Been meaning to write this post for a while, but it seems that my patience has paid off as I have even more chart evidence to work with. And, that is the main point of this post. Technical analysis is not a means for predicting the future. That may seem paradoxical to those who have seen me use technical analysis to attempt to determine buy and sell points. If technical analysis isn't for predicting the future, then how can it be useful?
What I mean by technical analysis isn't meant for predicting is that rigorous technical analysis isn't a means for front running what you "think" may happen to a stock. It's a means for gleaning information from an existing chart about an existing trend and helping the practitioner to make investments with underlying market supply and demand behind them.
I came across this post, "Are Sears Holdings shares headed for a correction?," at wallstnation.com. The thrust of their argument was that SHLD's shares had traced a head and shoulders pattern and that a significant correction was likely.
Here's a "text book" head and shoulders pattern:
- The stock rises to a peak and then declines.
- The stock rises to a second higher peak and declines, but does not breach support defined by the first decline.
- The stock rises to another peak but is unable to recapture the second peak.
- The stock breaches the support level formed by the first decline and confirms a reversal of trend
Wall St. Nation posted this chart of SHLD as of 9/30/09:
Wall St. Nation's contention is that a head-and-shoulders pattern is defined by the three peaks it labels. Wall St. Nation claims that this formation is "one of the most reliable trend-reversal patterns." What they miss is the fact that simply the formation of three "head-and-shoulders" style peaks is not the predictor of a trend reversal. The head and shoulders pattern must include a breach at the end of the pattern formation. That is step 4 listed above. The head and shoulders pattern is a trend reversal not just a "reliable pattern" for predicting a trend reversal. Once the full pattern is formed, one can typically expect a continuation of the downtrend after initial support is breached. In the case of SHLD, baseline support was never breached. In Wall St. Nation's own example, lows never breached the initial low in the formation and in fact have maintained an upward bias that they themselves identify in the chart.
In fact, in the chart they show, it is just as likely that a long term uptrend remains well intact and that the chart is merely in-between a series two "higher highs.' Well, a month later, what has happened?
Well, the stock ripped off an upward run and made a high beyond the "right shoulder" of the supposed head-and-shoulders pattern. At this point, it's hard to identify any near term trend at all. Generally speaking, it seems the 50-day moving average has provided rough support, MACD shows a bullish bias, and a lower low has not been seen all year.
In the case of the 3-year, weekly price chart, SHLD has just this year managed to return above its 50-week moving average which I believe signals a potential long term trend reversal.
