The discovery of factors in fixed income

Scott Longley

a pile of colored pencils

In similar fashion to the dwindling faith in active management within the equity space, managers are now coming to questions whether the same erosion of trust is taking place in the ability of fixed income managers to control interest and credit-risk exposures.

Professors Lionel Martellini, Riccardo Rebonato and Jean-Michel Maeso from the EDHEC business school suggest the steady decrease in interest rates in recent years has led institutional investors to question whether the active skills in this asset class can still work.

Moreover, they suggest that some asset managers, including AQR, Robeco and Blackrock, have begun to cater for this institutional demand by proposing products that "focus on the efficient harvesting of fixed income risk premia".

Martellini et al suggest the take-up rate "from a virtually standing start has been impressive".

According to a recent EDHEC ETF survey, the percentage of asset owners investing in smart beta products has increased from 25% in 2014 to 51% in 2019. In the same year, only 21% of the institutions interviewed did not invest, or did not consider investing, in such products in the near future.

The authors suggest that from a practical perspective, one major benefit can be expected from shifting the focus from asset classes to risk factors.

“In the ‘new view of the investment world’ assets are not fundamental, but just carriers of rewarded factor exposure,” they add. “Identifying these common exposures allows clearer risk control, and cheaper, more liquid and more transparent access to the underlying sources of returns. This is especially attractive in those markets where the traditional active investment vehicles has been found wanting.”

However, the factor approach still poses two major challenges – the identification of the rewarded factors and then the subsequent construction of the associated investable proxies.

“In equity markets, a relative consensus exists regarding the relevant factors, including notably, value, size, momentum, low risk or quality, and numerous funds exist that are designed to efficiently harvest the associated risk premia.”

But in fixed income, the consensus view is yet to “crystallise” leading to what Martellini et al suggest is the most exciting area of smart beta research.

“Industry players, such as AQR and Robeco, have highlighted the existence of risk factors in the US corporate bond universe such as momentum, value, carry, or size and have already developed two funds aimed at capturing the associated  rewards,” they suggest. “But a lot of research remains to be done, and a lot more research is needed before risk premia can be extracted in fixed-income as efficiently as in the equity space.”

Emerging differences

The professors point out, though, that some important differences between equity factors and fixed income are already emerging. While equity factor approaches have been mainly cross-sectional - playing one set of equity stock against another - there is ample evidence that directional predictability i.e. market timing in fixed-income is much stronger compared to equity markets.

“In addition, the traditional cross-sectional strategies seem to offer exploitable sources of profitability also in fixed income,” they add.

One conclusion form this is that investors are now at an important juncture in the road to successful investing and that the new techniques being deployed, if rooted in “economically motivated source of predictability” could usher in a new era of management of the exposure to change in the yield curve.

“Data mining and overzealous factor hunting are always a clear and present danger, and this is why it is necessary to combine the market experience offered by established managers with the rigorous and dispassionate economic analysis that only academia can provide,” the team write.

As a step in this direction, the EDHEC Risk Institute has been able to provide a first detailed, security-level analysis on two factors that explain a large fraction of the differences in the cross-section of bond returns, namely value and momentum, using economically justified proxies for these attributes.

“A lot more work needs doing, but we believe that we are truly witnessing a new dawn in fixed-income investing,” they conclude.


No ETFs to show.