Systematic, or factor-based, investing has become quite common in equities. Investor adoption in fixed income has lagged, at least when measured by the assets under management (AUM) in mutual funds and ETFs. At the end of 2020, $1.35 trillion in equity fund AUM1 was categorized as strategic beta by Morningstar. By contrast, just $14.36 billion of fixed income funds had the same designation.
In Dimensional’s case, systematic fixed income is hardly new; we have been managing fixed income portfolios since 1983. Our approach to fixed income shares many aspects with our systematic equity solutions. In both cases, our goal is to combine the best of indexing, such as broad diversification, low turnover, and transparency, with flexible active implementation to emphasize higher expected returns and manage risk.
The many parallels between Dimensional’s equity and fixed income approaches provide an opportunity to demystify systematic bond investing through the familiar lens of our approach in equities.
Systematic investing typically seeks to outperform markets by structuring investments around factors linked to differences in expected returns. This differentiates systematic investing from traditional indexing, which typically seeks to deliver market returns, and traditional active investing, which may seek outperformance by identifying so-called mispriced securities or timing markets.
Decades of research on stock returns has produced a vast number of published factors. Valuation theory helps us identify relevant factors by providing insights about differences in expected returns across stocks. It tells us discount rates or, equivalently, expected returns link the price investors pay with the cash flows they expect to receive. In equities, this motivates the use of price variables, such as market capitalization and relative price, and cash flow variables, such as profitability, to systematically identify differences in expected stock returns.
A similar principle applies in fixed income, but with the nuance that future cash flows are more clearly defined for bonds. Consequently, we can observe bond yields that link price to future cash flows. For a bond not held to maturity, there may be additional expected gains or losses from selling at a yield different from that at which it was purchased. A forward rate, defined as the sum of these two components, the yield and expected capital gain or loss, therefore provides systematic information about discount rates through the combined information about the price investors pay and the cash flows they expect to receive. Comparing forward rates of bonds with different durations, credit quality, and currency of issuance tells us about differences in their expected returns.
Armed with a grasp of the core drivers of expected returns, we assess evidence on new factors on the basis of whether they add to our understanding of the cross-section of expected returns. Specifically, do additional factors deliver a premium after controlling for the size, value, and profitability premiums? This is standard practice for rigorous academic studies evaluating patterns in average stock returns, such as Fama and French’s survey of well-known “anomaly” variables2 or Robert Novy-Marx’s deconstruction of low volatility strategies.3 In most cases, the return effect under investigation is attenuated or eradicated after controlling for the size, relative price, and profitability premiums.
Estimating expected returns from forward rates
Example: government spot curve, 1-year holding period.
As with equities, new variables in fixed income should be evaluated in the context of the known drivers of expected returns. In other words, do additional bond factors provide information about expected returns beyond what is contained in forward rates? Lee et al4 tested more than a dozen fixed income factors proposed in the literature and found that most return spreads attributable to these variables vanished once controlling for forward rates. In other words, these variables contributed no further information about expected bond returns beyond what was captured by forward rates.
Market prices change every day, meaning a daily process is essential to maintaining a continuous focus on higher expected returns. Past Dimensional research has documented the importance of a consistent emphasis to mitigate style drift5 and capture premiums when they appear.6 This is a key advantage over indexed approaches, which typically rebalance infrequently during arbitrarily prescribed events, such as index reconstitutions.
As with equities, we use current bond prices to identify segments of the market with the highest expected returns. As prices change, bonds characterized as securities with higher expected returns may change. The flexibility to rebalance daily is therefore critical for maintaining a focus on higher expected returns in fixed income.
The need for a daily process becomes apparent when looking at term spreads, or yield differences between bonds of different maturities but similar credit quality, and credit spreads, or yield differences between different tiers of credit quality but similar maturity. Dimensional’s research7 tells us the width of these spreads contains information about the subsequent premiums. We can see this in the charts below, which indicate higher average term and credit premiums during months when term and credit spreads are wider.
We believe broad diversification is the primary tool for controlling risk in both equities and fixed income, adding to the appeal of systematic investing. However, both goals and risks can be more clearly defined for fixed income relative to equities. Investors often use the fixed income component of their allocation to reduce uncertainty around meeting specific objectives. It’s important that a bond portfolio’s holdings are consistent with these goals. Maintaining an appropriate risk profile can be facilitated through a robust credit monitoring process.
For example, investors whose goals dictate an allocation to bonds with higher credit quality may select an investment grade fixed income strategy. Rating agencies such as Moody’s and S&P can provide credit ratings that are helpful gauges to broadly classify bonds and their issuers into tiers based on credit risk. But what happens when a bond is trading at a yield substantially higher than those of similarly rated peers? For example, a bond rated BBB (investment grade) that’s trading closer in price to one rated BB (below investment grade)? Our research8 shows that bonds with yields closer to those of the next tier down in credit quality have a higher frequency of being downgraded over the following year and, in the event of credit spreads widening, experience larger drawdowns on average than their peers—as one would expect of a high-credit-risk security. So outlier bond yields contain relevant information about expected returns and risk that may not be captured by credit ratings.
A lack of flexibility means index strategies hold whatever is in the index—including bonds that meet the index’s eligibility based on a stated rating but are trading like bonds below the index’s minimum rating. That means an investment grade index strategy may at times hold bonds whose risk profile and return behavior are non-investment grade.
Dimensional uses many inputs, including current market prices, to assess the credit risk of bonds each day. Issuers of bonds trading at markedly higher yields than those of peers may be assigned a lower internal credit rating than the stated rating, potentially impacting their eligibility for specific strategies. For example, a bond trading like one rated BBB may become ineligible for a portfolio restricted to securities rated AA and above, even if its stated rating meets the portfolio’s guidelines. Our flexible process allows us to incorporate this information in buy and sell decisions every day.
In many ways, the premise of systematic investing is simple. However, this impression belies the complexity of managing systematic fixed income at scale. Every day, Dimensional calculates forward rates across bonds from thousands of issuers, in a dozen currencies, and from over 20 countries—this amounts to over 20,000 expected return calculations per day. By using these myriad inputs to make informed decisions on how to increase expected returns and manage risk, our daily process seeks to deliver reliable outcomes to investors in a world of ever-increasing complexity.
Systematic fixed income may seem a promising recipe for more reliable outcomes than traditional bond approaches, but investors should remember the role of the chef in translating recipes to dishes. A focus on the key ingredients driving expected returns and risks combined with flexible and scalable implementation are both important when trying to deliver on the promises of systematic investing. For Dimensional, systematic investing is not a recent revelation but a core part of our investment philosophy, as it has been for more than four decades.