Scientific Beta

This paper discusses why robustness is essential for investors utilising smart beta strategies and describes the sources of deficiencies. It explains the need for robustness checks in performance analysis of such strategies and the various methods by which Scientific Beta improves robustness. It also summarises the measurement protocol that we employ internally to assess the robustness of smart beta strategies. The paper moves on to analyse the design of a set of competitors’ multi-factor strategies, focussing on the selection of factors, factor definitions and methodological choices. 

This paper discusses why robustness is essential for investors utilising smart beta strategies and describes the sources of deficiencies. It explains the need for robustness checks in performance analysis of such strategies and the various methods by which Scientific Beta improves robustness. It also summarises the measurement protocol that we employ internally to assess the robustness of smart beta strategies.

This toolkit of tests is especially useful to investors in their evaluation of smart beta strategies or their internal portfolios. It is based on an analysis of factor exposures and evaluates the quality of factor diversification of the strategy or portfolio under scrutiny. This is particularly important as only strong factor deconcentration can guarantee the robustness of factor strategies’ outperformance over the long-term, since such strategies will benefit from factor decorrelation and their long-term reward. It also enables an analysis of conditional performance in a multi-dimensional context (market, volatility, sector, factors, and macroeconomic variables). It also draws on an evaluation of robust statistical inference and on out-of-sample robustness tests of the performance and risk of strategies using long-term data.

The paper moves on to analyse the design of a set of competitors’ multi-factor strategies, focussing on the selection of factors, factor definitions and methodological choices. We contrast this with implementation decisions taken by Scientific Beta to promote robustness in the index construction. Finally, our analysis extends to the application of the robustness protocol measurements to this set of competitor multifactor strategies. The results allow us to identify the issue of poor factor diversification and factor conditionality they typically display. By contrast, Scientific Beta’s multi-factor indices benefit from good factor diversification, increasing confidence in the expected out-of-sample outperformance.