Accounting for Uncovered Interest Rate Parity - The Permanent and Transitory Decomposition Approach

Abstract

In most of the empirical research about the uncovered interest rate parity (UIP), estimation of the condition at short horizons rejects this theoretical proposition and presents just opposite outcome, while estimation at long horizons reports more consistent results with the proposition. Based on the belief that permanent components may convey more information in the UIP condition, we adopt in this thesis the permanent-transitory decomposition approach introduced by Gonzalo\r\nand Granger (1995) to estimate the common long memory components\r\nof interest rates and that of exchange rate. We then re-evaluate the UIP condition using the decomposed permanent and transitory parts, respectively. Our results reinforce the stylized facts existing in the literature, while it remains to be answered why empirical evidence does not favor the UIP condition.

In most of the empirical research about the uncovered interest rate parity (UIP), estimation of the condition at short horizons rejects this theoretical proposition and presents just opposite outcome, while estimation at long horizons reports more consistent results with the proposition. Based on the belief that permanent components may convey more information in the UIP condition, we adopt in this thesis the permanent-transitory decomposition approach introduced by Gonzalo\r\nand Granger (1995) to estimate the common long memory components\r\nof interest rates and that of exchange rate. We then re-evaluate the UIP condition using the decomposed permanent and transitory parts, respectively. Our results reinforce the stylized facts existing in the literature, while it remains to be answered why empirical evidence does not favor the UIP condition.