I've spent the past three issues explaining how you can recognize a reversal pattern. As you will recall, you can use a reversal formation to help forecast a change in trend from up to down or down to up. However, focusing too much on reversals can be dangerous to your financial health. Markets might be in a trend, or they might be in a consolidation. If you spend all of your time trying to identify reversals, then you will miss a much larger set of opportunities -- trading with the trend.

In today's report I am going to begin our coverage of continuation patterns. Continuation patterns occur when an up or downtrend takes a rest. In these cases, the financial instrument (stock, bond, ETF) you're examining will often move sideways (consolidate) for a period of time, or may even correct against the larger trend.

DANGER, DANGER!
I cannot go on without offering you a quick reminder. Just as you should not try to find reversals under every rock, you must also refrain from constantly seeking out consolidation and correction patterns. The trend is your friend, and if the market is moving in your favor, then your best bet is usually to stick with it. If you're constantly seeking a better price to enter into a trade, this leads to hope. Hope is your enemy. If you have a trend, then stick with it until it does something wrong. Only at that point should you try to determine whether you are in a correction/continuation or a reversal.

TRIANGLES COME IN THREE FLAVORS
One of the most common types of continuation patterns is the triangle. Technical analysts distinguish between three types of triangles: Symmetric, ascending and descending. A symmetric triangle consists of a pair of converging trendlines -- one falling (resistance) and one rising (support). An ascending triangle consists of a rising support line and a flat resistance line. And finally, a descending triangle is formed by a falling resistance line and a flat support line. To help you visualize these various patterns, I've included a set of stylized drawings below.

As you can see, I drew the ascending triangle with a bullish breakout and the descending triangle with a bearish resolution. Most commentators will tell you that ascending triangles are always bullish and that descending triangles are always bearish. While the strong tendencies for these patterns are as discussed, you need to consider where you are in the trend before making a final determination. (More on this later when I discuss the tie-in to Elliott Wave Theory.)

Symmetric triangles are often referred to as continuation triangles. However, statistical studies say that the probability of a continuation in the direction of the trend is barely above 50%. That said, understanding where you are in the trend will help you to more properly filter these signals effectively.

PSYCHOLOGICAL BASIS FOR TRIANGLES
Let's think about what happens as an ascending triangle develops. Initially, prices have been rising within an uptrend. Once they reach a certain price level the trend is then turned back. That level might have been from a previous high, a volatility band, a retracement, or seemingly from nowhere. Regardless of the cause, the important thing to note is that prices reverse.

From this high, the market then enters a retracement. If price momentum was at a new trend peak at the time of the initial reversal, then I would feel extremely comfortable that the ultimate breakout will be higher. At this juncture, however, we do not know with a high degree of confidence whether we have a triangle, a reversal, or something else.

Once prices slip and rally again, they should ultimately reach the prior peak. Prices don't have to hit the prior high to the exact penny, although that is possible. Instead, price might slightly exceed its old high or fall a bit short. However, as long as it comes relatively close to the prior high, then the pattern is okay.

Here is where it can get a bit tough. As prices reverse from the second high, it is impossible not to consider the possibility that the market might be tracing out a double top. Momentum will likely diverge. Volume will probably be a bit lower on the second peak. However, as prices fall, the losses will probably be grudging, and the volume will continue to fall as prices slip. (If it was a reversal pattern (double top), then volume would often increase after peaking a second time.)

The next hint -- and it will be a strong one -- is that prices will not fall to the prior low. (A double top is only confirmed when the interim low is broken.) Prices might then begin a run to a new high. However, there is a good chance that the lateral resistance will hold one more time, and that we will see another fall to the trendline (this line can be drawn from the two prior reaction lows). In fact, if you recall my educational article from September 8, 2003, I warned that triple tops are relatively rare. More often than not, prices will ultimately break higher. So, when you see that third peak and subsequent turn lower, that is usually another major hint that prices will ultimately continue higher.

Think about the behaviors here: Prices keep banging into the resistance level and fail. However, each reaction lower is met at higher and higher support levels and on lower and lower volume. Buyers are anxious to buy and are not patient enough to let price fall more. When it comes to an ascending triangle, that is all you need to understand. Ultimately, the shares are likely to break out higher.

WHEN SHOULD THE PATTERN BREAK OUT?
Darned good question there! Project the two lines in the triangle into the future and note where they meet. For example, look at the Financial SPDR (XLF) chart shown above. The start of the pattern was July 1, 2003 -- the bottom of the first low. If you moved the chart to the right, then you would see that the two lines meet on January 14, 2004. The breakout to higher prices should occur no later than two-thirds to three-fourths of the way through the pattern, and usually takes place sometime after the midpoint. This breakout occurred just before the midpoint, so I would not be surprised to see a test back inside the triangle before a final break higher later. If XLF does not manage a good breakout, then watch out! Pattern failures can be ugly.

