Uncovered Interest Rate Parity – A Central Concept in International Finance
- By Sam van de Schootbrugge
- 5 min read
Summary
- A new NBER working paper revisits the uncovered interest rate parity condition using survey data on exchange rate expectations.
- There are five findings, most notably: the Fama (1984) puzzle disappears in advanced economies.
- The UIP condition still does not hold among EM currencies, as global investors’ risk sentiment about a country’s idiosyncratic policy risk is priced into the interest rate differential.
Introduction
The uncovered interest rate parity (UIP) condition is a key concept in international macroeconomics and finance. Famously, it appears not to hold over one- to two-year horizons, creating positive excess returns from investing in international currencies. A new NBER working paper revisits the UIP condition using survey data and documents five novel facts:
- The UIP premium co-moves with global risk aversion (VIX) for all currencies.
- In emerging economies, the interest rate differential co-moves with the VIX, while expected exchange rate changes co-move with the VIX in advanced economies.
- The degree of policy uncertainty can explain the negative co-movement of the UIP premium with the VIX, interest rate differentials and country-specific capital inflows in EMs.
- There are no overshooting and predictability reversal puzzles – for any currency – when using survey data instead of realised exchange rates.
- The classic Fama puzzle disappears in advanced economies when using survey data, as opposed to ex-post exchange rate data, but remains for EMs.
The UIP Premium
The UIP premium comprises two components: the interest rate differential and the exchange rate adjustments (Chart 1). The former is the domestic interest rate minus the foreign short-term interest rate (the US in this paper). The latter deducts the expected exchange rate from the exchange rate in units of local currency per USD. When the two components offset each other, the UIP condition holds.
The Fama (UIP) Puzzle
Under the UIP, the home currency is expected to depreciate if the home interest rate is higher than the foreign interest rate, offsetting any interest rate differential. In practice, however, Fama (1984) documented an appreciation, implying positive excess returns on average (the equation in Chart 1 would be positive).
Other studies have since shown that these positive excess returns last between one and two years and that currencies then depreciate too much (overshooting), especially after crisis periods. However, all these studies use realised exchange rates to capture the theoretical counterpart of expectations (ste + h in Chart 1). They do so under the full information and rational expectations (FIRE) hypothesis, which means the expected exchange rate can be approximated with the ex-post exchange rate.
Constructing the UIP Premium Using Survey Data
Instead of using realised exchange rates, the authors use expectations of exchange rates from survey data. Specifically, they use the Consensus Forecast, which conducts a monthly survey about expectations on future exchange rates at one-, three-, 12- and 24-month horizons of major participants in the foreign exchange market. The survey covers advanced and EM economies.
The authors then use the IMF and Bloomberg to obtain exchange rate and interest rate data and construct the UIP at a 12-month horizon. In total, they gather data on 12 advanced and 22 EM economies over the period 1996 to 2018.
Methodology
After constructing the UIP premium, the authors examine, via regressions, how UIP dynamics are affected by changes in global risk, policy uncertainty and capital flows. For global risk, they use the VIX from the Federal Reserve Economic Data (FRED), and they take capital flow data from the IMF’s International Financial Statistics (IFS).
To proxy policy uncertainty for each country, they use the Economic Policy Uncertainty (EPU) Index. It gives an indication of investment risk premium, i.e., the excess compensation required due to country risk. They also use the International Country Risk Guide (ICRG), which reports several breakdowns for policy risk that capture the risk to global investors’ returns due to unexpected changes to central banks’ and governments’ actions.
For Fact 4, the authors examine the impact of an interest rate differential change on the UIP premium and expected changes in the exchange rate. This is to see whether the initial appreciation is followed by a larger-than-expected depreciation in the currency (overshooting). Finally, the authors test whether the UIP condition holds when using survey data on exchange rates.
The Facts
Facts 1-3
The authors regress changes in capital flows, VIX and policy uncertainty on the UIP premium. They find that capital inflows are associated with decreases in the UIP premium in EMs. Capital inflows lower the local interest rate and lead to an expected currency appreciation. Yet, the expected appreciation is insufficient to compensate for the lower interest rate differential and the UIP premium drops. For advanced economies, capital flows have an impact on the UIP premium.
Meanwhile, a strong correlation exists between the VIX and the UIP premium for both advanced and emerging economies (Chart 2). Quantitatively, if the VIX increases by 150% (as it did during the GFC), the UIP premium in EMs would increase by 9ppt. This strong correlation in EMs is driven by the co-movement between the VIX and the interest rate differential. Meanwhile, in advanced economies, the VIX leads to larger-than-expected changes in the exchange rate, and the interest rate differential channel plays no role.
Given the UIP premium in EM countries is driven by interest rate differentials, the authors argue it is related to an investment risk premium. That is, investors ask for an excess compensation that goes above their expectations of emerging markets’ currency depreciations because of the country risk inherent in EM investing.
Indeed, the authors find that policy uncertainty (EPU) linked to the credibility of emerging markets’ policies is key to explaining fluctuations in the UIP premium in EMs. An increase in the EPU Index is associated with a positive risk premium, channelled through a higher interest rate. And so, the authors argue that ‘global investors’ risk sentiment about a country’s idiosyncratic policy risk is priced in the interest rate differentials and, by this means, entails a UIP premium’.
Fact 4
The overshooting phenomenon refers to the U-shaped dynamics of exchange rates upon a domestic interest rate shock. Intuitively, a positive domestic interest rate shock increases the interest rate differential and leads to an initial appreciation. This is followed by a delayed depreciation, producing a U-shaped dynamic for the realised interest rate. Often, the depreciation is so large it creates negative excess returns.
The authors test this prediction using survey data. They instead find an inverted U-shape dynamic for EMs and no dynamics for advanced economies. For EMs, investors always expect a depreciation following an interest rate shock, but the expected depreciation is less than the initial interest rate shock and so leads to positive excess returns. These positive excess returns are again linked to policy uncertainty, as EM investors demand a higher premium because of higher policy uncertainty.
Fact 5
Lastly, the authors test the UIP condition to see if the Fama puzzle still exists when using survey data to proxy expectations. A simple eyeballing exercise shows the UIP broadly holds in advanced economies (UIP premium is on average zero) but not in EMs (Chart 3). Indeed, an OLS regression confirms these observations and indicates that the depreciation in EM currencies is insufficient to offset the interest rate differential, thereby creating positive excess returns.
Bottom Line
The co-movement between the UIP premium and VIX shows the importance of global uncertainty as a determinant of the UIP premium. Interestingly, the authors show that any interest rate changes in EMs lead to lower-than-required expected depreciations, creating persistent expected excess returns. This is not the case in advanced countries. And so, global investors expect and earn positive excess returns from investing in EM currencies which, according to the authors, arises from investors charging an ‘excess’ risk premium to compensate for policy uncertainty.
Citation
Kalemli-Özcan, S., Varela, L., (2021), Five Facts about the UIP Premium, NBER, (Working Paper 28923), https://www.nber.org/papers/w28923