Carry trade decomposition

All currency trades can be broken into their interest rate spread component and their spot rate component. Let’s first look at the interest rate spread component across a range of 29 selected currencies. Table 1 shows the currencies color-coded by status: major currency (green), emerging Europe, Middle East, and Africa (yellow), Latin American (orange), or Asia (violet). The carry trade returns calculated in the following formula are based on borrowing at the three-month LIBOR rate of the lower-yielding currency (LY3) and lending at the three-month LIBOR rate of the higher-yielding currency (HY3).

The returns on the higher-yielding currency are adjusted for the daily changes in the
spot rate for the lower-yielding currency (LYS). A 260-day trading year is used. Figure 1 shows what these interest rate carry returns have looked like since the January 1999 advent of the euro. Several observations are in immediate order. First, even on the logarithmic scales used, the low return on the JPY is an outlier; for this reason we will focus on it as a case study. Second, Argentina and particularly Turkey are outliers both in terms of their average daily returns and the standard deviation of those returns.

Third, several G-10 countries with high-yielding currencies, such as New Zealand, Australia, and the UK have unexpectedly low standard deviations of return. Finally, the opposite is true for several Asian countries, including Taiwan and Korea. It appears the central banks of the former group have learned the lessons of transparency, monetary policy, and predictability pioneered by the Federal Reserve, while the central banks of the latter group remain inscrutable.

A second set of observations emerges from a correlation matrix of these returns over the period in question, and it involves several of the countries noted above (Table 2). We might feel safe in assuming short-term interest rates worldwide are positively correlated (green cells) over long periods of time, as this is the nearly universal case for 25 of the 29 countries in question. However, the carry returns for the ARS, BRL, AUD, and NZD exhibit large swaths of negative correlation (yellow cells) against other currencies.

Argentina, which has been a proving ground for every misguided economic policy known to mankind for the past 60 years, is marching to a truly different drummer. Not only has its interest rate return been well above norm, it has been more volatile and more unrelated to the rest of the world as well. Its larger neighbor, Brazil, whose BRL was our topic recently (see “The stronger real: Don’t blame it on Rio,” Currency Trader, April 2007), has been a miniature version of Argentina in some ways, but often exhibits hope in others.

Australia and New Zealand have been in the unusual position of simultaneously having high levels of internal debt and surging export revenues. Their economies are linked increasingly to the East Asian boom, and their monetary policies quite literally must ride the tiger.