December 6th, 2012

More Reliable: Horizontal vs. Sloped Trendlines?

Over the past several years, charts have become more pervasive than ever in discussions, commentaries and prognostications about individual stocks and the market in general.  Even Cramer, who once considered technicians to be on the same level as astrologers or readers of tea leaves, no regularly refers to the analysis of one chartist or another.

One of my pet peeves, however, is that bloggers and media talking heads will insert trendlines in their discussions almost willy-nilly as they pontificate about the support or resistance they hope the line they drew will presumably offered them. Because the use of trendlines is so prevalent, it’s assumed that everyone understands their meaning and relevance; we rarely hear about the arbitrariness and subjectivity that goes into their selection.

Last week, I offered three examples of head-and-shoulders patterns (GLD, AAPL and FDX), each demarcated by a horizontal trendline, or the pattern’s “neckline”.  [As an interesting aside, I may have been one of the first to publish an alert about the possibility of AAPL forming a head-and-shoulder top which could result in a correction down to approximately 400 in my November 8 post, "AAPL Gets a Cold, the Market Gets …..?"  Now many are commenting about it and Cramer even had a segment tonight dismissing the stocks technical risks.] What makes the head-and-shoulder such a “reliable” (if you allow me to use that term in the context of something so subjective as the reading of stock charts) pattern is that the supporting trendline is horizontal.  I’ve seen sloping necklines but these never turn out to be as recognizable nor as accurate.

As I describe in my book, Run with the Herd,

What makes trendlines so confusing is that many trendlines that seemed so precise at first may lose their potency as new trading is tacked on.  As a matter of fact, as more transaction data over longer and longer periods of time with multiple trading days condensed into individual bars, you’ll usually find yourself drawing a plethora of trendlines.  Some trendlines are short and some long, some connect pivot points that once seemed compelling and inviolate become less significant and even irrelevant when viewed in a longer-term.  The support or resistance expectations implicit in short trendlines at one may become overwhelmed and irrelevant as more recent buying and selling emerges.

Trendlines are nothing more than an arbitrary, imaginary lines that visually connect two or more pivot points. Pivot points are those spaces in time and price where control is transfers between buyers and sellers, when one trend in one direction reverses and begins moving in the opposite direction. In reality, this transfer doesn’t occur in one transaction at one price but instead occur over a period of time, a large number of trades at a range of prices.  There’s no precise way of locating when that transfer is complete and the struggle continue even when a reversal appears to be complete.

Why is it important to locate these pivots? Because we believe that after having occurred several times at approximately the same point, the failure to occur at that same level sometime in the future means that the winner of the last battle has lost control of the trend and it now resides in the other side who will control the trend until the next struggle begins.  That transfer of control is the breakout.

Repetitive struggles at the same price make intuitive sense.  An institutional investor looking to sell its large holding in a stock will continue to do so as long as a stock’s price is above a certain level; when it drops to or below the level they hold their shares back from the market; they will continue to accumulate shares up to a certain price but not above that price.  But what can we say the same thing about pivot points at different levels?  Bottom line, they tell us little about what we can expect about where the next pivot might be and we can say little about whether that recent pivot is the beginning of a reversal or the continuation of a trend.

The above chart for LKQ presents a pretty channel but it offers little information about the risks of the trend failing, how much profit potential remains in the channel trend or when it might collapse.  It is easy to draw the channel trendlines after the fact but drawing those lines in 2010 would have produced the dotted trendlines.  Which more accurately defines the trend, the solid or dotted lines.  Is the stock currently within the trend boundary or is it outside the bounds and bound to correct.  The most common mistake when inserting trendlines is thinking that the recent pivot is critical in establishing a meaningful trendline.  In other words, trendlines are usually discovered within the time frames of the chart, rarely coming in from prior the chart’s beginning.  That’s why I always simultaneously look at a charts in three time horizons.

That error doesn’t occur when you look for breakouts across horizontal trendlines like this one for NEOG.  Is there any doubt that a cross above 48.00 indicates that the bears have lost control to the bulls who have launched a new push to higher stock prices?

Technorati Tags: