Navigating Rising Rates
- At some point rates will rise, but look for a gradual increase
- Remember that you own duration primarily for diversification
- True interest rate risk differs across bond market sectors — and duration can be an imperfect guide
- Positioning your bond portfolio
Since the start of 2013, we have been recommending that investors start thinking seriously about interest rate risk in portfolios. Over the last three decades, long-term government bonds rewarded investors with healthy real returns, relatively low volatility and good performance during economic downturns. But at current yield levels, it is hard to escape the conclusion that prospective returns look much worse than the solid performance of recent history. In this article, we discuss the outlook for interest rates in detail as well as the implications for investors’ portfolios.
Interest rate outlook
Long-term interest rates have declined for roughly 30 years but are bounded at zero. It is therefore natural to conclude that, at some point, the trend decline in rates will end and yields will revert back to a higher level. If something can’t go on forever, it won’t (Exhibit 1).
We share the view that rising rates are likely to be a meaningful headwind to high-quality fixed-income returns for the next few years. The U.S. recovery is on more stable footing, and the Federal Reserve (the Fed) looks to be gradually turning its attention away from pushing rates down and toward managing the monetary exit process. As a result, rates should have a tendency to rise. However, because of structural changes in the economy, we think the increase will remain modest by historical standards.
First, and most importantly, long-term inflation expectations remain low and stable. When asked in the late 1970s and early 1980s for their longer term inflation outlook, both households and economists responded with estimates around 8% (Exhibit 2). Today, longer term inflation expectations are roughly 2%–3%, depending on the survey, and they show very little volatility despite large swings in actual inflation.
More than anything else, this anchoring of inflation expectations underpins the low interest rate environment across developed-market economies. Low and stable inflation means lower compensation for the erosion of purchasing power and a lower risk premium because of reduced inflation uncertainty. Success in keeping inflation low over the last 15 years does not guarantee it will remain that way — past performance is not necessarily indicative of future results. But it does emphasize that a return to 1980s-like interest rates probably requires 1980s-like inflation expectations — including a breakdown of inflation-targeting and other institutional changes at the Fed.
Second, monetary policy transparency reduces the odds of big interest rate surprises. The very first time the Fed released a statement after an FOMC meeting was in February 1994 — and even then it did not provide a specific number for the size of the rate hike. Because officials provided virtually no guidance to markets in advance, each Fed action had much bigger implications for the bond market. Today the Fed is setting its funds rate policy with explicit reference to economic indicators — a 6.5% unemployment rate. There is still uncertainty about some of the details, but in general this systematic, rule-like behavior significantly lowers the uncertainty around future monetary policy.
Third, long-term interest rates today are not entirely free-floating. Instead, we see similarities to partially managed exchange rates: rates are market determined to a point, but they are also affected by direct government intervention and at least implicit policy goals. In the exchange rate world, economists often refer to this gray area as “dirty float.”1 Under this type of regime, Fed officials would probably tolerate a moderate glide path higher in rates but might intervene to prevent disorderly outcomes.
Many investors remain concerned about a 1994-like sell-off in the bond market. However, we believe this experience is less relevant today because of low and stable inflation expectations, monetary policy transparency and the Fed’s active intervention in the bond market. While we expect interest rates to rise, our best guess is that the pace of increase will remain gradual.
Duration for diversification
We often hear investors say something like the following: “I own stocks for growth and bonds for income.” But in practice, of course, that is not how it really works. Investors hold portfolios for total return but invest across asset classes for diversification. Diversification is still one of the most fascinating ideas of financial economics: one portfolio can have the same risk but higher total return than another simply by investing in different markets.
This basic principle should guide decisions about an issue many investors are struggling with today: the appropriate amount of bond duration in an environment of rising interest rates. For bonds themselves, the implications of rising interest rates are straightforward: higher rates reduce prices and erode total returns.2 But what do rising rates mean in a portfolio context? What is the right amount of duration in a world with lower expected bond returns?
The answer, it turns out, relates to diversification. Since 1992, longer term Treasury securities (7- to 10-year maturity) have delivered an annualized total return of 7.2% (Barclays). Because of rising rates, we expect total returns for longer term Treasuries to fall to around zero for the next 1–2 years. You can think of this as a downward shift in the distribution of returns — instead of a midpoint of 7.2%, the midpoint falls to zero (Exhibit 3). But the critical point to remember is that, in this example, only the midpoint is changing. Other aspects of Treasury returns — like their volatility and, most importantly, their correlations with other asset returns — need not change just because rates are rising. This means that Treasuries can still have diversification value — they can enhance portfolios’ risk-adjusted returns — even as rates rise.
Exhibit 3: Rising rates mean a downward shift in the distribution of Treasury returns (data since 1992-2012)
Consider a simple example in which an investor can only purchase two assets: an S&P 500 Index and Treasuries with 7–10 years remaining maturity. What is the combination of these two assets that maximizes risk-adjusted returns? Basic portfolio theory tells us that we need to look at the distributions of the two assets — their returns, volatilities and correlation. Since 1992 (shown in Exhibit 4), we would have maximized risk-adjusted returns by investing 21% of the portfolio in the S&P 500 Index and 79% of the portfolio in Treasuries. The portfolio would have had an annualized return of 7.5% and volatility of 5.4% — lower volatility than Treasuries with higher returns. Today a portfolio like this would have a duration of 6.2 years (calculated as the 79% Treasury share times its 7.78 year duration).
