A bond investor’s options today may seem like having to choose between a rock and a hard place. Either settle for low expectations in terms of what you may earn on a bond, or try to reinstate yesterday’s fixed income returns by reaching for yield earned through additional risk.
But what if there’s a third way to structure bond strategies? One that acknowledges interest rate levels yet realizes they are only part of the fixed income story? And one that harnesses what’s best about systematic approaches while relying on well-established inputs and daily flexibility to manage risks and increase expected returns? Our research suggests that such systematic approaches may well be the future of bond investing.
While factor-based approaches (those targeting historical drivers or sources of return) have become more common in equity portfolios, systematic approaches in fixed income have faced slower adoption. This is despite academic research that dates back to Nobel laureate Eugene Fama’s work in the 1970s finding that forward rates, as future expected bond yields, contain useful assessments of differences in expected returns among bonds not held to maturity.1 The first strategies that systematically exploited the link between forward rates and expected term and credit premiums—the expected return differences between risk-free bonds of different maturity, and corporate and risk-free bonds with the same maturity, respectively—were launched in the 1980s and now have almost 40 years of live performance. These strategies show that you can beat the market without trying to identify mispriced bonds or time interest rate changes.
As is the case with equity factor investing, investors need to look deeper when evaluating fixed income solution-based factors, such as bond and issuer characteristics. Given the large number of proposed factors, investors may be left wondering which ones to pursue and how to best implement a systematic approach to fixed income investing. One way to gain clarity is to consider what new insights or information an alternative variable contributes. One could ask: Does the variable in question provide new information about expected bond returns beyond what is already known from forward rates?
To answer this question, we undertook a comprehensive analysis of global bonds. Using both portfolio and statistical, or regression-based approaches, we tested many variables to determine what information, if any, they provide about cross-sectional differences in expected corporate bond returns.
A bond’s expected return can be divided into three parts (see Exhibit 1). Numbers (1) and (2) are observable and form the forward rate. Dividing a bond’s expected return in this way suggests that for a variable to contain additional information it must predict cross-sectional differences in changes in yields (3), default probabilities (4), or recovery rates (5). Hence, we examined the predictability of returns in excess of the forward rate.
We studied the performance of more than 17,000 global corporate bonds over the 21-year period from 2000 to 2020. Using third-party credit ratings to measure credit risk, our sample included both investment-grade bonds, those rated BBB and higher, and high-yield bonds, those rated lower than BBB, with credit ratings thereby ranging from AAA to B. Bond-level variables were a bond’s forward rate, the default-adjusted credit spread (sometimes referred to as “value”), as well as bond and credit momentum. Issuer-level variables present in equities asset-pricing research included market capitalization, relative price, and profitability. We further examined an issuer’s use of net debt, or leverage; distance to default; equity momentum, which we define as an issuer’s prior six- to 12-month returns, skipping the most recent month; and short-term equity returns.
Our tests show that most variables do not provide information about expected bond returns that isn’t already in forward rates. The one exception is short-term equity returns.
In contrast to the numerous other variables examined, we found recent performance of the issuer’s stock provides additional information about the cross-section of expected corporate bond returns. If an issuer’s stock underperformed the market in a given month, its bonds tended to underperform in the subsequent month. In other words, recent performance of the issuer’s stock contains information about future performance of the issuer’s bonds, and this information is not subsumed by forward rates. Additionally, short-term equity returns showed little correlation with forward rates.
While short-term equity returns are positively related to subsequent bond returns even after controlling for forward rates, the information contained in short-term equity returns tends to decay quickly—typically within a few months (see Exhibit 2). This time decay, or rate of decline (in information) over time, has direct implications: Implementing strategies based on short-term equity returns might require high levels of excessive trading, thus increasing trading costs.
In contrast, information in forward rates is relatively long-lived (see Exhibit 3). When compared to short-term equity returns, the information in forward rates decays much more slowly, suggesting suitability for cost-effective strategies targeting higher expected bond returns.
We also examined a common measure of “value” for corporate bonds, defined as default-adjusted credit spreads. This measure looks at a bond’s credit spread relative to the issuer’s model-derived default risk. We found this “value” variable does not add information beyond what is already contained in forward rates.
Consistent with prior empirical evidence, we find forward rates continue to be the most reliable and useful metric for identifying cross-sectional differences in expected bond returns. Many other metrics highlighted in research to date are not, in fact, useful as predictors of expected bond returns, providing little information beyond what is already contained in forward rates. Short-term equity returns contain information, although this information decays rapidly, suggesting high potential implementation costs.
Fixed income investing has long been dominated by active strategies that try to outguess market prices and passive funds that are too constrained to nimbly target higher returns. Systematic fixed income solutions go beyond indexing and provide the flexibility to navigate dynamic market conditions in a cost-effective and transparent way.
Just as multifactor equity investing has continued to gain prominence over the last decade, we believe investors will increasingly look to systematic fixed income solutions as reliable and transparent ways of accessing bond markets and targeting higher expected returns.
This piece first appeared in P&I Industry Voices with the title “Targeting higher returns in your fixed income investments? Factors can help, but choose carefully.”