Currency Carry

Today, I’m introducing one of the most crucial investment strategies in the forex market: Currency Carry. Understanding Carry often distinguishes professional investors from amateurs.

This article is divided into five parts. The first part briefly explains what Carry is, the second discusses the historical performance of the Carry strategy, the third part explores the characteristics of Carry strategy returns, the fourth part talks about optimising traditional Carry strategies, and the fifth part discusses strategy ETFs.

Part I: What is Carry?

I believe most readers’ first reaction is: What is Carry? The definition is simple. In the forex market, Carry means borrowing a currency with a low-interest rate to invest in a currency with a high-interest rate. For example, in Australia, the deposit rate is 1.5%, while in Japan, it is almost 0. So, you can borrow yen, convert it to Australian dollars, deposit it in an Australian bank, and earn 1.5% interest. When you need to repay the yen, you convert the Australian dollars back. If the exchange rate between yen and Australian dollars remains stable, the return from Carry is the interest rate differential, i.e., 1.5%.

Economics enthusiasts might object, citing the Uncovered Interest Parity (UIP) concept, which states that although the Australian dollar earns 1.5% more interest than the yen, the Australian dollar is expected to depreciate by 1.5% against the yen, resulting in no net extra return. This theory sounds plausible and logical, but real-world data does not support it.

The UIP example is intriguing, closely associated with the spirit of the University of Chicago, where it’s joked that anyone entering UChicago should forget about investing since it’s the birthplace of the Efficient Market Hypothesis (EMH). Most market anomalies originate from this place. UChicago has a grand saying:

So, whenever you hear about any analysis, no matter how reasonable it sounds, it’s worth validating with data—you might be surprised. Unlike natural sciences, finance and economics often face scenarios that defy equilibrium assumptions, such as the Volatility Smile post-Black Monday 1987 and numerous “Got the bet right and lost all the money anyway” situations (refer to "Uncertainty" for more on limited arbitrage and financial risk management).

Back to our topic. UIP is a beautiful theory that doesn’t reflect reality. Carry returns are not accompanied by equivalent currency depreciation; often, the opposite happens, with high-interest-rate currencies appreciating.

So, Carry strategy: Buy high-interest-rate currencies, sell low-interest-rate currencies.

Part II: Empirical Performance of Carry

Let's define a simple Carry strategy. From the G10 currencies (USD, CAD, AUD, NZD, JPY, EUR, GBP, CHF, SEK, NOK), pick the three with the highest interest rates and the three with the lowest. Go long on the former and short on the latter, with weights of 50%, 30%, and 20% respectively, rebalancing weekly (for reasons, see "Rebalancing"). Here is the historical return and Sharpe ratio for this Carry strategy:

From 1983 to 2009 (27 years), the annualised excess return of this strategy was 6.1%, with a volatility of 10%, and a Sharpe ratio of 0.61. During the same period, the Sharpe ratio of the US stock index was only 0.27 (I used 2% as Rf, which might be higher, so the US stock Sharpe ratio would be even lower).

Data before 1983 is less reliable, but calculations still show similar Sharpe ratios, indicating that the Carry strategy’s performance isn’t just data mining; its effectiveness is sustainable.

This strategy is just a combination of G10 currencies. Some might ask why not simply go long on the highest interest rate currency and short on the lowest? In theory, this should provide the strongest signal and best returns. However, this isn’t the case. Although returns slightly improve (8%), volatility increases significantly (16%), reducing risk-adjusted returns. Why? It’s the old adage: diversification is (almost) the only free lunch in finance (see "The Art of Diversification (I)" and "The Art of Diversification (II)"). Most professional investors rarely place heavy bets on strong signals. For example, Jim Chanos limits any position to below 5%, no matter how confident he is (refer to "Dedicated Short Bias").

Some might argue that weekly rebalancing is too frequent and incurs high transaction costs. Indeed, this is a valid consideration. Monthly or even quarterly rebalancing doesn’t affect risk-adjusted returns but helps reduce transaction costs, which is crucial. For most strategies, transaction costs can erode 20%-50% of profits. For forex, this factor is less impactful due to high liquidity, leading to low transaction costs, especially for institutional investors.

