Working Paper: NBER ID: w17278
Authors: Craig Burnside
Abstract: Carry trades, in which an investor borrows a low interest rate currency and lends a high interest rate currency, have been profitable historically. The risk exposure of carry traders might explain their high returns, but conventional models of risk do not work because traditional risk factors, used to price the stock market, do not price currency returns. Less traditional factors that are more successful in explaining currency returns, are, however, unsuccessful in explaining the returns to the stock market. More exotic models of "crisis risk" are another possibility, but I show that any time-variation in the exposure of the carry trade to market risk has been insufficient, in sample, to explain the average returns earned by carry traders. Instead, peso events remain a candidate explanation of the returns to the carry trade.
Keywords: Carry trades; Risk exposure; Currency markets; Peso events
JEL Codes: F31
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Traditional risk factors (G22) | Carry trade returns (G15) |
| CAPM model (G11) | Carry trade returns (G15) |
| Observed returns (G17) | Carry trade returns (G15) |
| Less traditional risk factors (G40) | Carry trade returns (G15) |
| Peso events (F31) | Carry trade returns (G15) |
| Traditional risk factors (G22) | Stock returns (G12) |
| No unifying risk-based explanation (D81) | Carry trade returns and stock returns (G15) |