RDP 2025-09: Forecasts of Period-average Exchange Rates: Insights from Real-time Daily Data 6. Conclusion

We construct a novel real-time dataset of daily bilateral and effective exchange rates, both nominal and real, for a broad set of countries. These data enable the first real-time evaluation of forecasts for period-average exchange rates and allow us to assess the effects of temporal aggregation. Our analysis yields three main findings.

First, the month-average no-change forecast benchmark substantially underperforms the end-of-month no-change benchmark for EERs. This questions the empirical results of the surveyed papers examining period-average exchange rates.

Second, incorporating information from daily or end-of-month exchange rates substantially improves the accuracy of real-time forecasts. This points to the potential for improving current forecasting practices, leading to better decision-making. It also highlights the need for official data providers to begin publishing end-of-month and daily measures of EERs, rather than the current practice of only reporting period-average EERs.

Third, we find that period-average EERs and bilateral real exchange rates are forecastable in real time for many countries. These findings motivate further exploration of temporally disaggregated methods and a reassessment of the predictability of EERs.

More broadly, the daily real-time dataset introduced in this paper creates the possibility for a wide range of applications, including the identification of high-frequency shocks, the analysis of dynamic responses, and the evaluation of economic theory and policy. Both the survey and daily data offer a foundation for new empirical work in macroeconomics.