Key Quant: New revelation of systematic management

This post is also available in: Français

“Key Quant.” The saga of Robert Baguenault Viéville and Raphael Gelrubin could be summarized in these two words, the name of the company they run. Specializing in systematic trend-following management strategies (trend following CTA); this young company is digging its wake in the world of fund management.

Founded in 2009 by two former quantitative research and risk management professional from Fidex Man Ltd, a subsidiary of the investment fund Man Group, KeyQuant manages nearly 80 million U.S. dollars with honorable performances from three years: +14.7% on average, with a peak at 44% in 2010, the first year. And even less successful in 2012 (-4%), the fund remains at quite enviable historical results.

At the origin of the emergence of these two entrepreneurs, an importance research work initiated in Man Fidex Ltd.. Their goal was to improve the risk profile of the systematic management model of the company, which sought to capture the beginnings of trends. After two years of intensive research, Robert and Raphael Baguenault Viéville Gelrubin make proposals for changes to Man, which the group denied. The new system would have a fairly high correlation with Man AHL, the flagship fund of Man Group. The two friends decided to resign, to create their structure and develop their own system. From the end of 2007, they sent their daily paper trading to Finaltis, which has real expertise in systematic quantitative management. In 2009, they take the leap and decided to create KeyQuant that gets its management approvals a few months later.

Years of research by Robert Baguenault de Viéville and Raphael Gelrubin led to the development of algorithms for detecting trends that go well beyond the traditional tools used in the systematic management today, as variations of moving averages and breaks, to anticipate and estimate trends.

Indeed, according to the two managers, “a lot of systems send binary signals: buyer, seller or neutral. The allocation of each asset is then derived from notions such as volatility risk, VaR, CVaR. “Them, they went on another approach:” The markets are continuous and there is not a day that goes from a market neutral to buyer, but instead, a trend will happen slowly, then confirm”. The intuition is that it is certainly possible to anticipate price trends much earlier even than do existing algorithms.

The two co-founders of KeyQuant have integrated in their tool the calculation of a predictive distribution function for each market, “to have state probabilities of increases and decreases, as well as normal and extreme risk assessments “. The other major difference between KeyQuant and the others, state Baguenault de Viéville and Gelrubin, is this indicator that they have developed for deciding on the allocation of portfolio risk in terms of quality trends.

Overall, KeyQuant has a significant correlation with the large international CTAs, Winton, Aspect or Blue Trend but their approach provides a decorrelation at key moments, allowing them to offer a higher return/volatility profile until day.

The results seem to live up to their expectations and assets under management are expected to continue to grow. KeyQuant has established a partnership with Finaltis, which since 2010 has delegated part of the management of their assets. The two managers also expect on forecasts of 3 to 5 billion U.S. dollars in management within three years, armed with their approach appreciated by institutional.

The adventure is just started for KeyQuant and only history will tell if these two young managers will be able to revolutionize innovative systematic management algorithms with their “new generation” of detecting trends. In June the company received the award for best new CTA at the ceremony of Managed Futures Pinnacle Awards held in Chicago by Barclay Hedge and CME. Bodes well …