Carry on investing
Campbell & Company's Susan Roberts (pictured) argues the case for why carry trades should not be viewed merely as an FX play but can, when applied across multiple asset classes, produce enhanced risk-adjusted returns.
Whenever investors hear the term ‘carry trade’ they tend to immediately think of FX carry, and the nasty left tail unwinds that can sometimes arise. But conflating ‘carry’ solely with FX carry and making certain assumptions about the return distribution is a mistake. As Susan Roberts, Product Specialist at Campbell & Company, a US systematic, quantitative investment firm, points out, there is “a fundamental misunderstanding” as to what carry strategies do.
“At Campbell, we have been trading carry strategies in one form or another for well over a decade,” says Roberts. “Though many investors tend to associate carry entirely with foreign exchange, it is a strategy that can be generalised across all asset classes (fixed income, equity, commodities). And furthermore, a multi-asset carry portfolio can have a very different return distribution than a traditional FX-only approach.
“It is a return distribution that has historically been complementary to trend following, and more broadly, to traditional asset allocations as well.”
In a white paper written earlier this year entitled An Introduction to Global Carry, Roberts set out to banish common misconceptions that, in addition to the FX characterisation, also include the idea that carry strategies use significant leverage, and that they tend to be relative value (that is, non-directional) in nature. These are not absolutes, something that Roberts was quick to address in the white paper written with the aim of helping educate investors and demonstrate the advantages of multi-asset carry strategies.
A carry trade will seek to derive returns from the net benefit of holding an asset in excess of the price appreciation/depreciation associated with that asset over time. Carry strategies will take long positions in markets with a positive net benefit and short positions in markets with a negative net benefit, for instance, net cost.
In this sense they are different from directional strategies that seek to generate returns purely from prices rising or falling. If the benefits exceed the costs it is regarded as positive carry and vice-versa. In FX markets, this is achieved by exploiting the difference in short-term lending rates between two economies.
As Roberts illustrates in her paper, on 2 January, 2010, the three-month interest rate in Australia was 3.7 per cent, while the three-month interest rate in the US was 0.1 per cent. An investor who borrowed USD at 0.1 per cent (cost) and invested the proceeds in AUD debt yielding 3.7 per cent (benefit) would have generated profits of 3.6 per cent (assuming the exchange rate was unchanged during the investment).
The same concept applies to other asset classes: fixed income, equities, and commodities. In equities, for example, provided the dividend yield is higher than the financing rate, it would signal an opportunity to go long; and if the opposite were true, it would signal an opportunity to go short.
“At Campbell, we have a standalone carry programme that uses a directional implementation rather than a relative value implementation and therefore doesn’t require the same amount of leverage. It also trades five major asset classes, including volatility, rather than just FX.
“Actually, foreign exchange has not been one of the best-performing sectors for carry strategies in a number of years. We have found more compelling opportunities in other asset classes, such as fixed income and commodities. We let the systems determine where the best risk/reward is at any given time,” explains Roberts.
The way that one defines carry is that it is an asset’s expected total return assuming its price is unchanged.
As such there is an inherent risk to carry, as an asset’s price will fluctuate over time, either enhancing or diminishing the total return. One way to address risk is systematically. With respect to FX, carry trades tend to work when risk aversion is low. As it rises, and market sentiment starts to deteriorate, there is a higher probability that the FX carry trade will unwind.
“One of the systematic things we do to address that left tail, for example, is to use a Risk Aversion Indicator (in our flagship managed futures portfolio, which also trades carry) to monitor market sentiment. It looks at things such as realised and implied volatility, swap spreads, and a number of other indicators that help us analyse the trajectory of market sentiment.
“When the Risk Aversion Indicator rises above a certain threshold, exposure to FX carry will be systematically reduced. We use our own proprietary indicator rather than a bank-produced indicator. This has historically proven effective and allowed us to dynamically adjust FX carry exposure to avoid some nasty unwinds that were on the horizon,” comments Roberts.
The flip side to this is that occasionally the indicator creates false positives. One can think of it as functioning much like an insurance policy on top of a carry strategy; it will step out of a trade from time to time when risk aversion is getting uncomfortably high, but the market may not experience a major unwind. “In which case, we leave money on the table. But we are more than happy to make that trade in order to mitigate tail risk,” adds Roberts.
A second way to address carry risk is by market and asset class diversification.
“There’s a natural tendency for investors to assume that carry strategies in other asset classes will have the same risk factors as foreign exchange carry; however, that is not the case, with the possible exception of VIX carry. Fixed income, equity and commodity carry have not historically demonstrated the same vulnerability to risk aversion.
“That is one reason,” says Roberts, “using a multi-asset class approach, in addition to being thoughtful about portfolio construction, can lead to a return distribution that does not have the same type of tail risk we see in FX.”
Roberts demonstrates using a Hypothetical Global Carry Strategy presented in the paper that in negative periods for FX carry, the correlation to fixed income carry was -4 per cent, while the correlation to commodity carry was even lower, at -15 per cent. The highest correlation was observed with equity carry, at +9 per cent.
Using historical data between January 1992 and November 2015, even though FX carry generated average losses of -4.7 per cent, the Hypothetical Global Carry Strategy actually generated average positive returns of +3.7 per cent.
One of the main reasons for using a generic Hypothetical Global Carry Strategy, says Roberts, was due to the fact that there were no publicly available indices that were a close enough representation of the multi-asset carry approach used by Campbell.
“Many publicly available carry indices focus on foreign exchange. Furthermore, because the implementation of carry can differ so widely, we observed a very low correlation between the multi-asset class indices we did identify, and our own. Carry is often thought of purely as a highly levered, relative value trade and I think many of the publicly available indices are based on that sort of definition,” says Roberts.
There is, she says, unfortunately a misconception that there is some way to access carry while also neutralising price risk but there is no such thing as a risk-free trade. There are implementations of carry that can mitigate spot risk, but in doing so, new risks are introduced.
“For example, a relative value calendar spread approach (for a particular market, go long the contract in the term structure with the highest carry and short the contract with the lowest carry) will neutralise exposure to parallel shifts in the term structure, but increases exposure to the shape of the term structure.
“We weren’t able to find any index that was representative of what we do in our Carry program and that’s why we built a Hypothetical Global Carry Strategy,” she says.
What Campbell found when they added a 20 per cent allocation of the Hypothetical Global Carry Strategy to the Barclay CTA Index (again, using the same historical data set) was that annualised returns, when compared to the 100 per cent CTA Index, increased from 4.7 per cent to 5.8 per cent. At the same time, annualised volatility fell from 7.3 per cent to 6.6 per cent.
“We weren’t too surprised by the result. Many in the trend following space use carry as a component in their flagship portfolio, largely due to the complementary nature of carry and trend.
“However, I don’t think people truly understand the extent to which carry can provide diversification and some investors who read the paper might be surprised at the portfolio benefits of using a multi-asset class approach. We encourage people to look at carry as part of a diversified portfolio rather than on a standalone basis,” suggests Roberts.
Similar risk-adjusted return benefits were discovered when the Hypothetical Global Carry Strategy was added to a traditional 60/40 portfolio.
“The key point is that when investors experience stress periods in their portfolio, we think the inclusion of a moderate allocation to multi-asset carry has the potential to help smooth out the ride and reduce the level of volatility,” concludes Roberts.