Half of this rule is strongly true and half of it is a coin flip. Knowing which half is which is the difference between a patient setup and a textbook-looking trade that works about as often as guessing.
A flip is a level that stops acting as a ceiling and starts acting as a floor once price has closed through it — because the traders trapped above it and the traders who missed the move both want to buy at the same price.
A price acts as resistance for a while: every time the market gets there, it turns back. Then one day it does not — price closes above and keeps going. Later it drifts back down to that same price.
The claim, repeated everywhere, is that the old ceiling now becomes a floor. Traders call this role reversal, a flip, or a flip zone.
There is a genuine mechanism behind it. Traders who sold at the level and got trapped want out at break-even, so they buy back there. Traders who missed the breakout wait for a pullback to that price. Both are buy orders around one price, and this time they are on the other side of it.
A ceiling in a building stops you going up. Knock a hole in it and climb through, and the thing that was blocking you is now the floor you are standing on. Same structure, opposite function, purely because of which side you are on.
The chart version has one important difference from the building: the floor sometimes gives way. That is what the numbers below are about.
Every confirmed swing level that an hourly candle later closed through, in five years of data, followed forward.
| Instrument | Levels broken | Price returned | Of those, held | Of those, failed |
|---|---|---|---|---|
| XAUUSD | 2,073 | 85.6% | 55.9% | 40.2% |
| GBPUSD | 1,966 | 87% | 39.1% | 47% |
| EURUSD | 1,926 | 87.2% | 39.7% | 43.8% |
| US30 | 1,300 | 85.4% | 51.4% | 41.1% |
| NAS100 | 1,394 | 84.1% | 57.5% | 38.1% |
What was counted: A confirmed swing level that an hourly candle later closed through. If price returned to that level within 48 hours, it held when price moved away in the breakout direction without closing back through, and failed when an hourly candle closed back through it.
Two very different findings sit in that table, and they need reading separately.
The first half of the claim is strongly true. Price came back to the broken level around 85.6% of the time, on every instrument. If you break a level and walk away, you will almost always get a second look at it. That alone is worth knowing — it means chasing a breakout is rarely necessary.
The second half is close to a coin flip. On gold the level held 55.9% of the time. On GBPUSD it held only 39.1% and failed 47% — more often than it worked. "Broken resistance becomes support" is not a law. It is a tendency, and on some instruments not even that.
The retest of a broken level feels like the safest trade on the chart. The story is clean, the level is obvious, and the stop can go just below it — which is exactly why so much size ends up there on a setup that works about half the time.
Worse, the stop placed just beyond the flip level is in the most trafficked patch of price available, for the reasons set out in the liquidity sweep lesson. A coin-flip setup with a stop in the busiest place on the chart is a poor combination dressed up as a textbook trade.
A fifty-fifty setup is not useless. It is useless on its own. The way to make it worth taking is to require something else to line up with it.
Kwame has watched GBPUSD fail at 1.3400 three times this week. On Thursday it closes at 1.3435 — a clean break.
What he does not do. He does not buy at 1.3435. Price returns to a broken level around 87% of the time on this instrument, so chasing is paying the worst price for a second-best entry.
What he waits for. Price drifts back to 1.3402 the next morning. Now the textbook says buy the flip.
What he checks first. On GBPUSD, flips held only 39.1% of the time — it fails slightly more often than it works. So the level alone is not a reason. He looks for a second one: is there a demand zone here? Is this also the weekly pivot? Is the first retest?
The two outcomes. If he finds a second reason, he takes it with a stop below the noise, not below the line. If he does not, he skips it — and skipping a textbook-looking setup because the numbers say it is a coin flip is the hardest and most valuable discipline on this page.
Gold and GBPUSD gave meaningfully different answers, which means the number that matters is the one for the market you actually trade. Replay it, mark every broken level, and score the retests. An hour of that beats any general rule.
Test flips in the free backtester →Run the drill in the free backtester → Free, no sign-in to begin.
Does broken resistance always become support?
No. Price returned to the broken level 85.6% of the time, which is the reliable half of the claim. But the level then held only 55.9% of the time on gold, and on GBPUSD it held 39.1% and failed 47% — more often than it worked. Treat it as a tendency, not a rule.
Why does the flip work at all?
Traders who sold at the level and got trapped want out at break-even, so they buy there. Traders who missed the breakout are waiting for a pullback to that price. Both leave buy orders around the same level, this time from the other side of it.
Should I buy the breakout or wait for the retest?
Wait, on this evidence. Around 85.6% of broken levels are revisited within two days, so chasing means paying the worst price in the move for a setup you were very likely to get a second look at anyway.
How do I improve a coin-flip setup?
Require a second, independent reason at the same price — a demand zone, an order block, or a pivot. A flip level with nothing else on it is fifty-fifty; a flip level that is also the weekly pivot and the top of a demand zone is a different situation.
Where should the stop go on a flip trade?
Beyond the instrument's ordinary noise, not two pips past the line. A stop just past an obvious level sits in the busiest patch of price on the chart — see the liquidity sweep lesson — and the Stop Loss Reality Checker will tell you how far past is far enough.
Does it work the same on all instruments?
No, and that is one of the more useful findings here. Gold held around half the time; GBPUSD held under half. The instrument you trade has its own number, and it is worth measuring rather than assuming.
Gold and GBPUSD disagreed by a wide margin. Replay the market you actually trade, mark every broken level, score the retests, and split the results by whether anything else sat at the same price.
Open the free backtester →Beginner exploration
Open each answer for a plain-language way to read When Support Becomes Resistance, test it carefully and decide what to explore next.
This page focuses on “When Support Becomes Resistance”.Role reversal tested across 8,659 broken levels. Price returned 85.6% of the time, but the level held only 55.9% on gold and 39.1% on GBPUSD. A tendency, not a law.For “When Support Becomes Resistance”, a beginner should identify what the learning guide measures, assumes or teaches before acting on its conclusion.Treat this page's account of “When Support Becomes Resistance” as a learning reference rather than a prediction, signal or promise of future performance.
For “When Support Becomes Resistance”, translate the idea into a definition you could apply the same way on two different charts.While exploring “When Support Becomes Resistance”, work through one example slowly and record which inputs or observations determined the result.Keep your “When Support Becomes Resistance” record honest: list the limitation or counterexample before using the concept in a trading plan.Before leaving “When Support Becomes Resistance”, practise the definition on unseen history and review consistency before judging performance.
Turn one idea from “When Support Becomes Resistance” into a rule with explicit inputs, dates, costs and pass-or-fail conditions.Ask AI to expose missing assumptions in that “When Support Becomes Resistance” test, not to guess the next market move.Use the FXAbsolute AI Backtesting Lab to inspect calculations connected to “When Support Becomes Resistance” and the assumptions behind them.Reproduce any important “When Support Becomes Resistance” result and reserve unseen data before deciding that an apparent pattern is useful.
Continue your exploration of When Support Becomes Resistance with the beginner AI prompt guide, or inspect public calculations in the AI Backtesting Lab.