Carry trades borrow in low-interest-rate currencies and lend in high-rate currencies, capturing interest rate differentials. Uncovered interest parity implies these excess returns should not exist if currencies adjust to equalize expected returns. However, persistent violations suggest risk premia or survey evidence of systematic currency forecast errors that traders exploit.
From your prerequisite on uncovered interest parity (UIP), you know the benchmark prediction: expected exchange rate depreciation should exactly offset the interest rate differential between two countries. If Brazil offers 10% annual interest and Japan offers 1%, UIP predicts the Brazilian real will depreciate 9% against the yen over the year, leaving investors indifferent between holding either currency's assets. The carry trade is a direct bet that UIP will fail — and historically, across many currency pairs and time periods, it has failed in a remarkably consistent direction.
The mechanics are simple. A carry trade borrows in a low-yield funding currency (classically the Japanese yen or Swiss franc), converts to a high-yield target currency (Australian dollar, Brazilian real, Turkish lira), and invests in target-currency interest-bearing instruments. The gross profit is the interest rate differential — the "carry." If the exchange rate stays flat or depreciates less than UIP predicts, you pocket the spread. If the target currency appreciates (the opposite of UIP's prediction), you earn the carry plus a capital gain. The strategy pays off as long as UIP does not hold exactly.
Why does UIP fail? The forward premium puzzle is the empirical finding that high-interest-rate currencies tend to appreciate, not depreciate, in the short run — the opposite of what UIP predicts. One framework interprets the carry premium as a risk premium: carry trades deliver small, steady gains most of the time but suffer large, sudden losses when carry crashes occur. This is the profile of selling insurance: you collect the premium repeatedly until the catastrophic event materializes. In 2008, as global risk appetite collapsed, carry traders simultaneously unwound positions — selling target currencies and buying yen — causing the yen to appreciate sharply and wiping out years of accumulated carry gains in weeks.
The carry crash risk explains why the strategy is not simply free money from a market inefficiency. Carry returns are negatively skewed and crash precisely when investors most need liquidity — during global recessions and financial crises. This makes carry exposure systematically undesirable from a portfolio standpoint: it pays positive returns in calm markets but is negatively correlated with the marginal utility of wealth during crises. Empirical Sharpe ratios for carry strategies may look attractive, but adjusting for the timing and severity of crashes reveals that the returns compensate for a specific kind of tail risk, not an exploitable anomaly.
Practitioners typically construct carry trade baskets — going long the highest-yield currencies and short the lowest-yield currencies across ten to twenty currency pairs simultaneously. Diversification reduces idiosyncratic currency risk while maintaining the aggregate carry exposure. Momentum overlays (entering positions when recent price trends confirm the direction) have historically improved performance by partially timing exits before crashes. The fundamental risk management discipline: pre-specify stop-loss rules. When a target currency begins depreciating sharply, it often signals the unwind is beginning, and mean-reversion bets in a deteriorating carry environment are how positions turn into catastrophic losses.
No topics depend on this one yet.