The behaviour of a small foreign exchange market with a long-term peg–Barbados

Abstract

This article is a first analysis of daily transactions in the foreign exchange market of Barbados, a small open economy that has had an unchanged peg to the US dollar for over 30 years. As a result of the credibility of the peg, we expect that capital flows will respond to differentials between US and comparable Barbadian interest rates, and that this will result in uncovered interest parity, when allowance is made for market frictions and large discrete events. The tests appear to confirm this.

Acknowledgements

The authors acknowledge with thanks the comments of Gene Leon, Galina Menchikova, Romain Veyrune, Warrick Ward and a referee from this journal, though we are solely responsible for the views expressed herein. The views expressed in this article are those of the authors and do not necessarily represent those of the IMF or IMF policy, or of the Central Bank of Barbados.

Notes

¹ In this journal, studies include Goh et al. (2006) and Camarero et al . (2002).

² Recent comprehensive surveys of this literature include Koedijk et al . (2006) and Sarno (2005). Koedijk, Lothian and Dijk attribute improvements in uncovered interest parity (UIP) tests to the use of longer time horizons, inclusion of emerging market data, improved measures of exchange rate expectations, testing with nonlinear relationships and allowing for heterogeneous beliefs. Sarno reports studies that allow for risk aversion (participants demand a risk premium); monetary policy reactions to expected exchange rate changes; risk premiums that vary with maturity of instrument; the existence of rational bubbles; markets that learn from experience; markets characterized by capital flight; information inefficiencies; the use of density forecasts instead of point forecasts; and the existence of a band of inaction where foreign bias does not invite speculative capital movements.

³ Aruba, The Bahamas, the Cayman Islands, Barbados, Belize, the Eastern Caribbean Currency Union (ECCU) and the Netherlands Antilles.

⁴ Bosnia and Herzegovina, Brunei Darussalam, Bulgaria, Djibouti, Estonia and Lithuania.

⁵ See IMF (2005).

⁶ The limit was reduced from 110% in October 2004.

⁷ Banks report their position every Wednesday, and they are required to clear any excess liability within 2 business days.

⁸ See Sarno and Taylor (2002), Chapter 2.

⁹ In contrast, in relatively closed economies, where the ratio of foreign transactions to the money supply is low, foreign exchange flows are insufficient to fund the excess demand (positive or negative) for money, and the central bank may therefore determine the domestic interest rate.

¹⁰ Sarno and Taylor (2001).

¹¹ These payments produced the outliers that appear in Figs 2–4.

¹² The product of the coefficient of 0.22 on Tseason and the mean net sale of $ 38 000.

¹³ The coefficient for Spayment is −28.19.

¹⁴ Based on the elasticity estimate of −4.23 * e⁻⁴.