djx Breakouts on the Dow - A Longer Term Perspective

Breakouts on the Dow - A Longer Term Perspective

Although breakouts have been very profitable during the most recent bull market, they have been less useful in bear markets. From 1900 through 1989, the performance of the breakout was not as promising as it has been over the past decade although most of this is because the breakout was a lousy indicator during a bear market (1968-1983). I used only breakouts from a 6 week base (look_bak = 6, as defined in the article), then measured how the market performed over the 4 weeks (close-close) following each breakout.

Here is a breakdown of how the Dow performed from 1900-1989 for the 4 weeks after a 6 week breakout (close - close[t+4] of a week in which the high of this week exceeded the last 6 week high):

	Average 4 week period:*	Average 4 week period after a 6 week breakout:
Maximum:	57.5%	42.4%
Minimum:	-37.2%	-22.8%
Median:	0.8%	1.2%
Average:	0.49%	0.54%
Total:	4678	241

* close-close percentage change (dividends ignored) for all overlapping 4 week periods in the study sample.

The markets may have changed fundamentally from the turn of the century to now (especially since 1982 with the advent of the personal computer) so take this into account before generalizing from these findings. Nevertheless, it is noteworthy that although the average 4 week period following a 6 week breakout as defined in the article was superior to the average 4 week period in the study sample (1.2% v. 0.8%), the difference is not striking, and when you look at the median 4 week returns (0.54% v. 0.49%) it almost disappears. What is encouraging, however, is that the lowest post-breakout 4 week drawdown was -22.8% v. -37.2% for the average 4 week period.

The reason these numbers are so discouraging (in terms of highlighting the usefulness of a breakout as a bullish indicator) is that they include several protracted bear markets. During these bear markets, the breakout indicator did not perform well; in fact, if you only bought the Dow following a 6 week breakout in the period from 1968 to the beginning of 1983, you would have lost money during the average subsequent 4 week period:

	Average 4 week period, 12/1/1968-February, 1983:	Average 4 week period after a 6 week breakout, 12/1/1968 - 2/1983:
Max:	17.2%	11.7%
Min:	-16.2%	-10.4%
Median:	0.3%	-1.2%
Average:	0.18%	-0.60%
Total:	744	42

Note that the average and median 4 week returns following a breakout were not only less than average (-1.2% v. 0.3% median and -0.60% v. 0.18% average) but were negative. There are several ways to interpret this information. First, it underscores the importance of determining what kind of market we are in before selecting an investing or trading strategy. Momentum has a poor track record in bear markets. Second, once you determine that we are in a bear market or a trading range, you might fade breakouts, selling the market short following these periods. Third, you might shift your emphasis to oscillating indicators during this period, attempting to catch price swings instead of major upward price moves.

As in the 1990's, the 1980's (after 1982) saw a relatively uninterrupted bull market in the Dow Jones Industrial Average. Therefore, we would expect that the breakout indicator would be helpful, and sure enough it is:

	Average 4 week return, 1/1983-10/1989:	Average 4 week return following a 6 week breakout, 1/1983-10/1989:
Max:	12.1%	6.4%
Min:	-31.0%	-5.5%
Median:	1.6%	3.2%
Average:	1.19%	1.83%
Total:	356	19

Although this was a very bullish period with an average 1.19% 4 week return (excluding dividends), the average 4 week period following a 6 week breakout was even better with a 1.83% average 4 week return. The median figures were also solidly in favor of the breakout: 3.2% v. 1.6% subsequent 4 week return. Note also that the biggest loss following a 6 week breakout was only -5.5%, far less than the -31.0% maximum 4 week loss during this study period.

Summary

Breakouts perform as advertised on the Dow Jones Industrial Average when the market is in a bull market, as has been the case for most of the past 2 decades. Longer term, however, the breakout - at least from a 6 week base - is less consistent and is indeed poor during bear markets. Determining what type of market environment we are in is important before deciding how to interpret a breakout.