The Exchange Rate in the Presence of Transaction Costs: Implications for Tests of Purchasing Power Parity