Role Reversal: A Broken Ceiling as a Floor
Role reversal is the idea that a broken resistance level becomes support, and a broken support level becomes resistance. The orders that created the original level are gone once it breaks, so whatever effect remains comes from the level being visible to everyone.
How it works
Price is turned away from a level several times. Eventually it breaks through. Role reversal says that level now works the other way round: the old ceiling becomes a floor.
The traditional story is about participants changing their minds. Sellers who defended the level now regret it and buy the retest; buyers who missed the break get a second chance. It is a tidy narrative and it has one problem.
Breaking the level means those orders were consumed. That is what a break is. The supply that created the ceiling has been bought. Whatever is at that price on the retest is new, placed by different people for different reasons.
The mechanism that survives
One thing does carry over: the level is visible. It is drawn on a great many charts, alerts are set on it, and it appears in commentary. A price a lot of people are watching produces real behaviour when price returns to it.
That is a weaker claim than the mind-changing story, and it is the one that holds up. It also explains the pattern’s own limits — it predicts that obvious levels work better than obscure ones, and that the effect fades as attention moves on.
Which is why the first retest is the one worth trading. Attention is highest immediately after the break. By the third or fourth visit the level has been absorbed into ordinary price action and there is nothing distinguishing it.
In practice
A level is never one price. The highs that formed it were at slightly different values, so it is a band, and how wide that band is gets decided by whoever is drawing it. A wide band is always “respected”; a narrow one is often “broken.” Choosing the width after seeing the retest is where most of the pattern’s apparent reliability comes from.
Fix the width first — a percentage, or a multiple of average true range — and the concept becomes something you can be wrong about.
Volume at the break is the one non-price input. A break on heavy participation means the supply was genuinely consumed; a break on almost nothing means the level was not really tested, and price frequently comes back to test it properly.
Higher timeframes have fewer and better levels. A daily level is visible to far more people than a five-minute one, which — given that visibility is the mechanism — means the effect should be stronger and rarer there. That is a testable prediction of the visibility account, and it matches most traders' experience.
A gap through the level removes the retest. Price is past it before anyone can act, and the orders that would have been placed there never were.
A stop just beyond the level sits with everybody else’s. The same visibility that makes the level work makes the stop crowded, which is the liquidity pool problem in its most common form.
The concept is symmetric, and in a downtrend it produces a sequence of broken floors acting as ceilings — each one weaker than the last as attention moves down with price.
Each test traded costs 2% of a typical bar’s range in round-trip costs on this history, whether the level holds or not.
What role reversal is not
It is not a memory in the market. The orders are gone. What persists is a number people are watching.
It is not a guarantee that a broken level holds. It is a reason to expect more activity there than at a random price.
It is not a breaker block as a separate concept. That is the same idea in the smart-money vocabulary, with a candle-based definition of where the level sits.
And it is not stronger for the level having held many times. More prior tests means more of the original supply was consumed before the break, which cuts both ways.
When it fails
In a range the concept fires everywhere. Price crosses the middle of the range repeatedly, and every crossing can be described as a level flipping role. The vocabulary keeps working while the meaning has gone.
The second failure is the band width chosen after the fact. If “close enough” is decided once you see whether it held, the pattern cannot be wrong.
A third is the late retest. The fourth visit to a level three weeks after the break has none of the attention that made the first one meaningful.
A fourth is trading the retest without the break’s volume. A level cleared on nothing was not cleared; it was drifted through, and the supply may still be there.
And a fifth is treating it as a reason rather than a location. The flip tells you where a decision point is. Whether to take the trade needs the same evidence any other entry needs — participation, structure, and a stop that is not sitting in the crowd.
The original data
On this site’s shared 576-bar history, of 39 closes above a 20-bar high, 85% closed back below that
level within ten bars; at a 10-bar lookback, 53 events and 70%; at 55 bars, 11 events and 100%. The
counts are in research/series-measurements.json, produced by site/measure_series.py.
Those figures are the retest’s base rate, and they are the reason this page treats role reversal as a location rather than a signal. On a series with no participants and no memory, price came back through a broken level most of the time — so the retest itself is the ordinary case, not evidence of anything. What the visibility account predicts, and what you can test, is that the effect should be stronger on levels more people can see: daily rather than five-minute, round numbers rather than odd ones, recent breaks rather than old ones. Splitting your own logged retests along those three lines is a real experiment, and it is the one this concept has been waiting for.
Related
Support and resistance covers what a level is before it breaks. Breaker block is the same idea in the smart-money vocabulary. And breakout is the event that creates the flip in the first place.
Role reversal is the piece of classical technical analysis I still use every week, and the thing that made it work was demoting the story. I stopped thinking about who changed their mind and started thinking about who can see the line, which is a smaller claim and a much more reliable one.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.