|
Written by and copyright © 2005-2018 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 the registered trademarks of their respective owners.
My book,
Getting Started in Chart Patterns, Second Edition  ,
has an entire chapter on support and resistance, starting on page 35. I picture a copy of the book on the left.
If you click on this link and then buy the book (or anything) at Amazon.com, the referral will help support this site. Thanks. -- Tom Bulkowski
$ $ $
New Research added 7/20/2018
What happens to performance when overhead resistance disappears? A study I conducted answers this question using thousands of chart patterns.
Overhead Resistance: Summary
New (2018) research shows that stocks with overhead resistance perform better than do those without overhead resistance.
A contradiction? Maybe. The two studies use different methods and time periods. In one test, analyzing data from the below study comes to the same conclusion as the other study, that stocks
showing overhead resistance outperform those not showing resistance. Click the link for more information.
Test 1: When stocks climb to new highs and are met by overhead resistance at (or above) the ultimate high, they averaged gains of 29%. When stocks broke out to new all-time highs (no overhead resistance at the ultimate high, but resistance might have been present after the breakout and perhaps quite near the ultimate high), the stocks averaged gains of 65%.
Test 2: A second test of resistance/no resistance at the top of the chart pattern shows startling performance differences between stocks with overhead resistance and those breaking out to new all-time highs. This test looks
at the effects of nearby overhead resistance, if any. The below table shows the results.
Performance of Stocks Over Time with and without Overhead Resistance
| Resistance? | 1 month | 3 months | 6 months | 1 year | 2 years | 3 years | 4 years | Unlimited |
| Resistance | 18% | 25% | 29% | 33% | 37% | 39% | 40% | 45% |
| No resistance | 49% | 52% | 54% | 56% | 57% | 58% | 57% | 40% |
For example, stocks that did not show overhead resistance above the top of the chart pattern during the prior month (or no prior resistance regardless of the hold time) averaged gains of 49%. Those with overhead resistance within the
prior month climbed just 18%.
In 2018, I added a test to the spreadsheet based on overhead resistance or no resistance without any time limit. That's the "Unlimited" column on the far left. Notice that this shows
stocks with resistance setup by prior price movement outperformed those without resistance. This result is the opposite of the others.
Overhead Resistance: Methodology Test 1
I found a new way to test the stopping power of overhead resistance. It uses 16,530 chart patterns found in 1,263 stocks from May 1988 to June 2014.
I only included chart patterns with upward breakouts in a bull market.
Here is the list of the 28 types of chart patterns included in the study.
- Broadening tops and bottoms
- Right-angled ascending and descending broadening tops
- Ascending and descending broadening wedges
- Cup with handle
- Diamond tops and bottoms
- All four varieties of Adam and Eve double bottoms
- Simple and complex head-and-shoulders bottoms
- Rectangle tops and bottoms
- Rising wedge
- Rounding tops and bottoms
- Ascending scallop and inverted and ascending scallop
- Ascending, descending, and symmetrical triangles
- Triple bottoms
- Big W
- Ugly double bottom
For each chart pattern, I found the breakout and ultimate high. A breakout occurs when price closes above the chart pattern, closes above a neckline, or closes outside of a trendline (depending on the type of
chart pattern). The ultimate high is
the highest high before the primary uptrend reverses and either drops by at least 20% or closes below the bottom of the chart pattern.
I looked back in the historical price series to see if the stock broke out to a new all-time high (a high price never before reached) when the stock peaked at the ultimate high.
For example, suppose Quibble is a new issue that's been trading for a month. The highest the stock has climbed is $10. A broadening top appears and price breaks out upward at $9. The stock continues higher
and rises to $15 before it reverses and drops back to $10. The ultimate high is $15. I looked back in time to see if the stock had ever reached $15 before. In this case, it had not.
I looked for resistance (where the stock previously had been at that price) or lack of resistance (a new all-time high), and then logged the gain from the breakout to the ultimate high.
In the case of Quibble, the stock was making new all-time highs so there was no overhead resistance at the ultimate high ($15, but it did have to plow through resistance until passing above $10).
The percentage increase (67%) between the breakout ($9) and ultimate high ($15) was logged for that
chart pattern.
Overhead Resistance: Methodology Test 2
This test used the same chart patterns and data as the previous test except that it looked for prior prices not at the ultimate high, but at the top of the chart pattern.
This time, however, I looked for prior overhead resistance within a month, 3 and 6 months, and yearly up to 4 years before the top of the chart pattern (or no resistance at all above the chart pattern).
