The single most transferable setup in trading. It works on every market and every timeframe because it is built on the most reliable behaviour there is: a broken level changes role, and the traders caught on the wrong side create the orders that defend it.
Why it works
When resistance breaks, three groups act at once on the retest. Traders who shorted the level are underwater and want out at break-even. Traders who missed the break are waiting for a second chance. Traders who bought the break want to add. All three place orders at roughly the same price, in the same direction.
That is why the retest so often holds — not because the line is special, but because the line is obvious, and obviousness concentrates orders.
The breakout tells you the level changed hands. The retest tells you the new owner intends to keep it.
Market conditions required
- A level that is genuinely obvious — two to four clean touches, visible on the daily.
- A decisive break, meaning a candle body closing beyond the level, not a wick.
- Momentum on the break. Expanding range, and higher volume where volume is available.
- Patience for the return. Perhaps half of all breaks retest. The other half are simply missed, and that is fine.
Higher-timeframe bias
Retests aligned with the higher-timeframe trend are substantially more reliable. A daily downtrend breaking a small intraday resistance and retesting it is a weaker long than the same pattern inside a daily uptrend.
The setup
- 1Mark levels before the session
Prior day high and low, weekly high and low, obvious swing points, round numbers. Levels drawn after price arrives are contaminated by what you want to happen.
- 2Wait for a body close beyond one
Not a touch, not a wick. A close.
- 3Let price come back
Do not chase. The whole point of this strategy is a better entry than the breakout traders got.
- 4Demand a reaction at the level
A rejection candle, a small consolidation that holds, or a lower-timeframe structure break in the new direction.
Entry rules
- Price closes beyond the level on your trading timeframe.
- Price returns to within roughly 0.25× ATR of the level.
- A rejection candle forms — a wick into the level, closing back in the breakout direction.
- Enter on that close. A resting limit order at the level fills better but risks catching a level that fails outright.
Stop-loss rules
Beyond the level, plus a buffer of roughly 0.3× ATR. The logic is exact: if price closes back through the level, the flip did not happen and the break was false. That is the cleanest invalidation in trading — the idea is wrong, not merely unlucky.
Take-profit rules
| Target | Where | Notes |
|---|---|---|
| T1 | Next marked level | Take partial; levels are where moves pause |
| T2 | Measured move — the height of the prior range added to the break | The core target |
| Runner | Trail behind swing points | Only when the higher timeframe is aligned |
Risk management
- This setup produces tight stops, which tempts oversizing. Keep risk at a fixed 1% and let the tight stop buy you a bigger reward multiple, not a bigger position.
- If the retest never comes, take no trade. Missing trades costs nothing.
- Two failed retests on the same level means the level is no longer meaningful — stop trading it.
Example winning trade
XAU/USD, 15-minute. Resistance at 2,384 tested three times over two days. At 13:45 a 15-minute candle closes at 2,391, clearly beyond. Price drifts back and at 15:30 wicks to 2,385 before closing at 2,390.
Long 2,390, stop 2,381 below the level plus buffer, nine dollars of risk. T1 at the next level 2,404 hit for +1.5R; measured move target 2,412 reached the following session for roughly +2.4R on the remainder.
Example losing trade
EUR/USD, 15-minute. Support at 1.0820 breaks with a close at 1.0812. Price retests to 1.0819 and prints a rejection candle. Short entry 1.0814, stop 1.0827.
Price grinds back up through 1.0820, closes above it, and the stop is hit for −1R. The break had happened on unremarkable range and no volume expansion — it was a drift through the level rather than a decisive break, and the retest simply resumed the prior range.
When NOT to trade it
- The break was a wick, not a close.
- The level had been tested six or seven times — heavily-tested levels are weakening, not strengthening.
- Price returns to the level but slices straight through with no reaction.
- The retest takes so long that the context has changed — a retest three days later is a new trade, not this one.
Common mistakes
- 1Entering at the touch rather than the reaction
Price reaching the level is not the signal. The reaction to it is.
- 2Drawing the level to fit the trade
Mark levels before you have a position or an opinion about direction.
- 3Treating every break as tradeable
Without momentum on the break, the retest is just a range boundary being revisited.
- 4Moving the stop through the level
The level is the thesis. Below it, there is no trade left to defend.
Backtesting considerations
- Define “decisive break” numerically — for example a close at least 0.3× ATR beyond the level — or hindsight will select only the breaks that worked.
- Record how often a retest occurs at all; that rate determines how many opportunities the strategy actually produces.
- Compare limit-at-level entries against rejection-candle entries; the trade-off between fill rate and win rate is the main tuning decision here.