Swing failure and why breakouts fail
Lesson 19 · about 10 min
The last lesson showed how a stop run works. This lesson names the chart pattern it produces (swing failure), separates it from a genuine breakout using order flow, and explains why so many breakouts fail even when the chart looked right. The two are the same subject seen from opposite sides.
The swing failure pattern
A swing failure is a move beyond a prior swing high or low that does not hold: price breaks the level, triggers the stops, and closes back on the original side.
Price
5015 | *
5014 | * | * <- 5014.00 was the swing high
5013 | * | *
5012 | * * | *
5011 | * * * | *
5010 | * v | *
+----------------------------
swing high break fail
at 5014.00 to 5015 close 5013
The pattern is old and well known. What order flow adds is the ability to tell, while it is happening, whether the break is failing or succeeding.
The order flow of a failing break
| Signal | Failing break (swing failure) | Succeeding break (breakout) |
|---|---|---|
| Volume beyond the level | Heavy but no progress; absorption | Heavy and price keeps stepping up |
| Extreme of the spike | Finished: near-zero aggressive volume at the very top | Unfinished or still lifting |
| Delta after the level | Positive spike then flips negative fast | Stays positive; stacked imbalances continue |
| CVD at the new high | Lower high than the prior push (divergence) | New high with price |
| Displayed size above | Offers appear or reload as price arrives | Offers pulled; vacuum above |
| Time above the level | Seconds to a couple of bars | Builds a new bulge of volume above |
| Return to the level | Fast, with selling | Level becomes support; buyers defend on retest |
You do not need all seven. Absorption beyond the level, a finished extreme and a fast delta flip are enough to call it. Volume building above the level with stacked imbalances continuing is enough to call it the other way.
Key idea: A breakout and a swing failure look identical for the first few ticks. The difference is whether the aggression beyond the level is absorbed or accepted. Only the footprint shows that in time to act.
Why breakouts fail so often
Put together everything from this module and the answer is mechanical:
- The break is fuelled by stops, not by new decisions. The burst of buying above a high is largely stop orders from shorts. Those buyers did not choose to buy; they were forced to. Once they have covered, they are gone and add nothing further.
- Large sellers are waiting for exactly that. A participant who needs to sell size has been watching the same level and knows the pool is there. They sell into the stop-driven buying.
- Breakout buyers arrive last. Discretionary traders who buy "the breakout" enter after the stops have fired, at the worst prices, and become the trapped longs that fuel the reversal.
- Nothing is left to push. Once stops are covered, the seller has filled, and late longs are under water, the only participants with a reason to act are the trapped longs, and their action is selling.
A breakout succeeds when step 2 does not happen: no large seller is waiting, or the initiative buying is larger than what they have to sell. Then the acceptance builds and the level flips to support. On a trend day, or after a long balance that has drained the responsive sellers, that is common. In a mature range on a rotational day, it is rare.
A numeric contrast
Two breaks of the same 5014.00 high on different days.
| Measure (first three bars above 5014.00) | Day A | Day B |
|---|---|---|
| Ask volume above the level | 3,200 | 3,400 |
| Bid volume above the level | 3,050 | 1,150 |
| Delta above the level | +150 | +2,250 |
| Ticks of progress above the level | 3 | 11 |
| Volume at the top row (bid × ask) | 210 × 4 | 90 × 260 |
| CVD vs prior high | Lower high | New high |
Day A: 3,200 of buying, 3,050 of selling into it, three ticks of progress, and a finished high (four contracts lifted at the top). The buyers were sold into almost one for one. Swing failure. Day B: similar buying, a third of the selling, eleven ticks of progress, unfinished top, CVD confirming. Breakout. On a candle chart the first bar above 5014.00 looked the same on both days.
Trading the failure
The swing failure is a reversal entry with an unusually clear invalidation:
- Wait for the break and the absorption signature (heavy volume, no progress, finished extreme).
- Enter short as price closes back below the level, or on the first retest of the level from below that is rejected.
- Stop above the spike high. If a fresh burst of aggression takes out that high with acceptance, the read was wrong and the cost is a few ticks.
- Target the opposite side of the range or the POC: the trapped longs' stops are the fuel, and they sit below the range's lows or the recent swing low.
In R terms: a stop of 6 ticks above the spike and a target 24 ticks away at the range low is a 4R structure. Whether the win rate justifies it on your product is what your log is for.
Trading the breakout
The mirror: wait for the break and confirm acceptance (volume builds above, stacked imbalances continue, no wall). Enter on the first pullback to the level that holds with responsive buying. Stop below the level; if it fails to hold as support, it was a swing failure after all. Entering on the initial spike is the worst version of this trade because you cannot yet tell which of the two it is.
Try it: On your product, collect ten breaks of an obvious swing high or low. For each, record delta beyond the level in the first three bars, ticks of progress, and whether the extreme was finished. Sort them into failures and successes by what happened over the next hour. See how well the three measures separated them. Adjust thresholds to your product.
Recap
- A swing failure is a break of a swing point that closes back on the original side; it is the chart pattern a stop run produces.
- Failing and succeeding breaks look identical for the first ticks; absorption versus acceptance beyond the level separates them.
- Breakouts fail because stop-driven buying is finite, large sellers wait for it, and breakout buyers arrive last and become trapped.
- Trade the failure on the close back through the level with a stop above the spike; trade the breakout on the first held retest, never on the initial spike.
- Use delta beyond the level, ticks of progress and whether the extreme is finished as your three measures.
See it drawn
Original diagrams for the ideas on this page. Illustrative, not real market data.