PurposeThe purpose of this study is to evaluate the use of a parallel market mechanism in forecasting the exchange rate of Myanmar Kyat (MMK) after the 2021 Myanmar coup. This paper seeks to answer how an official fixed exchange rate is influenced by a parallel market rate during the military coup. This study demonstrates both the benefit of understanding the impact of a market mechanism and the unique value of leveraging deep learning (DL).Design/methodology/approachThis study applies DL for the first time to forecast the daily official exchange rates and parallel exchange rates, covering the period January 5, 2021–August 22, 2022. This study shows the insights available from forecasting the market exchange rate of MMK in parallel markets that are sensitive to political changes for Myanmar.FindingsAfter the February 2021 military coup in Myanmar, the subsequent use of an untrusted exchange rate system in Myanmar resulted in a sharp gap between the official and the parallel exchange rates. The study shows that the actual forecasting of exchange rates is primarily a reflection of an active market mechanism rather than official exchange rates set by the monetary authorities.Originality/valueNo previous study has forecast the exchange rate of Myanmar since the latest Myanmar coup in 2021. This study contributes crucial predictions, provides a robust technique “deep learning”.