EDHEC-Risk Institute Says Duration-Timing Strategies Can Work

The abundance of theoretical and empirical research on factor investing in the equity universe contrasts strongly with the relative scarcity of research on the existence and exploitability of risk premia in bond markets. Recently, some managers have been focusing on risk premia in fixed income, but the academic knowledge remains limited and it seems that it is not possible to apply the same factors that have been identified for equities in a straightforward way.

It is within this context that EDHEC-Risk Institute has launched a dedicated research programme that aims to broaden the concepts of factor-investing in bond markets by i) analysing the risk factors that drive these universes, ii) finding whether they attract compensation or not, and iii) more generally, examining bond return predictability.

A new study produced as part of the Amundi research chair on ETF, Indexing and Smart Beta Investment Strategies focuses on the two factors that explain a large fraction of differences over time in bond returns, namely the level or slope of the yield curve.

Using not only yield curve data but also a comprehensive database of individual bond returns in the US over the 1973-2018 sample period, the publication Factor Investing in Sovereign Bond Markets – A Time-Series Perspective, explores whether it is possible to identify strategies, which, after transaction costs, generate excess returns by taking relevant signal-based level or slope bets when investing in a real US coupon bonds universe.

The authors of the study, Jean-Michel Maeso, Lionel Martellini and Riccardo Rebonato, draw three major conclusions from their work:

  • They confirm the ?nding that long-term bonds do appear to o?er a higher unconditional excess return over short-term rates, an excess return that is also known as the bond risk premium;
  • They find that a conditional version of a carry strategy based upon a time-varying exposure to the level of the yield curve can generate up to 200 basis points of excess performance;
  • They also find that a conditional version of a flattener strategy based upon a time-varying exposure to the slope of the yield curve can generate economically-significant additional performance, even though such excess performance is limited in implementation by the presence of leverage constraints.

Commenting on this research, Riccardo Rebonato, Professor of Finance at EDHEC-Risk Institute, EDHEC Business School, said Equity markets are notoriously difficult to time, but, as the factor literature has shown us for over 20 years, lend themselves to relative positioning. Treasury markets are, in a sense, their mirror image: relative trading is possible, but hard. However, evidence is mounting that new-generation factors can allow the timing of duration and slope exposure. This is an exciting new field, in which EDHEC-Risk Institute ambitions to position itself as a thought leader.

Bruno Taillardat, Head of Smart Beta & Factor Investing at Amundi, added his thoughts: Demand from investors to apply factors to the fixed income asset class is growing rapidly, opening a new phase for the industry. Since factor-based solutions are not yet as developed on this side as they are on equities, we are delighted to support the EDHEC-Risk Institute with this new academic research, to drive investors education, better understand factor investing in fixed income and build the right investment solutions.

In two companion papers, to be released in June, the same authors will present a related analysis of the two most popular cross-sectional factors, namely value and momentum, using economically justified proxies for these attributes.

A copy of Factor Investing in Sovereign Bond Markets – A Time-Series Perspective can be downloaded via the following link:

Factor Investing in Sovereign Bond Markets – A Time-Series Perspective

You can access the editorial – EDHEC-Risk on Factor Investing in Sovereign Bond Universes – published in the April edition of EDHEC-Risk Institute quarterly newsletter.