A 2018 test extended this test to have no minimum time to find resistance. It was either present or not when price moved above the top of the chart pattern.
For example, suppose Quibble stock had an upward breakout from a broadening top and the top of the chart pattern peaked at $9 on August 1. The stock continued moving higher until it hit $15 (the ultimate high).
I looked for the same or higher prices in the 31 days before August 1. If it had some, then the stock had overhead resistance to plow through. If it did not, then it was resistance free.
I used the same methodology for 3 months, 6 months, 1, 2, 3, and 4 years before the top of the chart pattern.
Once I had collected my data, I measured the move from the breakout to the ultimate high (in the Quibble example, that is the difference between $9 and $15 since a breakout from a broadening top occurs
at the pattern's high) and sorted the results according to resistance or no resistance for each period studied.
This test helped remove the effects of overhead resistance disappearing because of missing prior price data. See the warnings below.
Overhead Resistance: Warnings
Be advised that not all stocks had data going back to their original issue. The effect of this is unknown but is expected to be minor due to the effect of stock splits and the general rise of stocks over long
periods of time. The effects of dividends, mergers, splits, commissions and so on were not included. I only looked at price.
Do not expect your trades to match these results. These results are for thousands of perfect trades, trades that enter at the breakout price and sell at the ultimate high, the highest high before the trend changes.
Instead, use the results for comparison purposes and to help formulate trading strategies.
Overhead Resistance: Results Test 1
Stocks with overhead resistance (the stock at the date of the ultimate high had price reach that level sometime in the past) had gains that averaged 29%.
Stocks with no overhead resistance (the stock broke out to new all-time highs) had gains that averaged 65%. That is more than double the result of stocks with overhead resistance.
Overhead Resistance: Results Test 2
The following table shows the performance results of stocks with and without overhead resistance sampled at various intervals.
For example, stocks with no overhead resistance within one month before the top of the chart pattern (or no resistance at all, regardless of the time period) showed gains to the
ultimate high that averaged 49%. Stocks with overhead resistance had gains of 18%.
Performance of Stocks Over Time with and without Overhead Resistance
| Resistance? | 1 month | 3 months | 6 months | 1 year | 2 years | 3 years | 4 years | Unlimited |
| Resistance | 18% | 25% | 29% | 33% | 37% | 39% | 40% | 45% |
| No resistance | 49% | 52% | 54% | 56% | 57% | 58% | 57% | 40% |
In 2018, I added a test to the spreadsheet based on overhead resistance or no resistance without any time limit. That's the "Unlimited" column on the far left. Notice that this shows
stocks with resistance setup by prior price movement outperformed those without resistance. This result is the opposite of the others, but the results agree with a different test, using
a different method. See the next section.
Overhead Resistance: New Research
In a July 2018 test, I looked at stock prices and round numbers ($10, $20, $30...$200), not chart patterns. I found that performance improved when overhead resistance was present.
That finding appears to contradict the results of the study described in this article. Here are the differences between the two methods.
| This article | July 2018 test |
| I reworked the spreadsheet data to exclude the 4-year limit but it measures overhead resistance not at the ultimate high (which may be below the top of the chart pattern such as in a
symmetrical triangle) but at the top of the chart pattern. The results agree with the other study. Stocks showing overhead resistance had gains of 45% (12,863 samples). Stocks
with no overhead resistance showed gains of 40% (2,127 samples). | Conclusion agrees |
| Limited the look-back for resistance to 4 years | No limit |
| Used chart patterns | No chart patterns used. Used price in multiples of $10, from $10 to $200 |
| Includes all prices | Included trades which crossed a decade (10) number from $10 to $200 |
| Sample counts are high (up to 11,431). The 4-year study had 1,432 samples showing no overhead resistance | Few samples, from 12 to 519 for stocks with no overhead resistance. |
| Possibly outdated (using data from 1988 to 2014) | Current: July 2018 (using data from 1990 to 2018) |
| Stock data may not include its entire existence (making it harder to check for no resistance) | Same |
| Performance measures to the ultimate high which can be a day to years away. The median time to the ultimate high is 81 days (about 2.5 months), but the average is 182 (about 6 months). | Measures discrete intervals: 1, 2, 3 and 6 weeks, 1 and 2 months, so duration is limited. |
-- Thomas Bulkowski
Written by and copyright © 2005-2018 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 the registered trademarks of their respective owners.
An apple a day will keep anyone away if you throw it hard enough.
|
If you buy ANYTHING, they pay for the referral.
on 
