Nexus of Supply and Demand
You have probably noticed that I draw trendlines in nearly all of the charts I show you in this newsletter. Trendlines represent a nexus of supply and demand. In an uptrend, the trendline acts as support. Below the trendline, you should expect that sellers will become more aggressive than buyers. However, as long as share prices remain above the trendline, buyers are in control. In a downtrend, as long as prices remain below the trendline, sellers are in control. Above the trendline, the back of those whom have been acting as suppliers (the sellers) is broken, and buyers are likely to become more aggressive.

See the chart below for an example of what I am talking about. If the S&P 500 closes below the blue trendline, then sellers are likely to quickly pile on to any downward price action and overwhelm the buyers.

When you look at a chart, your eye, even without the line drawn, is likely to "see" an area below which (in the case of an uptrend) the trend has probably broken. That would be a place where price losses are too large to ignore as mere daily fluctuations. A trendline more accurately depicts where the market's behavior may change (I am using "market" here as a general term, but I could use the words "stock", "bond", "ETF", "currency" -- or any traded asset or liability, for that matter -- in its place.)

What about regression lines?
Why not determine the trend by computing a regression (a statistical method for determining the average price change over a period of time)? The regression will give you a nice straight line that will also approximate the price action for the time period over which it was computed. However, that regression changes every day, as prices change. Furthermore, by definition, prices will fluctuate around the regression line in the past, but might not continue to do so in the future, so you really will not know when the trend has changed by simply looking at regression lines.

Why draw trendlines?
The point of drawing a trendline is to help you find a possible price area where the trend may change. Change does not necessarily mean from up to down. It could also mean from up to sideways (or range trading), down to up, down to sideways or sideways to up or down. Recall from last week's lessons that the market can have one of three trends: up, down or sideways. So, a break of a trendline does not have to mean that the direction of prices must reverse. It could just mean that prices will consolidate for a while before resuming the prior trend.

How do I draw a trendline?
If you remember your grade school geometry, you can draw an infinite number of straight lines through a single point. However, there is only one straight line that can be drawn through two points. Therefore, the minimum number of points that you need to draw a trendline is two. When you draw an up trendline, you connect the lows of the bars or candlesticks on your chart. When you draw a down trendline, you connect the highs.

Take a look at the chart below. What do you think of the down trendline that I drew?

Three touches required to confirm a trendline
Clearly, prices are falling. In fact, they've been falling since the middle of June. Notice that there are only two touches on the trendline. The second day after the line begins (large red real body and large upper shadow) does almost touch the trendline, but I would typically ignore that day. Barring a huge acceleration lower (in the case of a downtrend), almost by definition, the trendline will be near prices on the second day. This trendline can only be considered a tentative trendline. You can draw an infinite number of trendlines between any pair of points on a chart. The trendline is only confirmed once the line has been touched (or nearly touched) three times. I cannot tell you how many times I see my friends in the media talk about a trendline that broke, and then show a chart on the television with a line that only touches two points. That is not a trendline.

Let's look at that OIH chart again:

Broken trendlines reverse roles
As you can see, I drew the first down trendline (in green above) off the high in June. Although prices continued their descent through July, the speed with which they were falling eased. In other words, the momentum of the price drop slowed. Notice also how the green line has acted as support the last few days. It is common for a broken down trendline to act as support as time goes by. Likewise, a broken up trendline will often act as resistance.

Let's think about the green trendline that broke in late June. After it broke, prices did not reverse higher. Instead, they meandered for a bit, but essentially continued lower (albeit at a slightly slower pace). When prices initially fall or rise very quickly after a trend change, then you will often need to redraw your trendline one or two more times before the stock, or market, settles into a more sustainable trend.

What is a valid trendline break?
I have always said that technical analysis is more of an art than a science. Although you can write trading systems based on indicators, recognizing reversal and continuation patterns, or determining whether a trendline has broken is often in the eye of the beholder.

I have seen many authors suggest that a break of a trendline is only valid if the break is still present upon the daily close. Others require that prices close through the old trendline for at least two or three days. Another rule of thumb is that prices must move through the trendline by at least 3%. The latter rule is of little use in certain low-volatility markets, or if you are day trading. For that matter, if you are highly leveraged, a 3% adverse price move against you could be highly damaging.

All of these rules of thumb are interesting to examine, but there are better ways of determining whether a trendline break is valid. I always look for additional evidence, such as:

• Increased volume
• Momentum in the break's direction rises
• Key moving averages are crossed
• The market gaps away from the broken trendline

A pet peeve
Back in the old days, everybody drew their trendlines by hand, using a pencil and ruler on paper. In these cases, you really didn't know the exact spot where the trend broke. The exact location of the trendline depended on such random factors as how close you held the pencil to the ruler, and how sharp the point on your pencil was. In these cases, analysts often took license and would draw trendlines that slightly broke extremes (in an attempt to find a better fit).

In this day and age of computer-based technical analysis programs, with their myriad of drawing tools, the art of technical analysis is often lost. I do not believe there is a great advantage in knowing, to the penny, the exact price level where a trendline might fail. As I noted above, there are many other factors that you must consider before taking action on a trendline break. Computer-based drawing tools make an approximation appear to be a scientific fact, but this can ultimately be dangerous to your profit and loss statement.

Using trendlines in your trading, investing and analysis
I am certainly not suggesting you should NOT use trendlines in your work with the admonition above -- quite the contrary. However, you should always remember to use trendlines in conjunction with other tools. Although it is beyond the scope of this article, trendlines have very definite meanings when used in Elliott Wave (see my book, Applying Elliott Wave Theory Profitably, for more information on trendlines in Elliott Wave).

If your trades were with the old trend, then a trendline break should have you cover outstanding positions as long as there is other supporting evidence. Reversing your trade may only be advisable if the evidence for a trend change is incontrovertible. Such evidence may include a sharp increase in volume, confirmation of a reversal pattern (head and shoulders, double top confirmation), or a breakaway gap. If you do not have this kind of evidence, then I would advise you look for a retest to the old trendline before reversing your position. Trendline retests are quite common, and will allow you to enter your new position with tight stops (typically, you would place your stops beyond the price extreme on the day prices first breached the trendline).

SUMMARY
I could probably write a chapter in a book just on trendlines. No technical analyst worth his weight in salt leaves home without them. However, proper usage requires understanding what a trendline represents and what a break of one means. Using trendlines is as much an art as it is a science. Experience will help you to better identify lasting trend reversals from failed trendlines.

A number of other topics that are closely related to trendlines, but these are beyond the space I have in this week's issue. I will cover channels, which are very important in Elliott Wave Theory, at a later date, as will internal trendlines and support and resistance lines.

Steven Poser
Editor
The ETF Authority
New York, NY