Monday, April 4, 2011

The Best Hedges

If you've come to this post hoping for a lawn trimming pictorial then I'm sorry if I mislead you.

Often when building a portfolio, investors look to 'hedge' their investments. Wikipedia defines a hedge (finance) as, '...a position established in one market in an attempt to offset exposure to price changes or fluctuations in some opposite position with the goal of minimizing one's exposure to unwanted risk.' The tools of quantitative analysis give us a powerful and simple way to quantify the effective of a hedge. The correlation between the price movements of two stocks tells us how similar the changes in price are. I've already covered the basics of correlations in a previous post, so I won't do it again here. The idea is though, that if we want the stocks to be ideal hedges of each other, we want their price movements to be anti-correlated (correlation coefficient = -1). In short,  if the price of one goes up one day the other should go down, and vise versa.

So today I'll provide you with the top 3 quantitative hedges (Nasdaq only) of some of the most widely held equities in three sectors. Then I'll do something interesting and try to see if these hedges make sense, or if they're some kind of funny coincidence.
First up are the 3 best hedges against Google (GOOG),

Symbol               Correlation
SCIL                    -0.8347                - Educational software

ALXA                  -0.8158                - Pharmaceutical company
SXCI                   -0.7897                - Pharmaceutical and healthcare IT

Next up are the 3 best hedges against Qualcomm (QCOM),

Symbol               Correlation
THOR                   -0.8936                - Medical device electronics

AMAG                -0.8336                 - Pharmaceutical company
STRA                  -0.7967                 - Post secondary education services


Finally are the 3 best hedges against Amgen (AMGN),

Symbol               Correlation
CRUS                   -0.4881               - Electronics
WINA                  -0.4842                - Value retail
VASC                  -0.4058                - Medical Devices

So here's the thing...I see no rhyme or reason for any of these ideal hedges. Perhaps they are merely coincidental hedges or outliers of statistical randomness. Or perhaps they reflect some underlying trading activity. What I hope you get from this post is an understanding that a diversified portfolio can come from non-traditional combinations of equities. Perhaps Jim Cramer is a bit too presumptions in his, 'Are you diversified?' part of the show. Either way, once again, an (anti)correlation analysis is a powerful mathematic tool that you can add to your stock market toolkit.

*There are more complicated ways to compute hedges and I will build on those in a later post. Namely, we don't simply want our portfolio to balance itself out and never profit, but we want to minimize fluctuation/risk while maximizing return (see: Sharpe Ratio). Stay tuned...


  1. Correlations like this are pretty silly without detrending the data in some way/at some frequency. You'll get surprisingly high negative correlations for an upward market from a stock that's falling, and the overall market does on average rise...

  2. I'll agree with that...see my last paragraph.
    But it's a lot to introduce at once and is more instructive to build on the basics. Also look at two of the stocks I real trend, the prices are more or less stationary.

  3. It might be useful to give time periods for the correlation (is this daily, intraday, weekly, etc?). By reporting all three you'll be able to see larger variations in the correlation and be able to better discern which hedges are best