A retest is price returning to a level it has just broken, to see whether the level now holds from the other side. Broken resistance becomes support; broken support becomes resistance. A retest gives you a tighter stop and a clearer invalidation than entering on the break itself.
It is the least glamorous entry in trading and probably the most useful one. Nothing about a retest is exciting: the move has already happened, the candle you wanted has already printed, and you are buying something that is briefly going the wrong way. That discomfort is precisely what you are being paid for.
What actually happens during a retest
Price breaks above a level at 100 and runs to 102. Then it stops and drifts back down. It reaches 100 again — the same price it just broke through — and this time it is bought, and it turns back up. That return trip is the retest.
The sequence has four parts, and skipping any of them changes the trade:
- A level that existed beforehand. Marked before the session, defended at least once. See how to find levels that are actually real.
- A confirmed break. A close beyond the level, not a wick through it. That is the subject of break and hold, and it has to happen first.
- A return to the level. Price comes back to roughly the broken price. Roughly, because levels are zones, not lines.
- A rejection at it. The level holds from the new side and price turns away from it. Without this the retest has not passed; it is just price going back where it came from.
Why a broken level flips sides
The polarity principle — old resistance becoming support and vice versa — is usually stated as if it were a law of nature. It is not. It is a description of what happens to the orders and the positions around a level once it gives way.
Think about who is at 100 after the break. Sellers who were defending it have been stopped out or are now holding losing positions and would dearly like to exit at breakeven, which means buying back at 100. Buyers who watched the break happen without them want a second chance at the same price. And traders who entered on the break have their stops just below 100, so they have a direct interest in defending it. Three different groups, all with orders in the same small zone, all pointing the same way.
That is the whole mechanism, and it explains the exceptions too. When a level breaks on genuinely overwhelming flow — a scheduled release, a large institutional order — there may be no meaningful group of trapped sellers, no orderly pullback, and no retest at all. The principle depends on the level having been contested. Levels that were never really defended do not flip, because there was nobody there to change their mind.
Valid retest versus failed retest
Not every return to a level is a retest that passes. The difference is visible, and it is worth being strict.
| Valid retest | Failed retest | |
|---|---|---|
| How price arrives | Drifts back slowly, on shrinking candles | Comes back fast, on a large candle in one move |
| What it does at the level | Touches the zone and turns; wick into it, body above it | Trades through and closes on the far side |
| Time spent there | Brief — a candle or two | Sits at the level for many candles, grinding |
| What it tells you | The break was real; sellers are gone | The break was a spike; the level never changed hands |
The row that carries the most information is the third one. A level that flips cleanly does so quickly — price touches it, finds buyers, and leaves. A level that price sits on for twenty minutes is not being defended by anyone; it is being negotiated, and the original break is losing credibility with every candle that fails to move away from it.
Where the stop belongs
This is the concrete reason retest entries are worth the wait. The stop placement is not a judgement call — it falls out of the trade's own logic.
The premise of a retest long is that the broken level now holds as support. So the invalidation is: it does not hold. That means a stop below the retest low, beneath the wick rather than the body, with a small buffer for spread. If price closes back through the level, the reason you took the trade has been disproved and you should already be out.
Compare the alternative. Entering on the break at 100.40 with a stop below the level at 99.90 gives you 50 ticks of risk. Entering on the retest at 100.05 with a stop at 99.85 gives you 20 ticks — and the invalidation is cleaner, because it sits behind an actual rejection that just occurred rather than behind a level you are hoping will matter. At a fixed risk per trade, that difference is not academic: it is the position size you can take, and it is the difference between a target being worth 1.5R and 3R. The mechanics are in how to set a stop loss that isn't just a guess.
The honest cost, and a caution about entry refinements
Two things need saying that most retest articles leave out.
First, the retest often does not come. Strong moves leave levels and never return. If your rule is to enter only on a retest, you will miss those, and they include some of the largest moves of the year. There is no version of this method that both waits for the pullback and catches the runaway. Pick the trade-off deliberately: fewer trades, tighter stops, higher reward-to-risk on the ones you take, and the certainty of watching some moves go without you.
Second, a better entry is not the same thing as an edge. This matters because it is easy to convince yourself that refining an entry has created something that was not there. When Sullivan, Timmermann and White re-examined the technical trading literature, they expanded the earlier universe of 26 trading rules to 7,846 and applied them to 100 years of daily Dow Jones data using a bootstrap reality check that corrects for data-snooping. Their conclusion was that once you account for the number of rules searched, no simple technical trading rule stood up (Sullivan, Timmermann & White, "Data-Snooping, Technical Trading Rule Performance, and the Bootstrap", Journal of Finance 54(5), 1999, pp. 1647–1691).
That is not an argument against retests. It is an argument about what a retest is for. It improves your risk on a trade you had already decided to take — a smaller stop for the same idea. It does not turn a bad idea into a good one, and adding a third and fourth entry condition on top will not either. If the level was wrong, entering at it more precisely just makes you wrong more cheaply.
Frequently Asked Questions
What is a retest in trading?
A retest is price returning to a level it has just broken, to establish whether that level now holds from the other side. If price broke above resistance and then pulls back to that same price and is bought, the old resistance is behaving as support. The retest is the market's own check on whether the break was genuine, and it happens without you having to guess.
Why does broken resistance become support?
Because the orders and the positions around the level change sides when it breaks. Sellers who defended the level have been taken out or are now underwater and want out at breakeven. Buyers who missed the break want a second entry at the same price. Both groups place orders in the same zone, and their combined interest is what turns an old ceiling into a floor.
Where do you put the stop on a retest entry?
Below the retest low for a long, and specifically below the wick rather than the body, with a small buffer for spread and noise. The logic is definitional: the entire premise of the trade is that the broken level now holds. If price closes back through it, that premise is dead and the position should already be closed. The retest is what makes this stop tight rather than arbitrary.
What if the retest never comes?
Then you do not take the trade, and that is the cost of the method rather than a failure of it. Strong moves frequently leave a level and never return, and waiting for a pullback means missing those entirely. This is a real trade-off: retest entries give a tighter stop and a clearer invalidation, at the price of a lower trade count and the fastest moves passing you by.
Bottom line
A retest is the market checking its own work, and it hands you a stop you can defend on logic rather than on preference. Wait for the confirmed break, treat the level as a zone, require an actual rejection rather than a visit, and put the invalidation beneath the wick that just formed. Accept in advance that some moves will leave without you — that is the price of the tighter risk, not a flaw in the method. Start with what a breakout actually is, make sure the break qualified under break and hold, and see how the room's callouts define entry, targets and invalidation before a trade is taken.