• Related Articles

• Currency Management and Correlation in Turbulent Markets

• 

  A Convergence of factors: Currency Management has it come of age

• 

  Accessing Currency Alpha

• 

  Tackling the question of multi manager selection

Carry - is it still a viable strategy?

By Adrian Schmidt, Head of Research at C-View

This year has seen a resurgence in higher yielding currencies, after a sharp decline last year when the credit crunch hit. Does this mean it's OK to step back into the "carry trade" assuming that the credit crunch is now behind us? Or should the experience of 2008 be taken to heart and discourage investors from seeking carry? We at C-view would argue that it is always important to be selective in carry trades, and that investors need to focus on value when deciding whether to enter. While there is a sensible rationale behind the concept of the carry trade, in general a value filter should always be applied.

Before examining the rationale behind the carry trade, it's worth looking at the performance of carry trades in recent years. The RBS naïve carry trade index shows an average monthly return of 0.43% in the last 20 years with a Sharpe ratio (assuming no capital charge) of 0.28. Over the last 10 years the numbers are better - 0.46% with a Sharpe ratio of 0.49. But in the last 5 years, the performance has been poor, recording a negative return of 0.4%. Last year was horrible for carry traders - the index lost 31.5% from December 2007 to December 2008. So while the strategy produces a positive return over the long run, just like other assets timing of entry, as well as other issues like asset selection, is very important to avoid major accidents.

What is the rationale behind carry trades? The basic benefit of yield is obvious, but there would be little point in buying higher yielding currencies if all the yield benefit were lost because the currency declined. Some idea of correct valuation is consequently necessary before entering a carry trade. In this sense carry trades in general are the same as any other asset that yields a return. If a currency is dramatically overvalued then the risk of capital loss will outweigh any carry advantage, just as dividend yields usually won't compensate you for buying equities at too high a level. So the value in a carry trade depends on the idea that, when interest differentials are taken into account, higher yielding currencies are not fundamentally overvalued.

Typically higher yielding currencies are regarded as more risky than lower yielders, traditionally because they belong to deficit countries, but nowadays also because they attract investors looking for carry! This tends to mean that there is a risk premium for holding them. This can be seen in the options market, where the risk reversals generally favour the lower yielding currencies, reflecting the greater risk of decline in the higher yielders. Consequently, a risk neutral investor should, on average, be able to gain from holding higher yielding currencies, providing they start at fair value.

There are two things to be concerned about in this story. First, "on average". Carry trades are typically risk positive, so in times of negative risk sentiment they will all tend to lose. This can mean substantial short term losses even from a starting point of longer term fair value. Second, and most important for the longer run, because this is the factor that has changed recently, the concept of "higher yielding". Most carry traders look at simple nominal interest rates, but this is not the logically correct measure, because valuations are affected by inflation. There is little value in 5% rates if a country has 10% inflation, because the currency would be (other things equal) expected to fall 10% a year against a country with zero inflation. For this reason many of the traditional carry trades no longer make sense at the moment. The table below shows the current spreads against the USD, CHF and JPY under different inflation assumptions. It is interesting that while nominal rates still show the yen as the lowest yielding currency, based on most measures of real rates it is now one of the highest yielding currencies among the majors. So as far as major currencies are concerned, it makes little sense to be funding carry trades in yen. The lowest yielder of the majors in terms of real rates is now sterling, based on current inflation, and this makes the pound more attractive as a funding currency, though from a combination of value and carry, the euro and Swiss franc look the most vulnerable.

However, this still leaves aside the issue of valuation. This is not an easy problem to resolve. There is little agreement about the best way to value currencies. Purchasing power parity provides a starting point for many, but is not particularly useful on its own as many currencies consistently trade a long way away from PPP. The reasons for these biases have to be examined to provide a sensible valuation model. There are several alternative approaches based on various concepts of equilibrium exchange rates. There are also purely empirical approaches, essentially assuming mean reversion in real trade-weighted indices. We prefer the approach starting with PPPs, as a purely empirical approach will assume a reversal of systematic real appreciation and undervalue currencies that have good fundamental reasons for showing trend improvement.

