Tuesday, August 04, 2009

Panic Rarely Pays

I was noticing today that anyone who sold the stock market late in 2008 and early in 2009 when thousands of stocks were making fresh lows for the year would now be losing significant money if they had held their short positions.

Selling into panicky markets may profit short-term from downside momentum, but overall it's a losing strategy at any time frame.

When markets get panicky, pretty much all the bears have gone all in. When there isn't further selling pressure to push markets lower, those bears have to cover their positions. That can lead to vicious rallies.

I went back to September, 2002, when I first began collecting new 20-day high and low data, and examined all occasions in which new 20-day lows exceeded 1500 across the NYSE, NASDAQ, and ASE. Five days later, the S&P 500 Index averaged a gain of .43% (107 up, 90 down). That compares with an average five-day gain of .02% for the remainder of the sample (841 up, 688 down).

Interestingly, it's not that markets are more likely to go up following extreme weakness; rather, it's that their gains tend to be much larger than their losses.

Because there are indeed fat tails of negative returns in which weak results can lead to much further weakness, the best execution of trades when weakness hits is to wait for evidence that selling is drying up--i.e., one of those transition patterns I've written about.

Good trading doesn't mean automatically following the herd or fading it. Rather, it means standing above the herd and recognizing when the herd is likely to head off a cliff.

More on transition patterns here and here.