Scientific Beta

In a new survey conducted among investment professionals between June and September 2017, EDHEC-Risk Institute and ERI Scientific Beta have analysed the interests and motivations for investing in new forms of equity factor strategies. The 114 respondents together have at least USD 2.5 trillion in AUM and span all regions of the world (52% from Europe, 28% from North America and 20% from other parts of the world). 

In the past few years, equity factor investing has become increasingly popular among institutional investors and their managers.

At the start, and following the work of Ang et al., one of the motivations for smart equity factor investing was to replace active managers who were considered costly with indices representative of a choice of factors that were well rewarded over the long term.

Since then, factor investing has corresponded to numerous practices and motivations:

• Modifying the factor exposure of a core active or passive cap-weighted portfolio (factor overlay);
• De-risking the equity allocation by adding a low volatility sleeve;
• Replacing the core passive cap-weighted benchmark with a multi-factor benchmark.

More recently, factor investing has given rise not only to total or partial substitutions of active managers or cap-weighted benchmarks but to implementation of dynamic long or long/short strategies that aim to set up genuinely original factor allocation solutions that are referred to as active in the sense that they aim to produce performance compared to static single or multi-factor indices.

The goal of this survey is to learn and understand the interests and motivations for investing in these new forms of equity factor strategies.

Amongst the key findings:

• There is a contradiction between score-based factor design choices and the statistical beta-based risk analysis. Analysis of the extreme risk of factor portfolios is still fairly basic and does not really allow the extreme risks to be appreciated. 
• Multi-factor strategies tend to be implemented in a passive investment context. Even when dynamic, factor investing is generally based on risk budget management rather than active views of returns.
• Investors have a good understanding of the impact of market beta, the conditionality of factors’ exposures to the market and the usefulness of measuring and controlling market beta. This concern largely dominates topics that are the subject of buzz in the market, but are not as important for investors, such as factor timing or the maximisation of factor intensity through portfolio concentration.
• In spite of its limitations, the score-based approach dominates. Even though factor investing was founded on analyses in terms of betas, the measurement of betas is still in a minority and is rudimentary.
• Although valuation-based methods have been widely criticised in academia both for the value bias introduced and for their effectiveness, and methods based on momentum are often highly sample-dependent and criticised for their arbitrary aspect, investors still favour these two approaches.
• When evaluating their control of dynamic factor investing, most investors think that they have poor control of the sophisticated techniques for measuring and integrating the variations in betas and premia in the allocation.
• Performance analysis is consistent with risk analysis and favours the multi-factor beta or score approach. The analyses combining factors with sectors and countries are only reasonably widespread, even though there are quite well developed factor investing offerings based on controlling these three dimensions.