I've been playing around with some factor analysis lately and stumbled across something interesting.

But first, a brief overview of the methodology I utilized.  I constructed long-only factor portfolios from the top 50% most liquid S&P 500 constituents (defined by trailing 3-month volume; indices are survivorship-bias free using a point-in-time database for relevant factor data like book-to-market).  Factor portfolio construction is defined as follows:

  • Momentum: Securities are ranked by 12-1 month trailing total returns (split & dividend adjusted); securities from the 60th - 90th percentile are selected and weighted by split-adjusted market capitalization (capped at 5% total weight)
  • Size: Securities are ranked by market capitalization; securities from 0-20th percentile are selected and weighted by split-adjusted market capitalization (capped at 5% total weight)
  • Value: Securities are ranked by book-to-market; securities from 0-20th percentile are selected and their weights are a linear interpolation from double weighted to half-weighted based on an initial equal weight scheme (e.g. deepest value gets 2x equal-weight, most over-valued gets 1/2x equal-weight)
  • Low Volatility: Securities are ranked by trailing 12-month volatility; securities from 0-20th percentile are selected and their weights are a linear interpolation from double weighted to half-weighted based on an initial equal weight scheme (e.g. deepest value gets 2x equal-weight, most over-valued gets 1/2x equal-weight)

Securities are selected and held for a quarter.  To prevent any date selection bias, for each factor, three different indices are constructed:

  • An index that rebalances January, April, July, and October
  • An index that rebalances February, May, August, and November
  • An index that rebalances March, June, September, and December

Each portfolio is initially given 1/3rd of the portfolio.  When one portfolio has its quarterly turnover, the whole portfolio is re-normalized so that each index is reset to 1/3rd of the portfolio.

Enough with the details!

What was the interesting find?  It seems that, relative to the S&P 500, the "premia" in "risk premia" was not realized for investors in the last several years.  It isn't unusual for a given factor to go through a period of under-performance (likely because of "risks" associated with this factor are realized) or an extended period of neutral performance.  What is unusual about this is that all of the common factors exhibit neutral performance over the period.  I highlight the period in the image below.

Great Factor DepressionIs this simply a case of factor specific strategies becoming more common investment tools (see the proliferation of factor-based ETPs)?  Or is it simply that in a risk-on/risk-off or QE/no-QE market, the premia are overwhelmed by larger market factors?  Perhaps there is a simpler explanation...

Corey is co-founder and Chief Investment Officer of Newfound Research, a quantitative asset manager offering a suite of separately managed accounts and mutual funds. At Newfound, Corey is responsible for portfolio management, investment research, strategy development, and communication of the firm's views to clients.

Prior to offering asset management services, Newfound licensed research from the quantitative investment models developed by Corey. At peak, this research helped steer the tactical allocation decisions for upwards of $10bn.

Corey is a frequent speaker on industry panels and contributes to ETF.com, ETF Trends, and Forbes.com’s Great Speculations blog. He was named a 2014 ETF All Star by ETF.com.

Corey holds a Master of Science in Computational Finance from Carnegie Mellon University and a Bachelor of Science in Computer Science, cum laude, from Cornell University.

You can connect with Corey on LinkedIn or Twitter.