When interest rates are likely to rise and the distribution of expected Treasury returns shifts lower (as shown in Exhibit 3), the share of Treasuries in the portfolio should decline. But interestingly, even as expected returns fall to zero, an optimized portfolio would still hold some Treasuries. For example, if we think that Treasury returns will be zero, but their volatility and correlation with equity returns will stay the same, then the portfolio with the best risk-adjusted returns would invest 73% in the S&P 500 Index and 27% in Treasuries. This portfolio would have a duration of 2.1 years. In other words, investors maximizing risk-adjusted returns still benefit from duration even if Treasuries themselves return zero.
More generally, we can show that the optimal duration of any portfolio will depend primarily on two factors: (1) the correlation between Treasury returns and the rest of the portfolio and (2) the difference in total returns between Treasuries and the rest of the portfolio. These two factors could be considered measures of the benefits and costs of diversification, respectively.
Exhibit 5 graphs optimal durations for a portfolio of Treasuries and equities, where the other values are set to those in the previous example. When expected Treasury returns fall, the optimal portfolio duration also falls. However, when correlations, between Treasury and equity returns decline (become more negative), the optimal duration rises. At very low correlations, expected Treasury returns matter very little (the lines are clustered closely together) because the benefits from diversification easily outweigh the costs. This means that it can be worth investing in Treasuries even if expected returns are negative — just like the insurance on your house, it can be worth paying up for things that pay off at the right time.
Exhibit 5: Optimal duration depends on expected returns and correlations
Assumptions: No change in interest rates; six month holding period
Therefore, the right amount of duration cannot be determined on a standalone basis. It has to be considered in a portfolio context, because the main value of duration exposure comes through diversification. With yields likely to rise, the cost of that diversification has increased3 and portfolio duration should fall. However, because of the negative correlation between duration and the returns of riskier assets, high-quality fixed income will still be a cornerstone of any disciplined portfolio.
Measuring interest rate risk
Whatever investors’ interest rate outlook, they must ultimately turn to the practical issue of how to manage interest rate exposure in their portfolios. This typically means a greater focus on duration — the most common measure of interest rate risk in fixed income. Unfortunately, duration can be a highly misleading measure of interest rate risk when making comparisons across products.
Ironically, the best example of the pitfalls of duration might be equities. Technically speaking, the duration of a stock is equal to one divided by its dividend yield. Therefore, a stock with a dividend yield of 2% will have a duration of 50 — even longer than the 30-year Treasury bond.
But no investor realistically believes that equities have more interest rate risk than Treasury bonds. The discrepancy relates to the definition of duration: it assumes that other market factors remain constant. In practice, however, when interest rates change, other determinants of equity prices — like the expected growth rate of dividends — often change too.
As a result, when comparing interest rate risk across fixed-income sectors, we prefer to use measures of “empirical” duration. The conventional measure of duration — which we will refer to as “analytical duration” in the discussion below — is grounded in the bond math of cash flows and discount rates. In contrast, empirical duration is based on the statistical relationship between benchmark interest rates and bond returns. It helps answer the practical question of what returns to expect if interest rates rise.
Exhibit 6 compares the average analytical duration of bond market sectors over the last 10 years to our estimate of their empirical duration.4 For empirical duration, the estimates are based on each sector’s price sensitivity to changes in the 10-year Treasury yield. By design, the empirical duration of 10-year Treasuries is nearly identical to the average analytical duration: if the 10-year Treasury rises by 100 basis points (bps), prices should be expected to fall by just over 8%, regardless of the measure we use.5
However, for all the other sectors the differences are quite significant. For example, investment-grade corporate bonds have an average analytical duration of over six years, but the actual sensitivity to changes in Treasury yields over the past decade has only been about half of that. High-yield corporate bonds and bank loans actually have negative empirical durations, meaning that they have historically posted higher returns during periods of rising Treasury yields.
As a general rule, true interest rate sensitivity tends to fall — the gap between analytical and empirical duration widens — as we move down the credit quality spectrum. For lower quality bonds, the impact of changes in credit spreads dwarfs the impact of changes in benchmark interest rates. The very low or negative empirical durations highlight that investors are not primarily taking interest rate risk when allocating to these sectors — credit risk matters much more.
Positioning your bond portfolio
Given the prospect for rising rates, bond investing becomes a delicate exercise. While all bonds will not react the same way to rising Treasury yields, choosing between bonds of different maturities and sectors can have meaningful impact on prospective returns. As a result, it can be informative to see how different areas of the bond market have performed when Treasury rates have risen in the past.
Despite the secular decline in yields, we found 44 12-month periods in which 10-year Treasury yields rose by at least one half of a percent over the past 20 years. These examples provide some context for fixed-income performance during periods of rising rates.
