Investigating whether factor premiums will survive higher inflation
With the strong economic recovery witnessed so far in 2021, led by the reopening of the global economy following the start of Covid-19 vaccination schemes, inflation worries have gained momentum. We investigate in our article1 how inflation has historically affected credit factor premiums. Overall, we find that while individual factor premiums in corporate bond markets can be somewhat sensitive to inflation changes, the multi-factor strategies provide all-weather performance across inflation regimes. Our conclusions are robust for the choice of investment universe (investment grade or high yield), for unexpected inflation changes and changes in inflation expectations, and after controlling for state-dependent market exposures. Ultimately, we find that investors’ alpha is best protected against rising inflation when combining factors in a multi-factor portfolio.
Co-Head of Quant Fixed Income and Lead Portfolio Manager
Ultimately, we find that investors’ alpha is best protected against rising inflation when combining factors in a multi-factor portfolio.
Defining the scope
We analyze the low-risk, quality, momentum, size and value factors that are an integral part of our multi-factor model and have been documented in the academic literature as well as numerous white papers. Our data covers the global investment grade and global high yield universes over the period from January 1994 to December 2020.
To investigate factor performance, we create long-short factor portfolios by going long the top decile portfolio and short the bottom one. Our focus is on the pure credit return associated with corporate bonds by measuring their performance in excess of duration-matched Treasuries.
To better understand the impact of inflation changes for long-only investors harvesting credit factors, we also analyze a multi-factor long-only portfolio. Here we rely on historical simulations for our flagship products, namely the Global Multi-Factor Credits and Global Multi-Factor High Yield strategies, which we complement with live track records to extend the analysis to the end of June 2021.1
Measuring inflation sensitivity
First, we run scenario analyses, i.e., we measure factor performances in months with rising or decreasing inflation expectations and with rising or decreasing unexpected inflation changes. A rising inflationary regime is defined as a positive change, while a decreasing state is defined as a negative change.
In Figure 1, we first investigate the sensitivity of factor premiums to unexpected inflation changes. The chart on the left shows the Sharpe ratios of the individual long-short factor portfolios conditional on a rising or decreasing inflationary regime, for both investment grade and high yield. We see that the Sharpe ratios of the factors vary with the regime and are generally lower in regimes of decreasing inflation. This lower Sharpe ratio in deflationary times is typically the result of a lower return, but especially a higher risk.
In the chart on the right, we investigate the sensitivity of the Sharpe ratios of the market indices and the information ratios of the multi-factor strategies. We find that the returns of the investment grade credit market are higher when inflation increases unexpectedly, while the Sharpe ratio of the high yield market seems insensitive to unexpected inflation changes. The long-only multi-factor strategies deliver highly significant information ratios in all environments, for both investment universes, that are higher when inflation surprises on the upside. The higher information ratios are mostly the result of lower tracking errors in inflationary regimes, while the outperformances are hardly affected.
Figure 1 | Sensitivity of factors to unexpected inflation changes
Source: Robeco, Bloomberg. Period: January 1994-December 2020 for individual factors’ Sharpe ratios (left chart) and January 1994-June 2021 for the market indices’ Sharpe ratios and the multi-factor strategies’ Information ratios (right chart).
With regard to regimes of rising or decreasing inflation expectations, our analysis shows this time that individual factor premiums are more sensitive to the regime. When investigating the sensitivity of the Sharpe ratio of the market as well as of the information ratio of the long-only multi-factor flagship solutions, we find that while both the investment grade and the high yield market are sensitive to changes in inflation expectations, the long-only multi-factor strategies deliver highly significant risk-adjusted relative returns in all environments, for both investment universes. The multi-factor models benefit from the diversifying inflation sensitivities of the individual factors and offers a robust all-weather performance.
Controlling factor premiums for inflation sensitivities
The previous analysis shed light not only on performance differences but also on risk differences across inflationary regimes. These findings suggest that inflation and market regimes are related and that the market sensitivity of factors might vary across inflationary regimes. In this second analysis, we therefore investigate whether the conditional factor performances are robust after controlling for market exposures.
We see in Figure 2 that all long-short factors generate positive and significant alphas above and beyond their state-dependent market sensitivity, when we differentiate between positive and negative unexpected inflation changes. The alphas are robust across inflationary regimes and investment universes, with somewhat higher alphas in times of positive inflation surprises. While the alphas remain positive in all inflation regimes, their statistical significance for some factors can be regime dependent. This is in contrast to the long-only multi-factor strategies, which earn a consistently stable alpha across inflationary regimes, after controlling for state-dependent market exposures. We benchmark these alphas against the market indices average performance conditional on the inflation regimes and find that the magnitude of the long-only strategies’ alphas is consistently larger than the conditional market performance.
Figure 2 | Alpha beyond sensitivity to unexpected inflation changes
Source: Robeco, Bloomberg. Period: January 1994-December 2020 for individual factors (left chart) and January 1994-June 2021 for the market indices and the multi-factor strategies (right chart). For the market indices, we report the conditional average return.
We perform a similar analysis when considering the sensitivity of factor premiums to changes in inflation expectations. We find similar results as in the case of unexpected changes in inflation.
1This article is a shortened version of the article “Are credit factor premiums robust to inflation?”, September 2021, Thibault Lair and Patrick Houweling. We refer refer the reader to this full version for a more detailed analysis.
2The live track record of the Global Multi-Factor Credits and Global Multi-Factor High Yield strategies start respectively in July 2015 and July 2018.