Currency carry trading is an investment strategy whereby traders invest in high-interest rate currencies and fund in low-interest rate ones. Historically, this trade has generated substantial returns. Typical analysis of the trade has focused predominantly on currencies that yield high interest rate differentials, as the Australian dollar on the long side and Japanese yen on the short side.

However, research by HKUST’s George Panayotov and his co-author Geert Bekaert has challenged these long-held assumptions about the currencies that in fact drive the profitability of carry trades. “While the prior literature takes for granted that the prototypical carry currencies drive carry trade profitability,” say the researchers, “we document the existence of ‘good’ and ‘bad’ currency carry trades.” Most importantly, the “good trades” do not actually use the prototypical carry currencies

Armed with 30 years of trading data for 1984–2014, the team imagined an investor who began trading G10 currencies in 1994. Over a 30-year period, this trader would have chosen the most profitable currency trades as indicated by data available at each trading date. Trades were evaluated by comparing their return and risk, and then classified as either “good” or “bad.” In the process, defined as “dynamic enhancement, certain currencies were exluded from the trades.

The findings were surprising. “Prototypical” currencies long favored by investors, such as the Australian dollar, Japanese yen and Norwegian kroner, were rarely associated with good trades. Instead, the authors explain, “these currencies tend to be excluded from the best trades over and over by the dynamic enhancement rule.” It seems that prototypical carry trade currencies actually offer investors the relatively lowest returns with the highest risk, in a notable divergence from previous studies on this topic. “Trades involving only the prototypical currencies have lower Sharpe ratios and more negatively skewed returns,” the authors explain. “We denote them as ‘bad’ carry trades. The trades using the remaining currencies preserve the desirable features of “good” carry trades.”

However, one finding did back up previous research: the US dollar was never among the currencies excluded from the G10.

The authors suggest that these findings challenge conventional wisdom in the field. “The distinction between good trades and bad trades significantly alters our understanding of currency trade returns,” they note, “and should provide a springboard for future investigations into risk-based interpretations of carry-based returns.”

What lessons, then, can investors take from the research? The authors suggest a starting point. “A carry trade that focuses on the prototypical carry currencies is rather unattractive,” they warn, “but its return properties can be enhanced by exploiting return predictability.”