Prevailing stock correlation favours quants over stockpickers, says Axioma
The inter-correlation of FTSE Europe index securities is at its second highest point for the past decade, creating a “shockingly unhelpful environment” for stock pickers, but possibly the dawn of a good era for quants and hedge funds, say risk and portfolio optimisation software providers Axioma.
Since August the average asset correlation of the benchmark has been 46%.
This is a sharp jump from the 31.8% average in the preceding five years, and a world away from the 17.5% average correlation registered over the five years to August 2005.
The correlation is at its second highest point since September 2000, and lower than only one peak it reached last year.
Olivier d’Assier, managing director at Axioma, said: “Correlation is a problem for stock pickers because if all the assets are moving together, you cannot pick winners from losers, nor differentiate yourself as a good stock picker from a bad one.
“There is not a single industry that is not positively correlated to another, whereas 10 years ago a lot were, especially during and after the dot.com bubble.”
Reflecting the flipside of the high inter-correlation statistics, the percentage of FTSE Europe stocks with negative correlations to other benchmark constituents has fallen from an average of 17% from 2000 to 2005, to just 2.2% now.
This makes it almost impossible, and very costly, to re-risk a portfolio via diversification by holding more assets, d@assier said.
And the ratio of positively correlated to negatively correlated securities has jumped from 1.03 in 2000-2005, and 5.39 between 2006-2011, to 20.91 since August.
High correlation is often a hallmark of falling markets, and d’Assier said the 60-day market return turned negative when the average asset allocation breached 35%. It is now 46%.
The odds now are therefore weighed firmly against stockpickers selecting rising shares, and in favour of them selecting falling equities instead.
D’Assier said, however, that high prevailing correlations were helpful to quant and hedge funds. “The good news about high correlation is that hedging works even better.”
Quants and hedge funds had a golden era from 2003 to mid-2007, before giving way to fundamental stock pickers. He said quants and hedge managers could now enjoy another “three to four years” as the ascendant style, “if this new environment of dominating systematic sources over stock selection becomes the new norm.
“Since March quants have been able to identify factors with stable relationships again, and they’re outperforming. These style [shifts] last three to four years on average.”
Factors fundamental managers typically use – growth and value, leverage, and liquidity, for example – have driven meagre returns this year, according to Axioma’s elegant analysis, with only growth making gains. Even then, the gains were less than 5%.
The factors of medium- and short-term momentum used by quant and hedge communities, by contrast, have formed stronger, stable trends. They have generated, respectively, a gain of nearly 5%, and loss of over 5%. Riding volatility as a factor meant a loss of over 15% so far this year – or a healthy gain from shorting the trend.
D’Assier said, given the change in climate, many fundamental managers were hiring quant teams, becoming hybrid businesses running “quantamental” approaches.
“Quants have the more disciplined portfolio construction process, using risk models to eliminate the risks they do not want. These teams can build overlays for fundamental managers, and work together with them to build high conviction portfolios designed to capture the stock picking skill of the fundamental manager, but hedged quantitatively for any other source of unwanted risk.
“You have the discipline of the portfolio construction process from quants, and fundamental stocks selection from an experienced manager.”