Bulkowski’s Study: Long versus Short

Written by and copyright © 2005-2008 by Thomas N. Bulkowski. All rights reserved.

A question I am asked frequently is why I do not recommend shorting a stock. The answer is simple. You can make more money from the long side than the short side, even in a bear market.

For proof, I used 500 stocks from 1991 to 1996 and another 739 from 1995 to 2005 and found 12, 385 chart patterns. I measured the rise from the breakout price to the ultimate high, which is the highest high before price drops by at least 20%. For downward breakouts, I measured the decline from the breakout price to the ultimate low, which is the lowest low before price rebounds by at least 20%.

Average Rise or Decline

The following table shows the average rise or decline of perfectly traded chart patterns without any fees deducted. I split the yearly price range into thirds and measured the performance of the 12,385 chart patterns, placing each in one of the three categories according to the breakout price. Comparing the long versus short numbers, going long in both bull and bear markets beats going short in either a bull or bear market.

For example, those chart patterns with breakouts within a third of the yearly low showed gains of 38% in a bull market, but shorting in a bull market produced a gain of just 19%. In a bear market, going long made an average of 32% but going short made just 26%.

Market condition, Trade direction	Lowest Third	Middle Third	Highest Third
Bull market long	38%	35%	36%
Bull market short	19%	17%	16%
Bear market long	32%	27%	23%
Bear market short	26%	23%	22%

Failure Rates

The following table shows failure rates over the same three yearly trading ranges. The easiest way to explain the failure rate is to cite examples. Using the lowest third column (those chart patterns with breakouts within a third of the yearly low), 4% of the chart patterns I looked at failed to rise at least 5%, and double that -- or 8% -- failed to drop at least 5%. The next two rows down: 12% failed to rise at least 10% and 24% failed to drop at least 10%. In all cases, shorting a stock after a chart pattern breakout means an increased risk of failing than if you were to go long.

Failure Rate	Lowest Third	Middle Third	Highest Third
5% long	4%	5%	7%
5% short	8%	9%	15%
10% long	12%	17%	18%
10% short	24%	28%	34%
15% long	22%	29%	29%
15% short	39%	46%	51%
20% long	32%	39%	38%
20% short	53%	63%	66%
25% long	41%	48%	47%
25% short	65%	74%	78%