Interest rate changes affect domestic markets as well as foreign currency exchange markets. An important area in international financial research has developed around the interest rate parity conditions between countries. This condition with its two types - covered and uncovered, offers a simple mechanism of understanding the impact that interest rates have on changes in currencies in two countries. The uncovered interest rate parity condition (UIP) states that the difference between expected and actual spot exchange rate on a pair of currency at a future point in time should be equal to the difference in the interest rates between two countries. The UIP suggests that countries with high interest rates should see their currency depreciate, whereas countries with low interest rates are expected to have an appreciating currency.
Most of the research developed shows the failure of the UIP, i.e. countries with high interest rates have appreciating currency. Lothian and Wu (2003) propose that the reason for the violation of the UIP is the particular period of 1970s-1980s when most of the studies were done and the use of USD as a numeraire currency. Their findings show that when the time period is long and one of the currencies is not USD the UIP holds. Regressions with large samples confirm the results that UIP holds in longer periods. Similarly, Bailie and Bollerslev (2000) demonstrate that regressions with small samples point to rejecting the UIP.
The tests of uncovered interest rate parity are usually done with currency pairs, one of which is the US dollar (USD). Mehl and Cappiello (2007) have found evidence that the UIP holds in long and medium horizons for USD versus other major currencies. Moreover, their test of the parity condition fails for USD versus currencies of emerging markets. They conclude that the bond yield differential cannot be a reliable tool for anticipating the direction of developing countries’ currencies in the time intervals. The authors also point to the existence of nonlinearities in the medium term of the USD vis-?-vis major currencies, but not vis-?-vis emerging economies’ currencies. Baillie and Kilic (2006) have also observed nonlinearities in the behavior of the US dollar versus other major floating currencies. They have found that the sign of the estimated slope coefficients depends on the sign of the forward risk premium calculated as the difference between forward exchange and spot rates. When the premium on the US dollar is positive the signs of the estimated coefficients for all currencies, excluding the Canadian Dollar and the UK pound, are positive. On the other hand when the difference is negative the estimates are all negative, excluding the Canadian Dollar and the Italian Lira. Similar to Baillie and Kilic, Bansal (1997) provides evidence that the violations of the UIP depend on the sign of the difference between interest rates. His findings show that when the interest rate differentials are positive the slope coefficients are estimated as negative. The reverse is true for negative interest differentials. According to Fama (1984), the reason for the negative slope coefficient is the variance of the forward risk premium.
Download
The Uncovered Interest Rate Parity Condition: A Puzzling Phenomenon
