As of 04/24/2024
Indus: 38,461 -42.77 -0.1%
Trans: 15,078 -358.61 -2.3%
Utils: 893 +6.79 +0.8%
Nasdaq: 15,713 +16.11 +0.1%
S&P 500: 5,072 +1.08 +0.0%
|
YTD
+2.0%
-5.2%
+1.3%
+4.7%
+6.3%
|
 39,800 or 37,150 by 05/01/202416,200 or 15,000 by 05/01/2024885 or 850 by 05/01/202416,700 or 15,800 by 05/01/20245,250 or 5,025 by 05/01/2024 |
As of 04/24/2024
Indus: 38,461 -42.77 -0.1%
Trans: 15,078 -358.61 -2.3%
Utils: 893 +6.79 +0.8%
Nasdaq: 15,713 +16.11 +0.1%
S&P 500: 5,072 +1.08 +0.0%
|
YTD
+2.0%
-5.2%
+1.3%
+4.7%
+6.3%
| |
|
39,800 or 37,150 by 05/01/202416,200 or 15,000 by 05/01/2024885 or 850 by 05/01/202416,700 or 15,800 by 05/01/20245,250 or 5,025 by 05/01/2024 | ||
In my book,
Swing and Day Trading
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.
$ $ $
This article discusses Joseph Stowell's cups, caps and the three bar net line as a position trading setup.
This trading setup results in some good performance numbers on the weekly scale (16.2% to 24.5% average gain per trade) but trades are few. Drawdowns can be a problem (13% to 24%) if you're not willing to hold onto the position until a sell signal appears. That's typical for systems that use weekly data.
I included a simple moving average to eliminate losing trades that occurred in a bear market. The 15 and 20-week SMAs seemed to work best, but the 10 week SMA also did well.
Be sure to test this setup yourself since implementation can be difficult. You may want to test only the three-bar net line for entry and exit signals instead of the complete system with cups and caps. In one test I ran, it showed promise.
Technical Analysis of Stocks & Commodities magazine interviewed Joseph Stowell in the July 1995 issue and included a sidebar on the three bar net line, cups, and caps. He says, "I use what I call a three-bar net line that I use to identify the medium-term trend in the bond market. It's an excellent device to use on a monthly stock chart." He goes on to explain that when the monthly price retraces back to the three-bar net line, "it's an excellent place to purchase [the stock]."
I decided to test the three-bar net line along with his cups and caps entry and exit signals on the daily and weekly charts, not the monthly scale.
I used 559 stocks and 104 exchange traded funds from March 12, 2001 to October 1, 2010. Those two dates have the S&P 500 index beginning and ending near the same price. In between those two dates, the market experienced two bull and two bear markets. Not all securities covered the entire range. I also made a few runs with data going back to 1990.
I used the following rules.
Begin looking for the setup by determining the price trend. Use higher highs and higher lows for an up trend and lower highs and lower lows for a down trend.
Once you have determined the initial trend in your data series, look for the three-bar net line. The figure shows a hypothetical example.
The image on the right is for price up trends. Here are the rules for up trends.
The next image on the right is for price down trends. Here are the rules.
These two patterns determine the price trend leading to cups and caps.
After determining the trend using the three-bar net line as described above, then look for a cup or cap trading signal. The cup and cap patterns are highlighted in red boxes in the figure.
Cups appear in an up trend and I show an example of that in the figure. Here are the rules for a cup sell signal.
Caps follow similar rules, but issue buy signals.
The setup to buy a stock or exchange traded fund followed these rules and they depend on the three-bar net line and cap pattern as described above.
The sell signal is given by these rules and it also includes the three-bar net line and cup pattern as described above.
* The rule regarding the cup or cap appearing at the end of a trend can be difficult to automate for testing, but do your best. My method is somewhat complicated. I found the highest high or lowest low before the three-bar net line to the cup or cap. You may want to be sure that the cup or cap is the lowest low or highest high of the past x bars for simplicity.
