Recently, fixed income markets, along with the alpha of many portfolio managers, have been driven by cyclical credit-oriented risk. As we wrote in a blog post, we think this highlights the need for allocators to have a clear view of their active managers’ biases in alpha generation. This is not to say the credit opportunity is over — we believe there is still a place for an overweight-credit-risk style of investing. But allocators also need to look at the bigger picture and how well diversified they are across their portfolios.
To seek adequate diversification, we believe allocators should consider adopting a robust portfolio construction and manager selection process designed to source complementary styles of alpha-generation to help “smooth out” the path of fixed income alpha. In this note, we’ll highlight two tools that could help: a factor-based framework focused on the pattern of active-manager alpha and a nimble barbell approach to capital allocation.
Applying a factor-based lens
To provide more balance to the alpha generation of fixed income portfolio allocations, our Fundamental Factor Team developed a “role-based” framework for understanding the pattern of active-manager alpha behavior across a variety of style factor categories. Commonly discussed in relation to equities, style factors matter in fixed income as well and may be used to pursue alpha in fixed income portfolios. We break them into five categories (across all asset classes) based on the potential portfolio role and behavior:
Risk seeking — Higher risk, wide range of outcomes, levered to the cycle
Mean reversion — Capital appreciation, believe current discount will revert to the mean
Trend following — Capital appreciation, believe current trend will continue beyond the mean
Risk aversion — Capital compounder, lower risk, countercyclical
Carry — Income, consistency, current income more important than capital appreciation
In fixed income markets, we believe these style factor categories can be naturally diversifying to one another. To illustrate this diversification potential in credit markets, we draw on our Market Environments Framework, a research tool for breaking down the behavior of an alpha stream based on different states of the world. As shown in Figure 1, we began with a broad credit market risk factor, represented by the excess returns of high-yield credit over the corresponding duration-matched Treasuries. We then looked at a how a mean reversion factor and a risk aversion factor within the US high-yield credit market performed (relative to the index) in different regimes for credit risk. The dark blue and light blue bars are the average monthly alpha generated by each factor when returns to credit risk have been leading/lagging (positive/negative).
With an eye to risk management, we also looked at the factor results under more stressed left-tail (orange bars) and right-tail (purple bars) market conditions. Left tail and right tail are defined as the worst 10% and the best 10% of credit market risk factor returns, respectively. The yellow bars (normal) highlight average alpha during the middle 80% of markets.
In “leading” environments (positive), the mean reversion factor realized average monthly alpha of 83 bps while the risk aversion factor averaged -3 bps. In “lagging” environments (negative), risk aversion averaged 35 bps of alpha while mean reversion lost 21 bps. This relationship was even more extreme in the “tails” of the distribution (the “stress” environments), shown in the chart on the right. For example, mean reversion served an important role during normal and pro-risk periods (right tail) while risk aversion provided balance and kicked in during risk-off periods (left tail) when mean reversion tended to underperform.