The term structure of uncovered interest parity in emerging markets

ABSTRACT

This note examines uncovered interest parity (UIP) at both short and longer horizons in emerging markets. In part due to currency market intervention, the duration of risk premium shocks, and more easily verified medium term trends, the UIP term structure in these economies may differ from that in advanced economies. Analysing ten major emerging markets between 2000 and 2020, the evidence indicates that the validity of UIP increases with the investment horizon, corroborating a similar finding for advanced economies.

KEYWORDS:

• Term structure
• uncovered interest parity
• risk-premium shocks
• investment horizon

JEL CLASSIFICATION:

• E40
• F30

Acknowldgement

I would like to thank the editor, David Peel, and an anonymous referee for the suggestions that have improved the content and contributions of the paper, Che-Wing Pang and Philip Hubbard at Consensus Forecasts for kindly providing details on data compilation and survey design, Sebnem Kalemli-Ozcan and Marcos Chamon for helpful discussions and Awais Khuhro and Hailey Ordal for excellent research assistance. Any views expressed in the paper do not necessarily reflect the views of the IMF, its Executive Board or IMF Management.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ Bansal and Dahlquist (2000) and Frankel and Poonawala (2010) document a positive relation between interest rate differentials and the expected exchange rates in emerging markets, although the relation is disparate from the UIP-implied unity. Subsequent work has found significant UIP deviations in emerging countries, in some cases even larger than in advanced economies; see Francis, Hasan, and Hunter (2002), Poghosyan et al. (2008), Ferreira and León-Ledesma (2007), Loring and Lucey (2013), Coulibaly and Kempf (2019), Kalemli-Ozcan and Varela (2020).

² Primitive considerations for time-varying risk premia include the carry trade (Brunnermeier, Nagel, and Pedersen 2009), habits (Verdelhan 2010) and rare disasters (Farhi and Gabaix 2016).

³ Singleton (2014) argues that trading activity results from an adaptive learning process in which speculators learn about economic fundamentals by observing market prices.

⁴ Forward rates from Datastream are for the same period as the data from Consensus Forecasts, with two major exceptions: China (whose available data for both 3- and 12-month are 2006:8–2020:1) and Russia (whose available data for both horizons is 2004:3–2020:1).

⁵ Notable exceptions are India and China. During much of the last two decades, the Chinese currency in particular has been relatively stable vis-à-vis the U.S. dollar (Ilzetzki, Reinhart, and Rogoff 2019), often attributed to either management of the exchange rate or precautionary FX reserves accumulation. The persistent stability has potentially given rise to expectations of future appreciation, despite consistently positive interest rate differentials from the U.S. Similar considerations are also relevant in India, albeit to a lesser extent.