First: the takeaways

  • Expected return for an asset class is a combination of carry and price appreciation
  • If carry is high enough, it can potentially dwarf price depreciation due to rising rates
  • Higher carry assets (e.g. high yield bonds, MLPs, REITs, bank loans, EM debt, et cetera) have historically had low-to-negative estimated duration profiles
  • High carry assets have a variety of associated risk factors, representing an ability to generate income and diversify away from interest rate risk

So what is carry?

In their paper "Carry," Koijen, Moskowitz, Pedersen and Vrugt define carry to be and asset's "expected return assuming that market conditions, including its price, stays the same."

Using this definition, we can decompose an asset's return into three categories: carry, price appreciation (or depreciation), and unexpected shocks.  The first two – carry and price appreciation – define the expected return.  Unexpected shocks are volatility.

Carry me away

For rate sensitive instruments – like bonds – a rising rate environment can present a very real risk to price return.  In fact, expecting a rising rate environment can lead to a negative expected return profile.

That is, of course, if carry is not sufficiently large to outweigh the impact of the rising rates.

In a prior post, I touched upon a simple formula to determine how rising rates might affect the returns of a constant maturity fixed-income portfolio.  The underlying tug-of-war was between carry and price appreciation: when rates rise, carry increases, but price decreases.  At a certain point, the increased carry outweighs the losses.

With Treasury rates near all-time lows, it is unlikely that carry will make up for price losses in the near future – but what happens if we start with high carry asset classes?


High-Carry Assets

To keep things simple, I am going to simply define carry as trailing 12-month dividend yield.  Traditionally, carry would include things like futures roll and short-term rates – but for this exercise, the exact carry value is less relevant than whether an asset is high carry or not.

Based on this definition, I've chosen to evaluate the following asset classes: high yield bonds, MLPs, bank loans, convertible bonds, emerging market debt (local currency and USD denominated), S&P 500 BuyWrite, preferreds, mortgage REITs, U.S. REITs and international REITs.

Scenario Analysis

I am not a big fan of fixed-period based analysis.  I find that you end up losing a lot of information about what happens intra-period based on some arbitrarily chosen start and end points.

October 2014 is my favorite recent example: U.S. equity markets fell over 7% intra-month before rebounding and closing the month with a positive 2.3% return.  Month-to-month analysis completely loses this volatility.

Fortunately, my co-portfolio manager, Justin Sibears, wrote an awesome algorithm to break a time-series down into different scenarios.  That's exactly what we're going to use here to look at changes in 10-year rates and see how these high carry assets reacted.

Below I plot the different scenarios the algorithm identified in 10-year U.S. Treasury rates from 12/31/2008 to today.

Rate Scenarios

We can see the algorithm identified 11 unique scenarios that breaks the series into distinct periods of large changes and flat movements.

Implied Duration

For each of these scenario periods, we compute the total return for the high carry asset class over the same period.  The 2nd column shows the returns for 7-10 Year U.S. Treasuries.

Rate Scenario Returns

At the bottom of the graph, we calculate a simple implied duration (based on a linear regression) based on these returns and the change in 10-year rates over the periods.

What we can see is that while 7-10 Year U.S. Treasuries have a duration of 7.5 – meaning that if rates go up 1%, we expect Treasuries to lose 7.5% – most of the high carry assets have low, or even negative, duration.  This does not actually make sense for many of these asset classes.  High yield, for example, certainly does not have negative duration – at least not in the instantaneous sense.

Since we are evaluating scenarios that unfold over several months, other aspects come into play.  First, the higher carry helps reduce overall return exposure to rate changes.  Second, many of these assets have large exposures to other risk factors, including credit, oil, and even currency.  The independent nature of these other factors helps diversify returns away from interest rate exposure.

Of course, when these risk factors converge, all hell can break loose.  We saw this in the 5/2/2013 to 9/5/2013 scenario: the "taper tantrum" and associated market hangover.  Rising rates typically aren't an issue if the associated economic outlook is positive – but when the market response is, "if rates are going up, then I'm not longer going to buy risky assets," then we get a nasty convergence of returns.

And the shorter the horizon of the scenario, the less impact carry can have in outweighing it.

Where lies the risk?

High carry isn't free: that high yield bond is higher yielding because it has greater associated risks.  But those risks are often different than just interest rate sensitivity.

And that is a diversification opportunity.

Similar to what we did with rate changes, we can break down other potential risk factors into scenarios and evaluate the sensitivities of the high carry assets.  Below are the uniquely identified scenarios for US equities, credit spreads, the U.S. dollar (against a broad international basket), and oil.

US Equity Scenarios Credit Spread Scenarios dollar-scenarios Oil Scenarios

Statisticians would probably want to run a multi-linear regression here and do all sorts of tests for collinearity within the risk factors.  Unfortunately, we cannot do that since the scenarios for each risk factor are disjoint.

This means that any sensitivities we arrive at from a simple linear regression run the risk of being due to a spurious relationship.  Hopefully, by analyzing over a long enough time horizon, we reduce that risk.

Scenario SensitivitiesIt should be noted that these are not apples-to-apples sensitivity numbers since the volatility of the underlying risk factor has to be considered.  A several hundred basis point move is astronomically large for credit spreads and 10-year rates – for equities, it's just another day in the market.  To make these numbers comparable to one another, I've multiplied them by the realized annualized standard deviation of each risk factor, allowing us to ask, "how much would we expect the asset to gain or lose based on a standard deviation movement in the underlying risk factor?"

Scaled Scenario Sensitivities

What we see is that high carry instruments that may have low-to-negative exposure to interest rates have enormous exposure to other risk factors.  Most significantly for many of these asset classes are credit spreads and broad equity market performance.  The high carry plus this diversified exposure may help buoy returns in a rising rate environment.


It is no secret that fixed-income is vulnerable to price depreciation in a rising rate environment.  The expected return from this depreciation can be offset, however, by the carry of the asset.  But in today's low yield environment, you've got low carry and high duration.

Enter high carry assets.  With a more significant yield profile and a diversified base of risk factors, they may present an opportunity to investors to achieve income while sidestepping material losses due to rising rates.

That isn't to say that high carry assets are not without their own risks: it is, after all, why they are high carry.  Creating a portfolio of such assets, however, may be able to diversify exposure across enough sufficiently independent factors to limit exposure to any single one.

At Newfound, we manage a diversified portfolio of high carry assets that embraces a positive-trend, high risk-adjusted carry approach.  You can learn more about the portfolio here.

Corey is co-founder and Chief Investment Officer of Newfound Research, a quantitative asset manager offering a suite of separately managed accounts and mutual funds. At Newfound, Corey is responsible for portfolio management, investment research, strategy development, and communication of the firm's views to clients.

Prior to offering asset management services, Newfound licensed research from the quantitative investment models developed by Corey. At peak, this research helped steer the tactical allocation decisions for upwards of $10bn.

Corey is a frequent speaker on industry panels and contributes to, ETF Trends, and’s Great Speculations blog. He was named a 2014 ETF All Star by

Corey holds a Master of Science in Computational Finance from Carnegie Mellon University and a Bachelor of Science in Computer Science, cum laude, from Cornell University.

You can connect with Corey on LinkedIn or Twitter.