.com/science/article/pii/S2405844024151432?via%3Dihub