Five Facts about the UIP Premium
We construct a forward-looking measure of the currency risk premium—the UIP premium—from survey-based expectations of exchange rates for 22 emerging markets and 12 advanced economies, and use it to characterize how currency risk is priced across the two groups. Our central result is a compositional asymmetry: in advanced economies the UIP premium is primarily an expectations phenomenon that co-moves with global risk, whereas in emerging markets it is primarily an interest-rate-differential phenomenon that co-moves with local risk tied to domestic policy uncertainty. We offer a unifying empirical model of currency risk premia, interest rates, and exchange rates, both in the cross section and in time series of emerging market and advanced economy currencies—summarized in five facts: (1) The average UIP premium for emerging markets is consistently positive, higher, and more volatile than the one for advanced economies; (2) A significant portion of cross-sectional and time-series variation in the UIP premium is driven by local risk factors in emerging markets; (3) The interest rate differential component of the UIP premium is more volatile and strongly correlated with local risk factors in emerging markets compared to advanced economies; (4) Local and global risk factors influence exchange rate expectations, which in turn predict interest rate differentials; (5) The strongest predictor of the local risk factor is country-time varying policy uncertainty, influencing expectations of currency depreciations and the UIP premium. Together, these facts provide empirical moments that discipline models of segmented currency markets with endogenous, time-varying currency risk premia.
-
-
Copy CitationṢebnem Kalemli-Özcan and Liliana Varela, "Five Facts about the UIP Premium," NBER Working Paper 28923 (2021), https://doi.org/10.3386/w28923.Download Citation
-
Published Versions
Forthcoming: Five Facts about the UIP Premium, Ṣebnem Kalemli-Özcan, Liliana Varela. in NBER International Seminar on Macroeconomics 2026, Frankel and Rey. 2026