There are, in our view, essentially four reasons why currencies deviate from PPP. The first is the level of GDP per capita, which itself reflects productivity, investment and education. Low income countries have weak currencies relative to PPP because they have lower productivity and lower wages and consequently lower prices. Tradable goods prices are largely similar across countries (transport costs excluded) but non-tradable reflect the price of labour. So the price of output in low income countries is lower than in high income countries. PPP, which equalises the price of output between countries, would therefore imply a much higher level of a currency in low income countries than actually exists. The main implication of this is that countries with strong productivity growth with tend to have appreciating real exchange rates.

The second reason for deviation from PPP is the terms of trade, reflected in trade and current account balances. Countries with persistent and large current account surpluses tend to have currencies trading above PPP, while those with big deficits tend to trade below PPP. This reflects the need to attract capital to finance these deficits, or the persistent current account flow into surplus countries. It can also be seen as a risk premium for deficit countries, as repatriation of capital will favour surplus countries.

The third major reason for deviation from PPP is market interference by government. Fixed exchange rates or various degrees of dirty floating can persistently bias exchange rates away from fair value.

Finally, currencies also deviate from PPP because of movements in real interest rates.

So valuation has to be assessed using all these criteria, and only then is it possible to decide whether an individual carry trade makes real sense.

Where does that leave carry trades in today's markets? As the table above shows, there is very little real attraction in the major currencies, where real interest rates are fairly similar, and there is also comparatively little dispersion in GDP per capita growth in the near term. Trend GDP per capita growth is also fairly similar among the major countries' currencies. There is some attraction in Australian yields, but after the AUD's rise this year, this is no longer an attractive entry level. Although the AUD and NZD have performed very well this year, we would argue that their strength has very little to do with the "carry trade". The gains have come from a combination of reversion to correct valuations after the sharp unwinding of carry trades in 2008, and the rise in yields in Australia (and to a lesser extent New Zealand) due to the relative strength of their economies. In other words, strength has had little to do with the level of yields, and much more to do with the change in yields, though the initial low valuation did make both the AUD and NZD attractive carry trades at the beginning of the year.

How to assess valuation? The chart below shows one of our favoured valuation methods. It plots current accounts as a % of GDP against currency deviations relative to PPP against the USD for the major currencies. This metric works well for major currencies because the differences in GDP per capita that make a big difference for emerging markets are less of an issue for the majors. Most of the persistent deviation from PPP among the major currencies is explained by the current account positions, which reflect things like productivity and terms of trade in the longer run The method uses a cross-sectional regression line as a value line, which effectively imposes the idea that the current account as a % of GDP should justify a constant scale deviation from PPP across currencies. Currently, this method shows that the USD and sterling are the cheapest currencies, while the AUD, JPY and CHF are the most expensive. Carry traders should focus on the NZD and NOK rather than the AUD based on this metric, funded by CHF and JPY rather than USD.

There is more attraction in emerging markets. Here, not only are real interest rates attractive in some places, but there is also the prospect of trend valuation improvement because of relative gains in GDP per capita. In many cases, initial valuation levels are also low. The initial valuation level is less of a concern in emerging markets if there is a good prospect of strong growth, as small initial overvaluations can quickly be reversed. Currently, the Israel and Poland offer good value in these terms, while in Latam Argentina, Chile and Mexico all look reasonable value.

However, while there is a good chance of these currencies appreciating in the long term, the road may well be bumpy. As we have seen in the last couple of years, periods of risk aversion can cause some big downdraughts, and while currency investors in for the long haul will probably emerge in reasonable shape, a wholesale buy and hold approach will lead to a lot of volatility and probably poor Sharpe ratios even if they generate decent absolute returns. We would consequently look at strategies which reduce volatility relative to return. This may involve finding funding currencies which reduce risk. For instance, the TWD is currently a reasonable funding currency given its low (negative) yields and the lack of prospect of significant appreciation in the absence of a Chinese revaluation, even though it looks very cheap on fundamental measures, while the AUD and NZD also look high relative to emerging markets, but with similar risk characteristics. The USD is one of the cheapest major currencies here, so funding exclusively in USD looks risky both because of the potential for a risk downturn and because of the potential of a USD recovery if US growth proves stronger than expected. The CHF on the other hand, looks expensive and still provides one of the lowest real yields.

In summary, carry trades can still make sense, but the last few years should have taught us that naïve carry strategies are a recipe for a lot of volatility and not necessarily positive returns. Considerations of valuations, awareness of the potential for risk downturns and a willingness to adjust portfolios, particularly in terms of funding currencies, is necessary if carry as a strategy is to provide a good volatility adjusted return.