The double top is a major reversal pattern that forms after an extended uptrend. As its name implies, the pattern is made up of two consecutive peaks that are roughly equal, with a moderate trough in between.

Although there can be variations, the classic double top marks at least an intermediate change, if not long-term change, in trend from bullish to bearish. Many potential double tops can form along the way up, but until key support is broken, a reversal cannot be confirmed. To help clarify, we will look at the key points in the formation and then walk through an example.

1. Prior Trend: With any reversal pattern, there must be an existing trend to reverse. In the case of the double top, a significant uptrend of several months should be in place.
2. First Peak: The first peak should mark the highest point of the current trend. As such, the first peak is fairly normal and the uptrend is not in jeopardy (or in question) at this time.
3. Trough: After the first peak, a decline takes place that typically ranges from 10 to 20%. Volume on the decline from the first peak is usually inconsequential. The lows are sometimes rounded or drawn out a bit, which can be a sign of tepid demand.
4. Second peak: The advance off the lows usually occurs with low volume and meets resistance from the previous high. Resistance from the previous high should be expected. Even after meeting resistance, only the possibility of a double top exists. The pattern still needs to be confirmed. The time period between peaks can vary from a few weeks to many months, with the norm being 1-3 months. While exact peaks are preferable, there is some leeway. Usually a peak within 3% of the previous high is adequate.
5. Decline from peak: The subsequent decline from the second peak should witness an expansion in volume and/or an accelerated descent, perhaps marked with a gap or two. Such a decline shows that the forces of demand are weaker than supply and a support test is imminent.
6. Support break: Even after trading down to support, the double top and trend reversal are still not complete. Breaking support from the lowest point between the peaks completes the double top. This too should occur with an increase in volume and/or an accelerated descent.
7. Support turned resistance: Broken support becomes potential resistance and there is sometimes a test of this newfound resistance level with a reaction rally. Such a test can offer a second chance to exit a position or initiate a short.
8. Price Target: The distance from support break to peak can be subtracted from the support break for a price target. This would infer that the bigger the formation is, the larger the potential decline.

While the double top formation may seem straightforward, technicians should take proper steps to avoid deceptive double tops. The peaks should be separated by about a month. If the peaks are too close, they could just represent normal resistance rather than a lasting change in the supply/demand picture. Ensure that the low between the peaks declines at least 10%. Declines less than 10% may not be indicative of a significant increase in selling pressure. After the decline, analyze the trough for clues on the strength of demand. If the trough drags on a bit and has trouble moving back up, demand could be drying up. When the security does advance, look for a contraction in volume as a further indication of weakening demand.

Perhaps the most important aspect of a double top is to avoid jumping the gun. Wait for support to be broken in a convincing manner, and usually with an expansion of volume. A price or time filter can be applied to differentiate between valid and false support breaks. A price filter might require a 3% support break before validation. A time filter might require the support break to hold for 3 days before considering it valid. The trend is in force until proven otherwise. This applies to the double top as well. Until support is broken in a convincing manner, the trend remains up.

The double top in Ford took about 5 months to form. Even after the support break, there was another test of newfound resistance almost 4 months later.

1. From a low near 10 in Mar-97, Ford advanced to 36 by Dec-98. The trendline extending up from Mar-97 is an internal trendline and Ford held above it until the break in May-99.
2. From the first peak, the stock declined around 15% to form the trough.
3. After reaching a low near 30 1/2 in early Feb, the trough formed over the next 2 months and there wasn't a rally until early April. This long drawn out low suggested tepid demand.
4. The rally from 30 1/2 to 36.80 occurred on fairly good volume, but money flows barely surpassed +10%. The high at 36.80 was about 2% higher than the previous high, but within the 3% threshold. The distance between the two peaks was about 3 months.
5. The decline from 36.80 occurred with two gaps down and increased volume. Furthermore, Chaikin Money Flow promptly moved below -10%. The speed with which money flows deteriorated indicated a serious increase in selling pressure.
6. In late May and early June, the stock traded for about 3 weeks at support from the previous low. During this time, money flows declined below -20%. Even though the situation looked ominous, the double formation would not be complete until support was broken.
7. Support was broken in early June when the stock fell below 28 1/2, which was more than 3% below support at 30 1/2. After this sharp drop, there was an equally sharp advance back above the newfound resistance level. While a test of broken support can be expected, it is usually not quite this early. The advance to 32 in late June may have triggered some unpleasant short covering for those who jumped in on the first support break. The stock fell to 25 and then began the retracement advance that would ultimately test support.
   
   

On the second chart, 30 3/4 marked the support turned resistance level and 31 marked a 50% retracement of the decline from 36.80 to 25. Combined with the price action in early June and early July, a resistance zone could probably be established between 31 and 32. The stock subsequently formed a lower high at 30 in Jan-00 and declined to around 22 by mid-March.