As of 10/23/2020
  Indus: 28,336 -28.09 -0.1%  
  Trans: 11,880 +131.58 +1.1%  
  Utils: 894 +2.40 +0.3%  
  Nasdaq: 11,548 +42.27 +0.4%  
  S&P 500: 3,465 +11.90 +0.3%  
YTD
-0.7%  
 +9.0%  
 +1.7%  
 +28.7%  
 +7.3%  
  Targets    Overview: 09/29/2020  
  Up arrow29,300 or 27,400 by 11/01/2020
  Up arrow12,000 or 11,000 by 11/01/2020
  Up arrow915 or 840 by 11/01/2020
  Up arrow12,800 or 11,300 by 11/01/2020
  Up arrow3,700 or 3,400 by 11/01/2020
CPI (updated daily): Arrows on 9/25/20
As of 10/23/2020
  Indus: 28,336 -28.09 -0.1%  
  Trans: 11,880 +131.58 +1.1%  
  Utils: 894 +2.40 +0.3%  
  Nasdaq: 11,548 +42.27 +0.4%  
  S&P 500: 3,465 +11.90 +0.3%  
YTD
-0.7%  
 +9.0%  
 +1.7%  
 +28.7%  
 +7.3%  
  Targets    Overview: 09/29/2020  
  Up arrow29,300 or 27,400 by 11/01/2020
  Up arrow12,000 or 11,000 by 11/01/2020
  Up arrow915 or 840 by 11/01/2020
  Up arrow12,800 or 11,300 by 11/01/2020
  Up arrow3,700 or 3,400 by 11/01/2020
CPI (updated daily): Arrows on 9/25/20

Bulkowski on Trading using Knots

 

Released 9/28/2020.

Trading Knots: Summary

Knots are useful for the following.

  1. Traders who wish to sell short now have an accurate gauge to determine how far price will drop after a breakout.
  2. Traders and investors, after price reaches a knot, can buy long and ride price upward.
  3. Investors: People who own a stock may wish to know how far price will drop before finding support. Maybe it'll indicate it's time to sell.
  4. Swing traders and investors: Those who wish to know that a large decline has a good chance of occurring. They may wish to sell a holding immediately or sell short.

I define a knot as a small congestion area of at least three price bars which has lots of overlap in a strong price trend. Here's what I discovered.

Trading Knots: Background

Picture of the a double top and knot.

I used the Patternz trading simulator to look for double tops, knots, and how they behave.

Pictured is a OHLC bar chart because it's easier to see the overlap in the price bars using this chart compared to a candlestick chart.

A double top chart pattern appears at the top of the chart. I didn't check any other chart pattern type (like triple tops, head-and-shoulders tops, and so on), nor qualify the type of double top (Adam and Eve combinations) in my search. If Patternz found a double top, then I used it.

In this chart, the double top appears at the end of an uphill run, which I show as the rise from A to B.

Along the way, a knot appears (circled in red). The knot is a place along the AB uptrend, priced below the breakout of the double top, where price moves sideways for at least three price bars. This one is a flag but it need not be. Often the knot is irregularly shaped, so it's not a flag nor a pennant, but I didn't qualify the knot as any type of chart pattern.

After the double top forms, it breaks out downward at C. Below C is where you start looking for a knot priced in the AB uptrend. I use C and not the confirmation price (the horizontal blue line) because it's important to see how far down price drops on the day of breakout.

Price drops from C to D, hits the price of the knot and pulls back to the confirmation price. The knot is the first one down along trend AB in the move from C to D. It's the first knot of support that the stock runs into as price drops.

Here's what to look for.

Top of page

Trading Knots: Methodology

I use the simulator in Patternz to find double top chart patterns and simulated trading them in a short sale.

I examined each chart and looked for a strong uptrend preceding the double top. Then I looked for a knot along that uptrend. If I found one, I started the simulation to see how far price dropped after confirmation of the double top.

I only looked at confirmed double tops, not twin peaks which didn't confirm.

