As of 09/22/2021
Indus: 34,258 +338.48 +1.0%
Trans: 14,001 67.19 0.5%
Utils: 898 0.44 0.0%
Nasdaq: 14,897 +150.45 +1.0%
S&P 500: 4,396 +41.45 +1.0%

YTD
+11.9%
+11.9%
+3.9%
+15.6%
+17.0%

35,600 or 33,800 by 10/01/2021
15,400 or 14,000 by 10/01/2021
930 or 890 by 10/01/2021
15,600 or 14,700 by 10/01/2021
4,650 or 4,350 by 10/01/2021

As of 09/22/2021
Indus: 34,258 +338.48 +1.0%
Trans: 14,001 67.19 0.5%
Utils: 898 0.44 0.0%
Nasdaq: 14,897 +150.45 +1.0%
S&P 500: 4,396 +41.45 +1.0%

YTD
+11.9%
+11.9%
+3.9%
+15.6%
+17.0%
 
35,600 or 33,800 by 10/01/2021
15,400 or 14,000 by 10/01/2021
930 or 890 by 10/01/2021
15,600 or 14,700 by 10/01/2021
4,650 or 4,350 by 10/01/2021
 
Revised, new statistics, 5/14/2021.
Price follows trends. When you draw an upsloping line along price valleys, the stock often touches the line and rises away from it without piercing it. The line is called a trendline because it shows the price trend.
$ $ $
My book, Trading Classic Chart Patterns, pictured on the left, has two, yes, two, chapters dedicated to trendlines.
If you click on the above link and then buy the book (or anything) while at Amazon.com, the referral will help support this site. Thanks.
$ $ $
Up Trendline Chart Pattern

Characteristic  Discussion 
Log scale  Use the logarithmic scale. Price will signal a trend change sooner on the log scale than on the arithmetic scale. 
Minor highs  Draw an upsloping trendline along the price valleys. That way, when the trend changes from up to down, you'll know with a trendline pierce. The numbers in the above chart show price touching the trendline at four minor lows. 
Spacing  Widely spaced touches (over the median 12 days each) suggest a more powerful move postbreakout. 
For the discussion which follows, I logged 3,720 trendlines with 3,172 of them happening in bull markets (only bull market results are reviewed in this article). I measured the move after the breakout to the ultimate low and after the third trendline touch to simulate trading them (more about that later). I used 699 stocks with trendlines priced above $5 per share from data starting in March 1991 through April 2021. Not all stocks covered the entire period.
Use the measure rule to predict how far price will tumble after a downward breakout (a price pierce) from the trendline. The figure to the right shows an upsloping trendline with price breaking out downward at point B.
If you'd like to use the classic measure rule, then from the breakout, find the prior minor low trendline touch. I show it as point A. Measure the tallest distance between those two points (re, A and B), measured vertically. In this case, that's the distance from C to D. Multiply that distance by 26% because that's how often this method works when a full height is used. Project the result downward from the breakout price (B)  the point where price pierces the trendline.
For example, if the high at C is 10 and directly below that at point D, the trendline is at 8, the difference is 2. Multiply this by 26% to get $0.52. Suppose the breakout at point B is at 9. That would give a price target of 8.48 (9  0.52). If the projected decline is less than 0, ignore the result.
Because the measure rule works only 26% of the time, I suggest you just use the price at the trendline start as the target. Price will reach that target 58% of the time, on average. Alternatively, you might look for underlying support where price might stop and reverse.
I was looking at trendlines, trying to discover if there was a way to trade them and make money. I thought that if I could buy a stock showing a trendline at the third touch, ride price higher until I got a downward breakout and then sell, I could pocket the difference.
The chart on the right shows an example. The 3touch trendline is in red. It ends in mid June. Four days later, buy at the open and hold until the day after price closes below the trendline (the orange/brown line). The difference between the buy price and the sale price is profit. If you have the courage to buy sooner at or after the third touch, then profits would likely increase.
I found that on 3,172 trades, you'd make an average of 0.7% or lose a median of 0.9%. The win/loss ratio was 37%. In other words, this method didn't work well.
Along the way, I learned features about trendlines. For the tests which follow, I found the third trendline touch and waited 3 days (price bars) before buying at the open the following day (4 days, total, after the minor low). This is because my software uses 3 bars after a minor low to assure that it is, indeed, a minor low on the trendline.
Because the first close below a trendline worked better than a delayed exit, all tests which follow use the first close as a sell signal (sell at the open the next trading day). I explain the delayed sale idea next.
Pictured is an upsloping trendline in a downward price trend. I found this type of setup happening frequently in the charts I looked at. As bad as it looks, the difference between an uptrendline in an upward versus a downward trend is just one percentage point in performance: 13% (in uptrend) versus 14% (in downtrends). That's the average drop after the breakout to the ultimate low.
A review of the statistics shows that the 5% failure rate is 30%. Almost a third of trendlines with downward breakouts will see price drop no more than 5%. Half (52%) will drop no more than 10%.
Briefly, I found the following based on the drop from the breakout price to the ultimate low (which is different than the trading method described above.
I spent 6 weeks looking at trendlines and studying them. I found several scenarios which appeared and tripped up trendlines. Here's a brief review.
I found other scenarios which happened infrequently.
Here are more tips, in slider format.
 Thomas Bulkowski
See Also

Support this site! Clicking any of the books (below) takes you to Amazon.com. If you buy ANYTHING while there, they pay for the referral.
No matter how cynical you become, it's never enough to keep up.  Lily Tomlin