Lower intrastock correlations favor alpha strategies

Columbia Management, Investment Team | May 14, 2014

From the Columbia Management Quantitative Strategies Equity Research Group

Chart: Trailing 12-month S&P 500 Index stock correlation and quarterly aggregate alpha

10-year timeline: Four periods of correlations and alpha 2003–2013

A: The pre-crisis period — Prior to the financial crisis, intrastock correlations were the lowest in the last 10 years, and for most of this time alpha was positive.

B: The financial crisis — Unsurprisingly, alpha was worst during the 2008–2009 crisis. Surprisingly, given the breadth of the equity rally, alpha was highest in the 2009 rebound that followed.

C: Post-crisis uncertain recovery — Macroeconomic worry dominated the post-crisis environment, overwhelming the virtues of individual stocks, thwarting stock-pickers.

D: Post-crisis stability — As intrastock correlations fell to their lowest level post crisis, alpha is once again mostly positive, rewarding active managers.



Quarterly aggregate alpha: Fund performance, AUM and style classification data is from Morningstar and FactSet. Only funds classified in the Morningstar nine style boxes were included, meaning many specialty funds were excluded. To compute the gross excess returns (alpha), the corresponding Russell benchmarks for the asset classes were used. Both inactive and active funds were included in the calculation. Index funds and enhanced index funds were excluded. Aggregate alpha for the Morningstar funds was calculated using the square root of the AUM-weighted mean, as this lessens the influence of very large funds on the result without completely discounting size.

Trailing 12-month stock correlation: The stock correlation is the average rolling 12-month daily pairwise S&P 500 Index stock correlations. The S&P 500 Index tracks the performance of 500 widely held, large-capitalization U.S. stocks. It is not possible to invest directly in an index.

The illustrations here are not intended to be representative of the performance of any particular investment. Such information analysis has inherent limitations and may not be indicative of future results. It is important to keep in mind that no formula, model or tool can in and of itself be used to determine which securities to buy or sell, or when to buy or sell them.