Amundi has provided further fuel to the debate around factors after it unveiled research that suggests factor-based investing can also be applied to investment grade bonds.
The new research shows that since the 2008 financial crisis, when looking at euro issuance bonds, the behaviour of investment grade corporate bonds can be explained by risk factors that have proved to be sustainable over time.
These are the traditional factors for debt instruments – duration, duration-times-spread and liquidity risk – as well as alternative risk factors such as value and momentum.
The long study, which backtests from 2003 to 2018, suggests moreover that the behaviour of the investment grade corporate bond market is better explained by risk factors than by the traditional Capital Asset Pricing Model (CAPM).
Yet, complicating matters somewhat, the research shows that up until 2008 the traditional factors explain the market in bonds. However, after 2008, the market is best explained by adding an overlay of the alternative factors provides an “even more relevant interpretation grid”.
In the case of value, it enables the identification of the relative value of bonds in relation to each other, considering other explanatory factors such as sector, geographical risk or duration being equal. For momentum, the factor consists of monitoring stocks that have tended to outperform in the near previous period, adjusted for duration.
Co-author of the study Jean-Marie Dumas, head of fixed income solutions at Amundi, said: “We have been able to both quantify the behaviour of our alternative factors in credit management as well as position them into a multi-factor framework integrating traditional factors.”
Looking at the difference pre- and post- the great crisis, Dumas suggests that in the 2003-08 era the traditional factors already “captured” the additional information captured by the alternative risk factors.
However, from 2009 onwards, the adding of the extra alternative factors “improves the explanatory power with limited collinearity, suggesting that there is additional information captured by Amundi’s alternative factors”.
Dumas et al add that when looking at the most prominent risk factors across the board to explain the euro-denominated investment grade bond market return, a “combination of traditional (duration-times-spread) and alternative risk factors (value and momentum, in this order) would be appropriate.”
“Because value and momentum display complementary pay-offs, combining them in portfolio construction can prove attractive for investors seeking a robust framework across the investment cycle,” they add.
In comparison, in the US dollar universe the authors suggest that value was already a significant factor for active managers already in the 2003-2009 period and remains significant in the 2009-2018 period, “although collinearity is higher than in the euro universe”.
How this kind of parsing of backtest results sits with modern factor theory might be open to question. As ETF Stream has discussed recently, experts at Scientific Beta have previously bemoaned the degree to which the commercial implementation of factors has led to a serious divergence from the evidence produced by academic studies.