We can also include emerging market currencies, requiring only two standards: sufficient liquidity and no capital controls. Expanding the currency pool significantly enhances risk-adjusted returns, as shown:

With emerging market currencies, the annualised return exceeds 16.4%, with a volatility of about 12%, and a Sharpe ratio of 1.30.

I hope these examples demonstrate the impressive risk-adjusted returns of the Carry strategy.

Part III: Alpha or Beta?

In a previous article, "The Art of Knowing Nothing," I discussed the need to identify the source of a strategy's returns. It should explain why the counterparties would forgo long-term returns; otherwise, positive returns might just be coincidental.

I mentioned five factors:

1. Risk aversion: People pay a premium to buy insurance, where the extra returns come from taking on additional risk. The most common strategy is volatility selling, demonstrated by the following graph.

   

2. Behavioural biases: Systematic mistakes made by people. For example, under-reaction to new information due to anchoring bias, followed by overreaction as the information gets reflected in prices, causing trend phenomena (see "A Better Way to Invest: Factor Investing").

3. Non-profit-seeking behaviour of some market participants: Common examples include central banks and commodity market hedgers. Speculators profit from the losses of these hedgers. Speculation is not pejorative in my view; retail investors' reckless gambling is not my definition of speculation, which should be an intellectual game benefiting the market.

4. Predictable short-term behaviours of some market participants: Extra returns come from opportunistic trades. Global macro investors often express views on macro events through direct risk exposure, like interest rates, but hedge with the cheapest bonds, often 7-year bonds. This predictable behaviour causes 7-year bonds to be overvalued relative to synthetic 7-year bonds (5-year and 10-year bond combinations). A simple way to profit is to long the 5-year and 10-year bonds while shorting the 7-year bonds, remaining market-neutral.

5. Market distortions due to regulation/risk management: Fischer Black believed that all regulations create opportunities for extra returns, a philosophy he maintained at Goldman Sachs (see "Academic Duels in Financial Markets").

Which category does the Carry strategy fall into? There is still debate. I believe Carry was initially an alpha strategy exploiting behavioural biases, gradually evolving into a beta strategy with systemic risks.

All explanations are based on phenomena, so I’ll list the characteristics of the Carry strategy for you to judge. Firstly, Carry goes long on high-interest-rate currencies and shorts low-interest-rate currencies, indicating its risk attributes. The classic Carry pair is AUD/JPY, with the former being a high-risk currency and the latter a "safe haven" currency. The former depreciates when market volatility and uncertainty rise, while the latter appreciates, revealing Carry’s beta nature. By the way, I don’t agree that the yen is a "safe haven" currency; its appreciation during crises is related to Carry. During economic booms (risk on), people short low-interest assets to buy high-risk assets for leveraged returns (buying houses follows a similar logic). In volatile times, high-risk assets suffer, and investors might sell high-risk assets and repay low-interest assets to reduce risk exposure. Yen appreciation during market volatility reflects increased demand to cover short positions during risk-off periods.

The Carry strategy also has a catastrophe insurance nature. It performs poorly when market liquidity declines and volatility rises: 36% drawdown in 2008, 28% in 1993, and 26% in 1986. Selling catastrophe insurance offers high long-term risk-adjusted returns, but the cost is bearing huge losses during disasters (see "The Curse of Insurance: Hedging Black Swans").

Another factor is strategy crowding. Ten years ago, Andrew Lo introduced the concept of Hedge Fund Beta. Hedge funds aim for alpha strategies, but high personnel turnover leads large funds to use similar strategies, making returns systemic. Funds simultaneously enter and exit the same assets, turning alpha into a degraded beta. Hedge Fund Beta is a risk factor behind the Carry strategy (see "Facing Death: The 2007 Quant Crisis").

Part IV: Carry Refinement

Previously, I mentioned two ways to improve the Carry strategy: 1) Reduce rebalancing frequency from weekly to monthly or quarterly, maintaining risk-adjusted returns while lowering transaction costs. In forex markets, transaction costs are low. 2) Include some emerging market currencies.

