The objective of this paper is to show that the proposal by Froot and Thaler (1990) of delayed portfolio adjustment can account for a broad set of puzzles about the relationship between interest rates and exchange rates. The puzzles include: i) the delayed overshooting puzzle; ii) the forward discount puzzle (or Fama puzzle); iii) the predictability reversal puzzle; iv) the Engel puzzle (high interest rate currencies are stronger than implied by UIP); v) the forward guidance exchange rate puzzle; vi) the absence of a forward discount puzzle with long-term bonds. These results are derived analytically in a simple two-country model with portfolio adjustment costs. Quantitatively, this approach can match all targeted moments related to these puzzles.