VCP: breakout from a tight base
This page documents a mechanical scan for volatility contraction patterns: a rising moving-average stack, relative strength, shrinking swings and a volume dry-up into a pivot. It is not a buy list. A later product will mail new setups when the scan finds them. The figures below are a backtest of that scan from 2016 through mid-August 2026.
US$337,037
12.1% per year
543 setups
What this backtest is
Lead slice: setup quality at least 3, ten stocks, one percent risk, no market-exposure throttle. Window 2016-01-01 to 2026-08-17, run on 17 August 2026. Starting capital is US dollars. 2026 is a partial year.
What the scan requires
Every stock must pass the trend filter first. Only then does the contraction pattern count, and only then is a quality level assigned.
Trend filter
Close above the 50-, 150- and 200-day averages, 150 above 200, 50 above both, and the 200-day average rising. Price at least 30 percent above the 52-week low and no more than 25 percent off the 52-week high. Relative strength at least 70 versus the rest of the universe that day.
Tight base into a pivot
Two to six contractions, each smaller than the last. The final swing is at most 10 percent. Volume dries up into the pivot. The pattern spans 15 to 325 sessions and the pivot is at most 25 sessions old.
Quality levels 1 to 4
Higher levels demand quieter recent volume, a closer stop, a younger pivot and stronger breakout volume. Level 3 is the lead slice because it still has hundreds of cases. Level 4 is tighter and, in this run, even stronger on fewer stocks.
Ten-slot portfolio by minimum quality
Same portfolio rules at every row: ten slots, one percent risk, 25 percent max weight, five basis points of slippage, trailing exit after a gain. Tighter minimum quality means fewer fills and, in this sample, higher equity.
| Min. level | Setups | Fills | End equity | Return | Max. drawdown | Profit factor |
|---|---|---|---|---|---|---|
| ≥ 1 | 2342 | 618 | US$186,575 | +87% | 50.3% | 1.31 |
| ≥ 2 | 1212 | 542 | US$176,921 | +77% | 43.7% | 1.34 |
| ≥ 3 | 543 | 380 | US$337,037 | +237% | 29.6% | 1.78 |
| ≥ 4 | 284 | 243 | US$345,762 | +246% | 19.5% | 2.14 |
The highlighted row is the lead slice. Level 4 reached a similar equity with a shallower drawdown and fewer trades. Past portfolio results are not a forecast.
Each setup on its own, by year
This table ignores slot limits. Every completed level-3-or-better setup is one row, so a weak year cannot hide inside a full portfolio. Alpha is versus ACWI over the same hold.
| Year | Setups | Win rate | Avg. return | Avg. alpha | Profit factor |
|---|---|---|---|---|---|
| 2016 | 36 | 39% | +15.2% | +12.5% | 4.69 |
| 2017 | 60 | 32% | +2.9% | −0.8% | 1.62 |
| 2018 | 57 | 21% | −1.2% | −1.2% | 0.78 |
| 2019 | 33 | 27% | −1.2% | −1.3% | 0.76 |
| 2020 | 48 | 44% | +7.9% | +4.7% | 3.15 |
| 2021 | 76 | 32% | +1.5% | +0.7% | 1.30 |
| 2022 | 14 | 7% | −4.2% | −0.5% | 0.26 |
| 2023 | 56 | 32% | +2.9% | +0.8% | 1.62 |
| 2024 | 77 | 38% | +1.3% | +0.4% | 1.28 |
| 2025 | 54 | 43% | +3.1% | +1.3% | 1.78 |
| 2026 | 30 | 23% | −2.2% | −3.7% | 0.56 |
2022 is the stress year: few setups and a poor payoff. 2026 stops in August. Win rate sits around one in three because losers are cut; the edge is in the size of the winners.
Where the level-3 setups sat
Counts only, no live stocks. Historical tickers are not the product. A later mail will carry new setups when they appear.
| Exchange | Setups |
|---|---|
| LSE | 207 |
| NASDAQ | 164 |
| NYSE | 93 |
| XETRA | 19 |
| MI | 12 |
| PA | 11 |
| OL | 8 |
| SW | 7 |
| MC | 7 |
| CO | 7 |
| AS | 6 |
| HE | 2 |
Mail when a setup appears
The live product is the scan, not this table. When a stock meets the rules, a later package will send the setup by mail. This page is the method and the historical flavour.
Frequently asked questions
What does VCP mean here?
A volatility contraction pattern: swings get smaller, volume dries up, then price leaves a well-defined pivot. We scan that structure with explicit numeric rules. This page is the scan, not a person and not a course.
Is this a recommendation to buy?
No. The scan marks a price structure. Position size, stops and whether you trade at all stay with you. It does not replace advice from a securities advisor or a bank.
Why is the win rate only about a third?
Stops take the losers. The backtest edge, where it exists, is a larger average winner than loser, not a high hit rate. A third is the expected shape, not a defect in the sample.
Will you send the stocks?
That is the planned product: live scan, mail on a new setup. It is not on this page yet. We do not publish today's open stocks here.
Limits
This is not investment advice and not an offer to buy or sell securities. The backtest uses one parameter set, five basis points of slippage and no taxes or borrow. A 30 percent drawdown happened in-sample. 2022 and the unfinished 2026 year belong in the record. Your results will differ.
PWP Leeway UG (haftungsbeschränkt)