For knots acting as support...

Picture of the a double top and knot.

I only looked at the knot closest to the double top (positioned at or below the breakout price, not the confirmation price).

If the knot showed support, I looked for price to pull back to the bottom of the double top (back to the confirmation price) or rise higher.

Here's what to look for to see if a knot will act as support. Reference the figure on the right.

In 173 simulations, I found that knots worked 64% of the time (I manually counted this, so your results may vary).

Top of page

For knots predicting a big decline...

Picture of the a double top and knot.

In my simulations, I discovered that knots located near the breakout price tended to predict a large decline.

I searched for a knot that was at or just below the confirmation price (the lowest price between the two peaks). In the chart, the knot is at B and the confirmation price is at E. Notice that the B to E distance is much closer than in the prior chart.

Once I found a candidate, I assessed whether the knot worked. If it worked, the price decline met or exceeded (eyeballing it) the measure rule for double tops. In other words, I looked at the height of the double top and checked if price declined 'height' distance below the bottom of the double top.

As an example, the height of this double top is DE and the target price would be E - DE. If E is 30, D is 35, then the target would be 30 - (35-30) or 25.

If the stock looked to meet or exceed the measure rule target on the way to the ultimate low, then I marked the knot as successful. If price first climbed above the top of the double top, then the knot was a failure.

Here's what to look for to see if a knot will predict a strong drop (reference the figure).

The decline to the ultimate low should be at least the height of the double top subtracted from the price of E. Sometimes, the drop is much farther. In technical terms, the decline meets or exceeds the measure rule target.

Tests show that price met or exceeded the measure rule target (and often much farther as price dropped to the ultimate low) in 90 of 129 tests or 70% of the time. I eyeballed the distance, so your results may vary.

Top of page

Trading Knots: Setup Form

Picture of the a double top and knot.

Here's how to setup the Patternz simulator (free software) to find knots and see how they behave.

After selecting stocks you wish to simulate trading on the Main Form, clicking the Simulator button near 2 (shown at the link), will take you to the Simulator Form.

Click the Setup button near the lower right of the form. A form similar to the one shown will appear, although I cut off a portion of the bottom of the form.

Circled in red is how I like to configure the simulator for this test.

I checked "Ultimate low found" to stop the simulator at the end of a perfect trade.

I checked "Auto set targets" to fill in the short sale targets for me, so I don't have to manually put something in.

The other settings show that the simulator will load a new chart and find the "breakout" (of a chart pattern) before pausing.

"Show circles" draws colored circles on the chart to show where I sold short, bought back, and the ultimate low are located.

Top of page

Trading Knots: Simulator Form

Picture of the a double top and knot.

This is the Patternz's Simulator Form, or a portion of it.

At A, circled in red, is the Short Sale check box. This tells the program to use short sales when calculating the target prices. Leaving this unchecked assumes you'll be looking for price to rise. We are expecting price to drop, so we check this. We want to sell the stock short.

At B, I set the simulator speed control at full speed, but often I slow it down some to let price scroll more slowly.

I checked "Skip" at C to skip stocks which don't have the double top chart pattern. Don't forget to click the Patterns button and pick "Double top" while there.

At D appears the colored circles I mentioned. D shows the ultimate low. DT is the double top. The red circle is the sell price (positioned on the date of sale), the blue circle is the buy price, and the green circle is the ultimate high.

To run the simulator, just click one of the double top choices in the list box (3 are shown, located above the Speed control at B). The chart will show the pattern at the breakout price (or wherever you told the program to pause).

Click Okay at the box which tells you to click Resume to begin the simulation.

Then click Resume, grab the popcorn while the trade unfolds.

The chart of Exelon shows a knot just under the double top. It suggests, but does not guarantee, that price will make a decent decline, which it does.

Top of page

Trading Knots: Warnings

Picture of the a double top and knot.

If a trend leading to the double top is long and a knot, such as a flag or pennant appears midway, price after the breakout may not drop far enough to touch the knot before pulling back.

