Economists Robert M. Dunn, Jr. and John H. Mutti note that financial markets may generate data inconsistent with interest rate parity, and that cases in which significant covered interest arbitrage profits appeared feasible were often due to assets not sharing the same perceptions of risk, the potential for
double taxation due to differing policies, and investors' concerns over the imposition of
foreign exchange controls cumbersome to the enforcement of forward contracts. Some covered interest arbitrage opportunities have appeared to exist when exchange rates and interest rates were collected for different periods; for example, the use of daily interest rates and daily closing exchange rates could render the illusion that arbitrage profits exist. Economists have suggested an array of other factors to account for observed deviations from interest rate parity, such as differing tax treatment, differing risks, government foreign exchange controls, supply or
demand inelasticity, transaction costs, and time differentials between observing and executing arbitrage opportunities. Economists
Jacob Frenkel and Richard M. Levich investigated the performance of covered interest arbitrage strategies during the 1970s'
flexible exchange rate regime by examining transaction costs and differentials between observing and executing arbitrage opportunities. Using weekly data, they estimated transaction costs and evaluated their role in explaining deviations from interest rate parity and found that most deviations could be explained by transaction costs. However, accommodating transaction costs did not explain observed deviations from covered interest rate parity between treasury bills in the United States and the
United Kingdom. Frenkel and Levich found that executing such transactions resulted in only illusory opportunities for arbitrage profits, and that in each execution the
mean percentage of profit decreased such that there was no
statistically significant difference from zero profitability. Frenkel and Levich concluded that unexploited opportunities for profit do not exist in covered interest arbitrage. Using a time series dataset of daily spot and forward USD/
JPY exchange rates and same-maturity short-term interest rates in both the United States and Japan, economists Johnathan A. Batten and Peter G. Szilagyi analyzed the sensitivity of forward market price differentials to short-term interest rate differentials. The researchers found evidence for substantial variation in covered interest rate parity deviations from equilibrium, attributed to transaction costs and
market segmentation. They found that such deviations and arbitrage opportunities diminished significantly nearly to a point of elimination by the year 2000. Batten and Szilagyi point out that the modern reliance on
electronic trading platforms and real-time equilibrium prices appear to account for the removal of the historical scale and scope of covered interest arbitrage opportunities. Further investigation of the deviations uncovered a long-term dependence, found to be consistent with other evidence of temporal long-term dependencies identified in asset returns from other
financial markets including currencies,
stocks, and
commodities. Economists Wai-Ming Fong, Giorgio Valente, and Joseph K.W. Fung, examined the relationship of covered interest rate parity arbitrage opportunities with
market liquidity and
credit risk using a dataset of
tick-by-tick spot and forward exchange rate quotes for the
Hong Kong dollar in relation to the
United States dollar. Their empirical analysis demonstrates that positive deviations from covered interest rate parity indeed compensate for liquidity and credit risk. After accounting for these
risk premia, the researchers demonstrated that small residual arbitrage profits accrue only to those arbitrageurs capable of negotiating low transaction costs. ==See also==