WHAT ELSE SHOULD I LOOK FOR?
When looking at stocks and indices, a successful breakout usually means a move of at least 3% above the resistance line (in the case of an ascending triangle). It is not uncommon to retest the breakout level. A break back into the triangle is not all that uncommon either. However, any such move should be on lower volume than when prices broke higher.

SO, WHAT CAN GO WRONG?
If any of the previously noted behaviors do not occur, then there might be a problem with the triangle pattern. For example, if volume increases as prices fall, you should be worried. If prices fall below the support line, that is a reason to be concerned that the pattern will not result in a bullish continuation. If momentum is stronger on the way down, that could be a warning that the ascending triangle will fail and prices will trade lower. However, as long as volume fades on the way down and the support line holds, the odds are very high that the ascending triangle will work out favorably.

BEWARE THE WEDGE!
A quick warning is worth noting. If you haven't already read it, then please take a look at last week's educational article on wedges (from our September 15, 2003 ETF Authority issue). If the three succeeding peaks result in slightly higher prices, then there is substantial risk that the formation you're looking at will turn out to be a wedge, NOT an ascending triangle. Recall that wedges ultimately result in reversals. Also, note that a break below support does not mean prices will reverse. However, there is a strong likelihood that a trading range -- sometimes called a rectangle -- will develop.

DESCENDING TRIANGLES
There is no real need to go through the descending triangle description. All you need to do is reverse the above price directions noted in our discussion of ascending triangles above. Volume typically falls throughout the pattern, just as it does in the ascending triangle. More often than not, the final resolution of a descending triangle is to lower prices.

PATTERN TIMEFRAME
Triangles usually play out over a period of several weeks, at least according to the books. I will tell you right now that there is really no difference between a triangle and what is often called a "pennant." Pennants are continuation patterns that last from intraday to several days in duration. (I'll cover these in a future issue.)

Although triangles usually take a few weeks to complete, I have seen them drawn out on weekly and monthly charts as well. Beware of underlying volume trends, however, when looking at such long-term charts. What I mean by this is that you need to make sure you do not eliminate the triangle as viable option just because the volume pattern is not correct. I am pretty sure that as Intel (INTC) became a household name, even if it formed a triangle, volume kept rising as the company came to dominate the booming microprocessor market. If there is strong growth in interest in a stock, or we are talking about a new product, such as the ETFs we cover in this newsletter, then volume might continue to rise even though the fund, stock or commodity you are looking at is tracing out a valid triangle.

SYMMETRIC TRIANGLES
As I noted earlier, a symmetric triangle occurs when two trendlines converge. Volume typically falls throughout the pattern. The timeframe for a breakout is the same for a symmetric triangle as it is for an ascending or descending triangle (two-thirds to three-fourths of the way through the pattern, and generally not before the midpoint).

The main difference between symmetric triangles and ascending or descending triangles is that it is much harder to determine which direction prices will ultimately resolve.

SO, HOW DO I FIGURE OUT WHICH WAY THE MARKET WILL BREAK?
Look for other markers. Notice the chart above. It is not a stock, but rather shows how many US dollars you would have to pay to buy one Euro (the currency used through much of Europe). As you can see, RSI, a momentum measure, diverged at the low in late 2000. That was a strong hint that the market might reverse instead of continue. Also, momentum confirmed the breakout in 2002 (there was a new high in momentum).

WHAT ABOUT ELLIOTT WAVE'S INTERPRETATION OF TRIANGLES?
Triangles are always corrections in Elliott Wave (EW) parlance (forget about wedges/diagonals, as they are not true triangles even though that is the term used in EW). This means that if you are in an uptrend, and in wave-4 (a correction in an uptrend), then even if a descending triangle forms, the breakout will ultimately be to higher prices.

This, of course, does not directly allow for a situation such as what I showed you in the dollar versus the Euro chart. In EW, that would not really be termed a triangle. I would have to find a complete wave count at the 2000 low and then count a bullish move from that bottom. In my opinion, this is a much better way of determining the trend, as it allows you to essentially forecast what kind of pattern to expect. Unfortunately, it takes a great deal of effort to learn Elliott Wave properly, so that analysis would be beyond the scope of this report. However, I know of a great book that should help! ("Applying Elliott Wave Theory Profitably," by yours truly.)

SUMMARY AND STRATEGIES
Triangles are usually consolidation patterns. If you understand where you are in a trend, then you can, with a high level of confidence, determine which way a triangle will break. However, the pattern is not confirmed until prices break out higher. If the breakout occurs on strong volume, then the best strategy is to enter the trade on a close past the breakout line. In the case of an ascending triangle, that means above resistance. For a descending triangle, it means selling below support. The break can be in either direction in the case of a symmetric triangle. When I trade these types of patterns, I generally place a stop about 1-2% inside the triangle. Meanwhile, I usually compute the price target by finding the height of the triangle -- the price difference between its high and low point. I then either add (for an ascending triangle) or subtract (in the case of a descending triangle) that difference from the point at which prices broke out. I prefer to use percentages to calculate the height and target, but most people use actual dollars and cents.

Good trading!

Steven Poser
Editor
The ETF Authority
New York, NY