Here are a few more potential improvements:

1. Tail Hedging: Carry acts like disaster insurance, performing poorly during crises. To avoid tail risk, use options for tail hedging. Although I previously advised against buying insurance in investments ("The Curse of Insurance"), forex option premiums are low and don’t significantly impact risk-adjusted returns. Historically, equity index options were similar until Black Monday 1987, after which premiums soared.

2. Strategy Diversification: In "The Art of Diversification (II)," I discussed Carry as a static strategy excelling when markets are stable, while trend-following is a long-tail option strategy thriving in volatile markets. Combining both provides robust diversification and high risk-adjusted returns, akin to value and momentum diversification (per Fama French definitions). Simple combinations yield great returns, but better yet, find currencies with both Carry and trend-following traits. Integration is better than addition. Specific methods are proprietary, but such strategies are excellent for hedge funds.

3. Carry Timing: Timing factor strategies is challenging, but Carry timing holds value. Position sizes can be adjusted based on forex option implied volatility or market risk indicators like Morgan Stanley’s Global Risk Demand Index. Here are some examples:

   

4. Interest Rate Adjustment: The initial strategy was based solely on interest rate differentials. Standardizing these with implied volatility yields better historical performance (similar to Information Ratio adjusted alpha). The original Carry strategy had a Sharpe ratio of 0.61, improved to 0.7 with implied volatility adjustments.

5. Avoid High Inflation Currencies: A subtle but important point, discussed further.

These improvements are worth trying. Simple optimizations often outperform highly mathematical hedge fund strategies developed by PhDs.

Part V: Carry ETFs

I discovered two Carry strategy ETFs/ETNs: ICI (iPath Optimised Currency Carry ETN) and DBV (PowerShares DB G10 Currency Harvest ETF), with expense ratios of 0.65% and 0.8%, respectively—reasonable for strategy ETFs.

A quick look reveals unsatisfactory performance for both.

Below is the performance of the benchmark (World Carry Excess Return Index USD ER).

Consider adding these ETFs to your portfolio to see if they improve risk-adjusted returns.

My view: ETFs are positive developments, replacing many non-value-adding active managers. However, ETFs lack customization. If we want to tweak Carry settings, ETFs can’t offer our desired version. Strategy ETF expense ratios typically exceed 0.75%, sometimes nearing 2%. Compared to hedge funds, this isn’t high, but implementing these strategies yourself incurs much lower costs, often under 0.1%. Another issue is due diligence lagging behind ETF growth, with some ETFs showing significant tracking errors and disappointing results. ICI and DBV are examples. I haven’t done detailed analysis, only reviewed recent performance.

This is a brief introduction to the Currency Carry strategy. There’s much more to discuss, with many opportunities behind it. However, overly detailed introductions don’t benefit me and may give readers a false sense of systematic knowledge—articles are mostly fragmented knowledge. True understanding comes from extensive reading, data analysis, thinking, experience, and backtesting.

I hope you found this helpful.

References:

• Asness, C.S., Moskowitz, T.J. and Pedersen, L.H., 2013. Value and momentum everywhere. The Journal of Finance, 68(3), pp.929-985.

• Balchunas, E., 2016. The Institutional ETF Toolbox: How Institutions Can Understand and Utilize the Fast-Growing World of ETFs. John Wiley & Sons.

• Bhansali, V., Davis, J., Dorsten, M.P. and Rennison, G., 2015. Carry and Trend in Lots of Places. Available at SSRN 2579089.

• Brown, A., 2015. Financial Risk Management for Dummies. For Dummies

• Ilmanen, A., 2011. Expected returns: An investor's guide to harvesting market rewards. John Wiley & Sons.

• Jurek, J.W., 2014. Crash-neutral currency carry trades. Journal of Financial Economics, 113(3), pp.325-347.

• Narang, R.K., 2013. Inside the Black Box: A Simple Guide to Quantitative and High Frequency Trading (Vol. 883). John Wiley & Sons.

Disclaimer: The data and information mentioned are from third-party sources, and accuracy is not guaranteed. This article shares information and views, not professional investment advice. Consult professional advice before making investment decisions.

This article was originally written in Chinese and posted on my WeChat platform in 2016. The Chinese link can be found here