What To Do About Increased Correlation

There has been a report floating around from State Street that included the following interesting nugget about correlation;

When returns (utilizing US and non-US equity markets) are greater than one standard deviation above the mean, the correlation is 17%. However, if the returns drop one standard deviation below the mean, correlation is 76%. With such a high correlation, investments are likely to fall in a collapsed market.

Actually that quote is from a post by Wall Street Post Game, there was also a very similar article at All About Alpha that was posted a couple of days earlier on this as well. Essentially someone crunched some numbers around the generalization that was so popular in 2008 that correlations all went to one. While not necessarily shocking I believe the numbers support two points I have tried to make in the past and implement in client portfolios.

First is the top down notion that in a down market it is better to get out, more practically (whipsaw, commission drag, tax consequences) reduce net long exposure, than try to find the few stocks that might go up in a down 30% world. This is of course a building block of top down management. If conditions favor a downtrend (like after a breach of the 200 DMA) the portfolio is better off being defensively positioned or otherwise hedged.

The other point that I think is supported is narrow based investing. A point made here repeatedly was that select countries would not avoid going down a lot but that they would do so on a different time table and thanks to better fundamentals would recover sooner or at the very least have done better than to be down 31% from the peak as the S&P 500 is now. Brazil and Norway kept going up for another seven months after the SPX' peak in 2007 and Chile bottomed out with a 30% drop versus 56% for the S&P 500. Compared to that 31% drop for the SPX mentioned above Chile is up 30%, Brazil up 8% and Norway down 18%. Interestingly Australia, another favorite investment destination appears to be down a hair more than the SPX in that time.

You should be able to get more on country returns for varying time periods from Bespoke Investment Group. I would submit that the decade numbers are more important than the three year numbers. While it is definitely preferable to be down less than the 31% of the SPX for the last three years the more important number is where you stand over a longer period of time and even more important is whether you have enough when you need it. It is unlikely that a three year result will be the make or break for any retirement plan. In that context the point then becomes about trying to smooth out the ride as best as possible which means avoiding the full brunt of down a lot and capturing most of the up cycle.

The importance of luck should not be underestimated either. A true story from yesterday; we went on our normal Sunday morning six-mile hike yesterday morning with a good friend and four of our dogs. When we go on this hike we let Roscoe go without a leash.

Yesterday as we were almost done, maybe a half mile from the car, we saw a woman standing on the trail with her mountain bike and she yelled "there is a rattlesnake (not pictured) here, you might want to leash your dog." As she was saying this Roscoe ran to her anyway and she grabbed him by the collar and I leashed him up. Just as this was happening the snake, about 10 feet off the trail under a manzanita tree, made some noise. We rarely see anyone (or any snakes for that matter) on this trail and that there was someone right by a rattlesnake that Roscoe would have unquestionably gotten too close to was an incredible stroke of luck.