Janus Henderson’s Portfolio Construction and Strategy (PCS) conducted an assessment of the key challenges of fixed income in retirement plan lineups. Here, Senior Portfolio Strategist Damien Comeaux explains why using traditional methods to assess the risks associated with certain diversifying fixed income categories may prove to be problematic.
Janus Henderson’s Portfolio Construction and Strategy (PCS) team developed a diagnostic report specific to retirement plans that provides a comprehensive analysis on the risk and reward of fixed income investments in plan lineups. In our conversations reviewing this diagnostic report with DC-focused financial professionals, we identified four areas we believe are the most crucial to consider in order to avoid the complications of implicit and unintended risk.
We cover each of those challenges in detail in our four-part blog series. In a previous post, we looked at some of the specific risks certain diversifying fixed income categories can present to a portfolio. Here, we discuss the troublesome issues presented by the fact that these risks are difficult to assess and understand using traditional methods familiar to us for equity investments and even other fixed income categories.
The categories that contribute the most potential diversification for a plan lineup were created to be intentionally different from the traditional categories, making them a challenge to judge using traditional benchmarks and their standard statistics of risk and return. This creates quite a conundrum for investment committees as they attempt to provide these diversifying options to plan participants.
While traditional Modern Portfolio Theory statistics are generally helpful, it is important to note the shortcomings of benchmark-relative statistics in the evaluation of categories that were explicitly constructed to look very different from the benchmark in order to improve the yield and duration risk/reward profile.
For example, R-Squared is a standard metric used to determine the statistical significance of a data point, with a score above 90 being an indicator that you’re comparing a fund to a “good,” or relevant, benchmark. The median Core bond fund has an R-Squared of 92 relative to the U.S. Aggregate Bond Index. This indicates that 92% of the median Core bond fund’s performance can be explained by the behavior of the benchmark. This is to be expected given the similarities in composition.
However, as more flexibility is granted to managers in a category, this number drops to 70 for Core-Plus and to just 14 for the multi-sector bond category. Of course, that figure shouldn’t come as a surprise; after all, this is exactly what the multi-sector category was designed to do. But this metric does indicate that benchmark-relative statistics have little to no relevance in evaluating these types of funds, since a low R-Squared shows a low level of statistical significance.
Regardless of the additional challenge this may pose, we believe the extra due diligence required for a category like multi-sector bond is well worth it. The category effectively helps address the “too much duration for too little yield” problem we outlined in a previous post in this series.
As demonstrated in the table below, the median multi-sector bond fund has meaningfully less exposure to low-yielding government securities, therefore providing less rate risk with a higher yield over the past five years. That positioning has translated into higher returns that categories with more rate risk had difficulty matching.
Source: Morningstar, as of 6/30/21.
How should DC-focused financial professionals proceed in evaluating this important asset class for potential addition to a plan? One solution could be to isolate discrete time periods of particular risk/reward patterns and compare and contrast manager success solely in that vacuum.
Discrete time periods like 2015 and 2018 are useful to help understand “worst-case scenarios” when examining these categories. In addition, comparing risk-specific stats, such as standard deviation and max drawdown, can help generate useful insights into these investments. Understanding a fund’s correlation to equities is critical as well. While these diversifying categories will naturally take on higher correlation to equities, based on typically lower exposures to government securities (again, as intended), it is important to consider funds with a reasonable correlation to equities as acceptable and those with very high correlations to equities as unacceptable.
Finally, for these strategies designed to look very different than standard benchmarks, understanding the level of risk taken in non-benchmark sectors is yet another way to glean insights for these investment types beyond traditional approaches.
Wondering what critical gaps could potentially be lurking in your plan menu? Download our full assessment of The Four Challenges of Fixed Income in Retirement Plan Lineups.
Core Bond portfolios invest primarily in investment-grade U.S. fixed income issues and hold less than 5% in below-investment-grade exposures.
Core Plus portfolios invest primarily in investment-grade U.S. fixed income issues, but generally have greater flexibility than core offerings to hold non-core sectors and up to 35% in below-investment-grade exposures.
Multisector Bond portfolios seek income by diversifying their assets among several fixed-income sectors and typically hold 35% to 65% of assets in below-investment-grade exposures.
Bloomberg U.S. Aggregate Bond Index is a broad-based measure of the investment grade, US dollar-denominated, fixed-rate taxable bond market.
This information is not intended to be legal or fiduciary advice or a full representation of all responsibilities of plan sponsors and financial professionals.
Janus Henderson is not affiliated with Plan Sponsor Council of America.