Start with government bonds. Investors may be surprised to learn that in these periods of rising yields, the typical return on the broad Barclays U.S. Treasury Index was actually positive, posting a median return of 0.28%. This small but surprising gain comes from the combination of a median price loss of 5.14% and a median coupon return of 5.42% — the coupon income modestly exceeded the capital losses.
Today, however, there is much less coupon cushion to absorb price losses if rates rise. If, for example, Treasury yields were to rise in line with historical experience, a 5.14% price drop would dwarf the 0.89% income and lead to a total return loss of 4.25% over a year. It’s a basic but important fact: with low coupon income, duration matters more for total returns. Exhibit 7 illustrates this point across the yield curve. Historically, increases in rates have delivered price losses to any bonds greater than three months, and investors owning bonds with maturities five years or longer experienced negative total returns.
Exhibit 7: Rate rise impact on Treasury curve
Median returns across the Treasury curve in periods when rates have risen (1992-2012)
In Exhibit 8, we illustrate these historical returns across sectors. While all sectors posted a positive median total return, most sectors did experience price losses in the mid-single digit range. However, there are some examples of sectors that have experienced price gains even while Treasury yields rose. Emerging market debt stands out as a strong performer in these environments as sovereign fundamentals in developing economies are less dependent on U.S. rates. Floating rate bank loans have also seen modest price gains, as interest in the asset class has tended to increase as investors flee longer duration assets in rising rate periods.
History demonstrates that bond market performance can have tremendous variance when yields begin to rise. Duration is a significant driver of return, given the price sensitivity of long maturity, fixed-rate issues. In addition, sector exposure matters, as idiosyncratic factors can cause performance of different parts of the bond market to diverge. Lastly, fundamental research can help investors avoid potholes. For example, it may seem that emerging market bonds are the best defense to rising Treasury rates, generating a median return of 14.6% during these historical periods. Keep in mind, however, that 14.6% is the median return, within a range of -19.8% to 37.7%. Therefore, investors should keep a diversified approach, focusing on bonds that offer enough yield (or risk premium) to compensate for potential price volatility.
About Columbia Management
Columbia Management is committed to delivering insight on subjects of critical importance, including insight on financial markets, global and economic issues and investor needs and trends. Our investment team examines the issues from multiple perspectives, and we’re not afraid to take a strong stand or point out opportunities — even when there is no clear consensus.
By turning knowledge into insight, Columbia Management thought leadership can provide:
- A deeper understanding of investment themes, trends and opportunities
- A framework for more informed financial decision-making
Access the insight, intellectual strength and practical wisdom of our experienced team.
Find more white papers and commentaries in the market insights section of our website columbiamanagement.com/market-insights
To learn more about the support and services available to you, contact your Columbia Management representative at 800.426.3750.
1"Dirty Float" in the Bond Market.” Columbia Management Perspectives, March 25, 2013.
2With the caveat that benchmark interest rates and credit spreads often move in opposite directions, which will dull the impact of rising Treasury yields on corporate bonds and other spread sectors.
3That being said, the benefits have arguably increased, too: in recent years Treasury returns have been more negatively correlated with risky assets than in the past.
4Empirical duration estimates shown here are based on historical data and derived from regressions. It is also possible to develop forward- looking estimates of empirical duration, but that is beyond the scope of this article.
5We ignore convexity throughout this discussion.
The views expressed are as of the date given, may change as market or other conditions change, and may differ from views expressed by other Columbia Management Investment Advisers, LLC (CMIA) associates or affiliates. Actual investments or investment decisions made by CMIA and its affiliates, whether for its own account or on behalf of clients, will not necessarily reflect the views expressed. This information is not intended to provide investment advice and does not account for individual investor circumstances. Investment decisions should always be made based on an investor’s specific financial needs, objectives, goals, time horizon and risk tolerance. Asset classes described may not be suitable for all investors. Past performance does not guarantee future results and no forecast should be considered a guarantee either. Since economic and market conditions change frequently, there can be no assurance that the trends described here will continue or that the forecasts are accurate.
Risks include prepayments, foreign, political and economic developments and bond market fluctuations due to changes in interest rates. When interest rates go up, bond prices typically drop and vice versa. Lower quality debt securities involve greater risk of default or price volatility from changes in credit quality of individual issuers.
The Barclays U.S. Treasury Index includes public obligations of the U.S. Treasury that have remaining maturities of more than one year.
The Standard & Poor’s 500 Index (S&P 500 Index) is an unmanaged list of common stocks that includes 500 large companies.
The Barclays U.S. Aggregate Indexis an index comprising approximately 6,000 publicly traded bonds, including U.S. government, mortgage-backed, corporate and Yankee bonds with an average maturity of approximately 10 years. The index is weighted by the market value of the bonds included in the index. This index represents asset types that are subject to risk, including loss of principal.
Diversification does not assure a profit or protect against loss.
Duration — A measure of the sensitivity of the price of a fixed-income investment to a change in interest rates.
Correlation — In finance, a statistical measure of how two securities move in relation to each other