The following table shows all of the tests that I ran using this setup, sorted by the average gain (second column on the left).
| Win Loss | Avg Gain | Avg Max Drawdown | Hold Time Loss | S&P | Hold Time Days | Trades | SMA | Description |
| 41% | 1.0% | 5.8% | 3.7% | -1.1% | 40 | 1768 | N/A | 88 long only ETFs with stop below cap, daily |
| 58% | 1.7% | 10.5% | 8.6% | -1.7% | 85 | 1435 | N/A | 104 ETFs, daily |
| 60% | 2.0% | 9.9% | 8.2% | -1.5% | 84 | 1306 | N/A | 88 long only ETFs, daily |
| 40% | 2.7% | 10.0% | 6.1% | -1.3% | 52 | 9468 | N/A | 559 stocks with stop below cap, daily |
| 64% | 3.1% | 8.1% | 6.5% | -0.9% | 82 | 907 | 100 day | 88 long only ETFs, daily |
| 61% | 5.3% | 15.8% | 12.3% | -1.8% | 118 | 4373 | 50 day | 559 stocks, daily |
| 61% | 5.3% | 16.8% | 13.0% | -1.7% | 120 | 6736 | N/A | 559 stocks, daily |
| 74% | 16.2% | 17.5% | 13.6% | 1.4% | 409 | 207 | 10 week | 104 ETFs from 1990, weekly |
| 68% | 17.3% | 24.4% | 20.2% | 0.7% | 414 | 1139 | 10 week | 559 stocks, weekly |
| 81% | 21.4% | 14.3% | 12.1% | 5.1% | 417 | 138 | 10 week | 88 long only ETFs, weekly |
| 84% | 23.8% | 13.7% | 10.3% | 5.8% | 412 | 127 | 20 week | 88 long only ETFs, weekly |
| 83% | 24.5% | 13.9% | 10.9% | 6.7% | 431 | 126 | 15 week | 88 long only ETFs, weekly |
The first test includes a stop loss placed below the cap (the cap signals a buy in a down trend). The reason for this test is that there were a number of large losses and I wanted to see if using a stop would cut their size. The average gain was the worst of the bunch, but it did cut the maximum drawdown per trade, averaged over all trades (shown as Avg Max Drawdown) and it also cut the hold time loss.
The hold time loss is how far price drops below the buy price during the trade, averaged over all trades.
The daily or weekly phrases in the description refer to daily or weekly price data used in the test.
Looking at the some of the test results, I decided to add a moving average to filter out trades in a bear market. I tried various values as the table shows. Including a moving average tended to boost the win/loss ratio and the average gain, but the drawdown and hold time loss also increased.
Notice that the higher the average gain, the fewer trades completed. Thus, don't look for huge gains or high win/loss ratios with the setup when using the 10-week to 20-week simple moving averages (SMA).
Since the trade count was so low, I tested the same setup using 559 stocks (4th row from the bottom versus 3rd row from bottom) and found that the win/loss ratio dropped from 81% to 68% and the average gain dropped from 21.4% to 17.3% using almost ten times the number of trades. Thus, the larger trade count gives more realistic results.
If you realize that a 10-week moving average is the same as a 50 day SMA, then you get four times the number of trades. The win/loss ratio remains close (68% versus 61%) but the average gain plummets from 17.3% to 5.3%. The drawdown drops (which is good), however. Thus, this setup works best on the weekly scale (using weekly data) providing you can keep the drawdown in check.
Adjusting some of the setup conditions may improve performance and increase the number of trades. You might use only the three-bar net line as an entry and exit signal and forget about cups and caps.
As an example of the cup and cap system, I show a picture of the Dow Utilities on the weekly scale. At point A begins the red box inset which shows the three bar net line for up and downtrends. Points 1, 2, and 3 correspond to the three-bar net line in an uptrend. However, price closes below the three-bar net line, flipping the trend to downward. That happens at the phrase "Trend Down" on the chart.
Following that, points 4, 5, and 6 show a three-bar net line that confirms a trend change from down to up. That occurs at "Trend Up."
A cap pattern appears at the bottom of the down trend, shown in the blue box inset. Thus, we have all of the ingredients of a buy signal except for the 10-period SMA. If you draw in the SMA, you'll find that price is above the SMA, so an entry signal is given on 11/30/2009 at 379.92. I show that as the green line.
For the exit, I only show the cup pattern, but the three-bar net line is the same as the red bar. You can see the three higher lows clustered together in the three days before the sell signal. The sell signal occurs on 11/15/2010 at 398.52 for a gain of $18.60 per share. That's not much considering a drawdown of 12.6% during the trade and a hold time loss of 8.7%.
-- Thomas Bulkowski
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.
Legal notice for paid links: "As an Amazon Associate I earn from qualifying purchases."
|
My Stock Market Books
|
My Novels
|
You never know how many friends you have until you rent a house on the beach.
on 
