The Dynamics of Arbitrage: Evidence from the Forward Markets

Abstract

Covered interest parity (CIP) is the theoretical relationship that explains the price difference between spot and forward exchange rates in terms of the interest rate differential between the home and the foreign currency. CIP arbitrage maintains the parity pricing between a host of financial products traded in currency, interest rate and derivatives markets with different denominations and maturities. However, market frictions cause the parity price to oscillate within a trading band, which is found to vary in size over time. We investigate the drivers behind the arbitrage dynamics using a three regime Bivariate Threshold AutoRegressive (BTAR) model where the bivariate pair is the implied and actual forward exchange rates and the threshold value is the difference between the two. When studying different maturities of the US dollar-Japanese yen relation, one state represents times when US dollar borrowers have a comparative advantage, one state represents times when Japanese yen borrowers have an advantage, and a third state represents white noise around the theoretical rate. The arbitrage profit is asymmetric and varies with maturity as well as state: the largest profit arises when US dollar borrowers have the advantage, while the largest profit variance occurs when yen borrowers have the advantage. Finally, we offer a more precise explanation of asset volatility and the manner it affects the CIP relation: intraday US interest rate (LIBOR) volatility and not exchange rate volatility appears to be the main driver of the trading band apparent in the three regimes. This suggests that CIP should continue during periods of exchange rate turbulence but will likely come to a standstill during periods of interest rate turbulence. These findings support the importance of maintaining liquidity in interest rate and credit markets as additional mechanisms to adjust exchange rate disequilibrium during times of crisis.

Keywords: arbitrage, covered interest parity, international financial markets, forward market, lead-lag relationship, threshold autoregression (TAR), non-linearity test

JEL Classification: C22, C32, E31, F31, G15