What’s hot: a report from our men at the Joint Spring Seminar with Inquire UK at the Coombe Abbey Hotel in Warwickshire, United Kingdom
Inquire Europe and Inquire UK are two sister, cooperative, non-profit professional organisations dedicated to developing quantitative research in investment management. A joint conference is organised annually bringing together practitioners and academics. The 2015 joint spring conference took place near Coventry at the spooky Abbey Coombe hotel.
For me the conference started on Sunday 15th March afternoon with the programme committee that I joined just recently and which is chaired by Hélène Harasty from Lombard Odier. We discussed the programme of future conferences including this year’s autumn conference to be held in early October in Athens, which will mark the 25th anniversary of Inquire Europe.
This year’s spring conference was attended by some 110 delegates from sponsoring institutions and included representatives with strong quantitative finance backgrounds from a number of leading asset owners and asset managers, mostly from Europe and UK, but also from the U.S. and from the Middle East.
The theme of the conference was “1/N and other ‘Smart’ Portfolio Construction Approaches”. Smart Beta was thus often in the limelight. In my view the most striking presentation was made by Raman Uppal from the EDHEC Business School, who presented a paper showing that the equally-weighted portfolio applied to a stock universe can be fully explained by market, small-cap and value factor exposures, much in line with our own research on this topic. What he added is that the frequency of rebalancing the equally-weighted portfolio adds an exposure to the reversal factor that increases as the frequency of rebalancing increases. So, if the equally-weighted portfolio performs so well it is not because of diversification. It is simply because it is exposed to factors that have paid in the past, something we have been sayingg for a number of years. Despite that, Bernd Hanke from Global Systematic Investors presented an approach based on combining equally weighted sector portfolios with equally weighted stocks within sectors without referring to the factor exposures inherent to the strategy.
Felix Goltz from the EDHEC-Risk Institute re-emphasised this showing that popular smart beta strategies actually bet the farm on a few factors making performance heavily dependent on market conditions. He defended multi-factor strategies as a more robust approach in particular when using different approaches to build the factors as a way for diversifying against parameter estimation errors, data mining risk and to as a way to diversify across cycles.
Frank Siu from Axioma delivered another enlightening presentation which again corroborates well our own research. He discussed the problem that, despite the raise in Smart Beta ETFs, when it comes to Smart Beta there is no one-size fits all. Focussing on the Minimum Variance strategy he showed that investors of different sizes require different forms of the strategy that should take into account capacity constraints. Indeed, the market capitalisation index has the largest capacity. All other portfolio strategies have necessarily lower capacity. This is something we have also emphasised in our research and which presents a difficulty for index providers and for the managers replicating those index strategies, something Frank recognised.
Altaf Kassam from MSCI alluded to this same capacity problem. He highlighted the fact that MSCI has now created high capacity factor indices as a complement to the standard factor indices, (henceforth to be known as high exposure factor indices), they had already introducede. The new high capacity indices typically invest in all stocks and take smaller active risk than their lower capacity standard equivalents.
Peter Gunthorp from FTSE gave an overview of what is today available in index format and what index providers are working on these days, in particular new factor indices, leveraged and long-short factors indices, factor rotation approaches and multi-asset class consistent approaches. He also noted that fixed income continues to offer a bigger challenge when it comes to Smart Beta mainly due to the price discovery mechanisms and the liquidity issues with fixed income securities, also raising implementation challenges. Interestingly, Claudio Ferrarese from Fidelity Worldwide Investment introduced a Smart Beta strategy for global high yield based on equal-risk contribution at country level. The equal-risk contribution strategy was also the focus of attention for Thierry Roncalli from Lyxor who presented a framework for introducing expected returns in risk parity strategies.
Finally, on multi-factor combination, Stephen Brown from NYU Stern, in his general presentation of the historical context of Smart Beta highlighted some of the dangers of too much focus on maximising the Sharpe ratio and the possible unreasonable outcomes that may arise from it. Interestingly he defended Bayesian approaches as a better way forward, something that again falls in line with our own research and the framework we developed from multi-factor combination. The topic was also discussed by Semyon Malamud from the Swiss Finance Institute who presented a Black-Litterman approach with options and transaction costs.
Raul Leote de Carvalho, co-Head of Financial Engineering team, BNPP IP, Paris, France