For example, look at the situation shown in the adjacent chart. Price bottoms at A, rises to form a knot at B and then continues up to a double top at CD. This confirms as a double top when price closes below E.

In this case, the knot at B is too far for price to drop on the way to F.

Instead, the stock finds support at the prior knot H, and rebounds to pullback to G.

I have seen situations where support at H is missing (such as when the stock is making new highs in a straight-line run up. Pretend everything to the left of the green line is missing) and price will drop partway down the AC uptrend and pull back.

Even if the stock does reach the knot, the knot might be too far away from the confirmation price (E) to see the stock pull back that far.

In this case, the stock after D drops all the way down to B, finds support there, but can't climb back up to E, so a pullback, technically, never completes.

Top of page

Trading Knots: Examples

The following series of slides shows examples of knots and how to use them to determine how a stock will behave.

1 / 14
chart pattern
Knot works 64% of the time.

This is an example of how a knot should work. In a strong uptrend, where price breaks out to a new high, a double top forms (DT). Along the uptrend from A to DT, a knot (B) forms.

The knot is the closet one to the breakout day, D. Price drops to C, hits support at the knot, and pulls back to the chart pattern. In this example, the stock rises to E. This is how a knot should behave. Notice the distance between the bottom of the double top and B. To trade this, you'd wait for D, the breakout, to appear before looking for the knot. In some cases, the breakout might drop past the knot, rendering it useless.

Another example on the next slide.
2 / 14
chart pattern
Knot works.

This example shows a knot at B providing support to price when it drops to D. The knot appears in a strong move up from A to DT (a double top chart pattern). Price breaks out downward from the double top at C. A pullback occurs which sends price up to E.

In this example, there is prior price movement well above the top of this chart pattern, but it's not shown on the chart.

Another example on the next slide.
3 / 14
chart pattern
Knot works.

This is another example of where a knot (A) provides support at B. Notice that there is prior support at C which didn't affect the drop from the double top (DT) to B.

Another example on the next slide.
4 / 14
chart pattern
Knot works.

Price forms a tight knot at A. Tight means lots of overlap and they often work best to support price. At C, the stock breaks out downward from the double top (DT). Price drops to find support at B, at the price of A. After that, the stock rises to D and eventually makes its way upward.

The next slide shows an example of a knot that didn't work.
5 / 14
chart pattern
Knot fails.

The double top is at DT. The breakout day is C. The closest knot to C is B. The stock should have dropped to B and turned higher, but it didn't. The next closest knot is A and the stock gapped around that and continued lower, too.

In this case, the knot is awfully close to the bottom of the double top, so this knot might really signal not support, but a large decline to come. And that's what happens. We'll see more of this kind of setup later in slides 8 to 11.

The next slide shows an example of a knot that didn't work.
6 / 14
chart pattern
Knot fails.

In this example, the breakout from the double top (DT) is at A. The closest knot below A is at B. I would expect the stock to turn there, but it didn't. The stock does find support at B, shown to the right of the red line (where price moved sideways for more than two weeks), but the stock continued lower instead of attempting or completing a pullback to C.

The next slide shows an example of another knot that didn't work.
7 / 14
chart pattern
Knot fails.

Price forms a knot at A. This one looks loose (less overlap, more meandering of price) compared to some of the knots appearing in prior slides. After the breakout (B) from the double top (DT), the stock drops to B and pulls back to D. I recommend placing a target at the top of the knot (at the red line). Price usually will drop to the top of the knot immediately after the breakout before a pullback begins. In this case, the pullback happened before the stock dropped to C.

Also notice that the stock dropped below the bottom of the knot, signaling this knot was a failure.

The next slide shows how a nearby knot of support indicates a big decline.
8 / 14
chart pattern
Knot predicting a big decline. This works 70% of the time.

