The technical support that "everyone" was watching in the S&P 500 was 1050, which coincided with the June closing low and a break of the Head & Shoulders pattern.

That pattern has gotten as much media attention as the "death cross" that we looked at last week.  It's probably just as (in)effective as a predictor...but let's actually test it instead of relying on textbook "shoulds" and vague opinions.

Some interpretations of the pattern differ, but if we're going to test this thing then we need hard-and-fast rules.  So what we're looking for is any time the market rallies, then falls back (creating the left shoulder).  Then it rallies even higher (creating the head) before falling back again to a level near the first pullback.  Then it must rally once more to a point that's relatively close in price to the left shoulder, and finally it must break the trendline that joins the trough between the two shoulders.

Once that occurs, we would sell short on a violation of that trendline.  The profit target is the difference between the head and the trendline, while the stop loss would trigger if prices rally back above the head.

If we test the Head & Shoulders pattern on the 500 stocks in the S&P 500 since 1995 by selling short when the pattern triggered, we get a total of 383 trades.

Out of that sample, only 27% of them were winners.  That means that almost 75% of them did not decline enough to meet the profit target.  Overall, the average return from these trades was -1.2%.  Not good.

For those trades that did work, it took a median of 26 trading days to hit the profit target.  For those that failed, it took a median of 58 days to get stopped out.

For the stocks in the Nasdaq 100, the win rate was 50% with an average return of -3.3%.  The higher winning percentage is likely due to higher volatility, and thus higher likelihood of hitting the price target, even though the 50/50 winning percentage wasn't enough to make the strategy even remotely profitable.  It took a median of 43 days to get stopped out, and a median of 24 days to hit the profit target.

There is some art in defining the Head & Shoulders, which makes it difficult to test, and looking through some of the trades, I personally wouldn't have considered them valid.  But the test gives us a general feeling for how successful the pattern has been across a wide swatch of stocks, without relying on a handful of cherry-picked examples.

 Disclaimer  |  Privacy Policy