Break and retest strategy: what a retest is in trading

A breakout is price clearing a support or resistance level. The retest comes afterwards, when price returns to touch the level it has just cleared. It is a simple setup to describe, a hard one to trade cleanly, and above all an easy one to confuse with a failed breakout.

• A retest is price returning to a level it has just broken, before moving on in the direction of the break.
• The level swaps roles: broken resistance becomes support, broken support becomes resistance.
• Nothing tells you in advance whether a retest will hold or the breakout will fail, except context and price behaviour.
• The return is not guaranteed: on US stocks, Thomas Bulkowski records it after 58% of chart pattern breakouts, not after all of them.

Break and retest is a classic of technical analysis. Its logic is clear: a level that capped price for weeks can become a floor once cleared. StockCharts' ChartSchool, on its support and resistance page, puts it this way: the breakout shows that demand has overwhelmed supply, and if price returns to that level, demand is likely to increase there. But between the logic and the execution sits everything that makes this job difficult.

This article describes the setup, what validates it, what invalidates it, and above all how to measure it on your own trades rather than judge it on an impression. No promise of results: it is a setup, not a recommendation.

What is a retest in trading?

A retest is the move by which price returns to touch a level it has just cleared, before continuing in the direction of the break. If resistance at 100 breaks upwards, the retest is price coming back towards 100, then bouncing higher. The level has not moved, it has changed role. Retesting in trading happens on every timeframe, from a five-minute chart to a weekly one, and the logic does not change. Thomas Bulkowski's Throwbacks page on ThePatternSite calls this return a throwback after an upward breakout, and the Pullbacks page on ThePatternSite calls it a pullback after a downward one.

That role swap is the heart of the idea, and its classic explanation is short. While 100 was resistance, sellers were active there. Once cleared, those who sold the level are in trouble and those who missed the break are waiting for a way in. Both groups may act in the same place. In the currency market, a study by Carol Osler for the Federal Reserve Bank of New York (Staff Report No. 125) examined the stop-loss and take-profit orders of a large dealing bank: they cluster strongly at round numbers, which are often used as support and resistance levels.

Not every level comes from a past high or low. Pivot points are calculated each morning from the previous day's high, low and close, and many platforms plot them by default: the lesson on pivot points gives their formulas and their limits.

What is a retest in swing trading? Exactly the same mechanic as on short timeframes, with one practical difference: it can take several sessions to appear, or never appear at all. On US stocks since 2000, Bulkowski counts a throwback after 58% of 10,305 upward chart pattern breakouts, inverse head and shoulders included, and a pullback after 58% of 8,765 downward ones, within a month of the breakout by convention. The round trip takes ten days on average in the first case and eleven in the second. So roughly four breakouts in ten do not come back in that window: that is the first source of frustration, and the first reason the setup gets missed.

The break and retest strategy, step by step

First, identify a level that matters: a high or low touched several times, not a line drawn at random. A level tested several times is better established, and its break signals that a balance has given way. Some traders start from a supply and demand zone instead, a price range left by a decisive move, or from a VCP pattern, whose successive contractions tighten price before the break: the rest of the method does not change.

Second, let the break happen without entering. This is the hardest part to hold to: entering once the breakout is under way means buying a price that has already moved away from the level, so accepting a wider stop for an invalidation that sits in the same place.

Third, wait for the return. Fourth, only take the trade if price reacts at the level rather than slicing through it as if nothing were there. That reaction, in whatever form, is the only observable difference between a retest and a breakout that fails. When that return settles into a small channel tilted against the move, it is a flag pattern, and the break of the channel trades with the same stop rules.

What separates a real retest from a false break

Honestly: nothing, in advance. A failing breakout looks identical to a retest right up to the moment it fails. That is why a stop is mandatory on this setup, and why it belongs on the other side of the level, where the thesis is invalidated.

Bulkowski's counts on US stocks give an order of magnitude. Once price has returned to the level after an upward breakout, it recovers 65% of the time and keeps dropping below the chart pattern 35% of the time. After a downward breakout, it resumes the decline only 47% of the time and continues rising above the pattern 53% of the time. A return to the level is therefore far from a safe entry.

What you can check before entering is the wider context: does it run with the break or against it? A bullish retest inside a broader downtrend is a bet against the current. Higher-timeframe market structure, which the structure and liquidity lesson covers, helps settle that question.

The second checkable element is how price broke. A slow, hesitant clearance with a tiny candle body does not say the same thing as a decisive one. It is not a guarantee, and volume does not settle it either: Bulkowski finds that an upward breakout on volume above its 30-day average throws back 70% of the time, more often than breakouts as a whole. It is one more piece of information, and it goes in the journal to be measured later.

Where the stop goes, and why that is the whole point

The entire appeal of waiting for the retest sits in the stop. Entering on the return to the level lets you place the stop just behind that level, far closer than entering on the break. At equal risk in currency, that means a larger position and a better reward-to-risk ratio for the same target.

It is also what makes the setup measurable. If you record your entry, your stop and your target, the difference between retest and break becomes a number rather than an intuition. Position sizing then works off a short, known distance instead of an estimate.

Measuring whether this setup works for you

A setup is judged neither on a video nor on ten trades. It is judged on a large enough population, with the cases separated: retests taken with the higher-timeframe trend against those taken against it, retests on a three-touch level against a single-touch one, retests that happened same-day against those that took several sessions.

That is exactly what a journal allows and memory does not. Tagging each trade with its setup, then comparing sub-populations, turns "I think break and retest works well" into a measurement with a sample size. And the answer is sometimes: it works, but only in one of the two contexts.

Frequently asked questions

Frequently asked questions