What are the implications of rapid global adoption of factor (smart beta) investing in single-factor, multi-factor and dynamic multi-factor strategies, most notably via equity exchange-traded funds (ETF). In their September 2018 paper entitled “Smart-Beta Herding and Its Economic Risks: Riding the Dragon?”, Eduard Krkoska and Klaus Schenk-Hoppé summarize the current state of smart beta investing, providing a concise overview of academic research, investment community reports and financial media coverage. They address evidence and implications of investor herding into smart beta vehicles. Based on the body of research and experience, they conclude that:
- Smart beta strategies offer some of the return benefits of active strategies with some of the implementation advantages of passive strategies.
- Most investment professionals who are experts on smart beta:
- Accept traditional factor strategies (value, size, momentum and perhaps a few others) as adequately argued and tested, with size relatively out of favor.
- Reserve judgment on more exotic factors.
- Implementation details are very important to smart beta strategy success.
- Investment professionals report that multi-factor strategies offer considerable performance improvement compared to single-factor approaches, but flawed implementation can results in factors cancelling or duplicating each other.
- Both behavioral (exploitation of irrationality) and risk-based (acceptance of some risk) forces probably drive smart beta returns. The former is persistent but subject to bubbles/crashes. The latter is subject to weakening via market adaptation (arbitrage).
- Offerors and financial media often treat smart beta strategies as exciting innovations, attracting investor attention/money and arguably inducing irrational herding into these strategies.
- Dynamic multi-factor strategies that assess factor valuations may support exploitation of such herding.
- The capacity of the market to absorb smart beta flows, especially those involving short sales, may limit herding.
- There are concerns that smart beta is attractive largely due to data snooping that will not carry over to live performance.
In summary, evidence suggests that investors are herding toward smart beta products, with growing probability of market reactions against this flow.
The paper includes many citations and summaries of findings from relevant research/reports.
Cautions regarding conclusions include:
- Much of the academic research on factor returns (premiums) ignores implementation costs and constraints. Some or many factors may not have net premiums.
- As covered in the paper, much research on factor/smart beta performance ignores the direct and inherited snooping bias derived from: (1) intensive reuse of the same source data; and, (2) frequent tweaking of past research. Some or many factor premiums may not survive correction for this bias.
For practical perspective, see:
