European academic think tank – EDHEC Risk Institute (ERI) – has released a study suggesting that smart beta performance reports should be more robust and not from data mining and non-robust weighting methodologies.
The Robustness of Smart Beta Strategies publication is from ERI Scientific Beta, the index arm of ERI, and looks at the importance of robustness, discussing how best to measure and assess robustness when analysing the performance of smart beta strategies.
According to ERI Scientific Beta it is important to trust academic consensus as the number of smart beta indexes and factors are being launched. It says: “The good idea of factor investing should not be transformed into factor fishing and data mining.”
The report states that the “lack of relative robustness arises mainly from data mining and non-robust weighting methodologies, while the lack of absolute robustness comes from undiversified factor exposures.”
The study explains that variables such as sales, dividends, book value and cash flow are used as risk factors by many fundamental factor-based funds and indexes with the idea that fundamental indexes could be high-performance smart proxies for the Value factor.
This highlights the importance of measuring robustness correctly. Measuring robustness correctly relies on the transparency of track records and the availability of instruments to actually measure it, such as the probability of outperformance.
“The probability of outperformance enables the smart beta index’s capacity to outperform to be measured for a chosen investment duration, whatever the investment period,” ERI said in a note.
The study finds that relative robustness can be improved by reducing all sources of unrewarded risks with the use of a consistent framework (to prohibit data mining), robust parameter estimation techniques, weight constraints and strategy specific risk. Absolute robustness can be achieved through allocating across several rewarded factors.
Their results show that the single factor indexes have a high degree of relative robustness, but they are not robust in absolute terms. The multi-beta allocations, on the other hand, are highly robust in absolute terms.
EDHEC Risk has been a controversial voice in the indexing space arguing for free data within the indexing sector. It has prompted several commercial index providers to speak out against the sentiment, arguing that indexing is a commercial business and that there are high costs associated with maintaining and running an index.
Most recently EDHEC teamed up with Amundi in a strategic partnership, which combined ERI’s smart beta platform, ERI Scientific Beta, and Amundi’s expertise in ETFs.
In February EDHEC launched its multi strategy smart factor indexes, which provides risk strategies in developed countries.