The knot at A is long and yet tight. You'd expect the stock to hit that and bust the downward breakout (by soaring). Instead, the stock plummeted to the ultimate low at C (see the price scale). Notice the position of A in relation to B. B is the confirmation price (not the breakout). The knot is at or just below B and that's important. Also note that this works 70% of the time, but only by predicting a minimum drop below B of the height of the double top. (Find the height of the double top, subtract it from B and the result becomes the minimum price target.

The next slide shows another example of how nearby support indicates a big decline.
9 / 14
chart pattern
Knot predicting a big decline.

This is a recent example of a knot at A, circled in red, B marks the confirmation price, and C is the breakout price. Notice the position of knot A to confirmation price B. On the breakout to C, the stock reaches its target easily (the pattern's height subtracted from B). A pullback sucks price up to E before resuming the downward move. The stock reached the ultimate low (not shown) at 177.

The next slide shows another example of how nearby support indicates a big decline.
10 / 14
chart pattern
Knot predicting a big decline.

This chart shows a familiar setup. The knot is at A, right at the confirmation price of B. After the breakout, the stock drops to C, the ultimate low.

The next slide shows another example of how nearby support indicates a big decline.
11 / 14
chart pattern
Knot predicting a big decline.

This is the same stock as the last slide except about 4 years later. The knot at A is loose looking and yet it still works well. It's located slightly above the confirmation price, B. The stock drops and finds the ultimate low at C.

The next slide shows the failure of a nearby knot to predict a big decline.
12 / 14
chart pattern
Failure of a knot to predict a big decline.

The knot is at A, circled in red. The double top (DT) has a confirmation price at B, so the knot is properly positioned at or just below the confirmation price. The knot looks tight but somewhat sloppy with price making tall price bars. I don't know if that's important or not. C is the breakout but along the way to C, peak D appears. That changes the potential double top to a triple top.

Even though price breaks out downward at C, the stock vaults to E and then tumbles to F and continues lower (not shown). The stock did make a big drop, but you'd have been stopped out first. This triple top double busted.

The next slide shows the failure of a nearby knot to predict a big decline.
13 / 14
chart pattern
Failure of a knot to predict a big decline.

This is a setup similar to the last slide. The knot is at A, taking the shape of a flag tilted upward. The double top is at DT, confirmed when price closes below the low at B. That happens at D, but along the way, C makes a third top, changing the double top to a triple top. It makes me wonder if triple tops fail more often with nearby knots than otherwise (I haven't checked this, but maybe it's true).

Anyway, the stock dropped from the price of B to E, which was less than the height of the pattern (roughly C to B). So it didn't meet my definition of a big decline (at least the height of the pattern below the confirmation price). In this case, the stock climbed more than 20% on the way to F, even though it doesn't look like a big move. E marks the ultimate low for the stock.

The next slide shows the failure of a nearby knot to predict a big decline.
14 / 14
chart pattern
Failure of a knot to predict a big decline.

The knot is the loose looking congestion area at A. The double top is a tiny one at DT which confirms as valid when it breaks out downward at B, on the way to the ultimate low at C. Technically the drop from confirmation to C is taller than the height of the double top, so this does qualify as a big decline. However, your pocketbook or wallet won't be enhanced by the meager 2% drop and that's if you trade it perfectly. The stock rises to D, busting the downward breakout.

Notice that this time we're using a double top and not a triple top.

The End.

See Also

 
Top
 

 

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.

My novels:  Bumper's Story Head's Law

Chart Patterns: After the Buy Getting Started in Chart Patterns, Second Edition Trading Basics Fundamental Analysis and Position Trading Swing and Day Trading Visual Guide to Chart Patterns Encyclopedia of Candlestick Charts Encyclopedia of Chart Patterns 2nd Edition Trading Classic Chart Patterns

Copyright © 2005-2020 by Thomas N. Bulkowski. All rights reserved.
Disclaimer: You alone are responsible for your investment decisions. See Privacy/Disclaimer for more information.
Some pattern names are registered trademarks of their respective owners.

Home Advertise Contact Donate Links Privacy/Disclaimer