WhitmanTrading

Breaker Block vs Break of Structure

Breaker blocks mark an area price broke through before returning to it from the other side, which requires structure to have failed. A break of structure is that failure itself, so the two are cause and consequence rather than competing setups.

These are not two options. One is an event and the other is the area that event leaves behind, and the second cannot be marked until the first has happened. Because the event can be defined exactly, the area can be too, and then it can be counted: how often price came back to it, and what it did there.

What each one is

A break of structure is the close that clears a swing level. Used in the narrow sense, it means the trend carrying on: an uptrend closing above its last swing high. Break of structure covers it.

A breaker block is the area left on the far side of that failure, which price later revisits from the other side. This site’s breaker block page draws it from the last candle of the opposite color before the break (for a break downward, the last up candle before price broke through the level) and uses that candle’s whole range as the block.

The break is the precondition. Without a structural failure the area is a level that has not been tested, and calling it a breaker is premature.

Where they differ

Which is the event. The break happens; the breaker is what remains afterwards. One is a moment and the other is a place that persists.

When each is identifiable. The break the moment the close clears the level. The breaker as a place at the same moment, and as a trade only once price comes back to it from the other side, which is later and does not always happen.

What each requires you to define. The break needs a swing definition. The breaker needs that, plus which candle forms the area and what counts as a return.

Where each sits. The block is usually not the broken swing level. On SPY the block lay wholly on the far side of the swing level from the break close in 325 of 365 cases and straddled the level in the other 40, so a breaker and a retest of the broken level are different trades at different prices.

What the break and its block share

Both depend on the same swing definition. The bar count, the wick rule and the timeframe decide the break, and with it every breaker drawn from it.

Both are frequent. Swing points sit close together when the market keeps reversing after a few bars, so structure breaks often in both directions and a block is left behind almost every time.

And neither supplies a stop. An invalidation just beyond a breaker or just beyond the broken swing has to survive whatever the next wide bar does, and neither concept says how wide the instrument’s bars get.

Which one to use: Breaker Block or Break of Structure?

Read the break first, as the event. It establishes what has changed, and without it the area you are about to draw has no meaning.

Use the breaker as the place, second. After the break confirms, the block is somewhere to act with a defined invalidation rather than chasing the move.

Use the breaker when you missed the break itself. It is a way of entering after the event rather than during it, which is its most honest application.

And when you cannot state what counts as a break, do not use the breaker. The second concept inherits every ambiguity in the first, plus its own.

What the block did when price came back

Price came back about as often as it would to any band. Of 365 breakers, 249 (68.22%) were revisited within 20 sessions, against 66.86% for an ordinary band of the same size at the same distance from the close on other days (p = 0.62). The median wait was 4 sessions, and 59 came back the very next session.

On the return, the block caught price more often. The returning session closed inside the block on 51.00% of returns against 41.04% for the ordinary band (p = 0.002), and closed right through it on 8.43% against 13.24% (p = 0.02).

But it did not turn price back more often. The return closed back on the break side, the outcome a breaker trade needs, on 40.56% of returns against 45.72% for the ordinary band (p = 0.11). A block that stalls price without rejecting it gives a trader a place to wait, not a reaction.

A stricter baseline shifts one result. Placing the same band after the other breaks in the same direction, instead of after any session, it was revisited 65.56% of the time; on the return it closed back on the break side 48.52% of the time, inside 40.62% and through 10.86%. Against that, the blocks still stalled inside more often (p = 0.001) and turned price back less often (p = 0.01), but closed through about as often (p = 0.26), so the lower through rate above may owe more to what follows any break than to the block.

Horizontal bars for SPY breakers from 1996 to 2026 against an ordinary price range: price returned within 20 sessions 68.22% against 66.86%; on the return it closed back on the break side 40.56% against 45.72%, and inside the breaker 51.00% against 41.04%.
SPY breakers after five-bar breaks of structure, January 1996 to October 2026, against an ordinary price range of the same size and distance on every session that was not a break. Source: Yahoo Finance, SPY daily bars (m71-spy-breaker-blocks-after-structure-breaks-1996-2026.csv).

A breaker after a break: SPY, 28 April 2023

The break. SPY closed at $415.93 on 28 April 2023, 21 cents above the $415.72 swing high of 18 April, an upward break of structure.

The block. The last down candle before the break was 26 April, two sessions earlier: open $406.72, close $404.36, range $403.78 to $407.84. The break close was well above the block’s top, so the block counts.

The return. On 2 May the low reached $407.82, two cents inside the block’s top, and the session closed at $410.84, back above it. By the rule here that is a hold, the outcome a breaker trade wants, decided by two cents of overlap. Two sessions later, on 4 May, the low was $403.74, four cents under the block’s bottom.

The original data

The bars. Yahoo Finance’s daily SPY bars with opens, 29 January 1993 to 2 October 2026. Breaks are counted from 2 January 1996: in 1993 to 1995, 69 of 737 recorded opens (9.4%) simply repeat the previous close, against 28 of 1,011 (2.8%) in 1996 to 1999, and candle color depends on the open. Structure uses the five-bar rule of the break of structure page (a swing high is a high above the five highs on each side, known five sessions later; a newer swing replaces it; a close beyond breaks it), reproduced row for row against that page’s event file.

The breaker rule. For an upward break, the block is the last down candle (close below open) before the break; for a downward break, the last up candle. Its whole high-to-low range is the block. 388 breaks of a live swing level, counting continuations and changes of character alike; in 23 the break close did not clear the block, leaving 365 breakers. The block candle was a median of 2 sessions before the break and the session immediately before it in 111 cases; the block’s median width was 1.02% of the break close, and the break close sat a median 1.24% beyond it.

Return and outcome. A return is the first of the next 20 sessions to trade back into the block from the break side; the outcome is where that session closed: back on the break side, inside the block, or through its far side. The comparison band takes the same percentage distances from the close of every session that was not itself a break and runs them through identical rules; tests are exact binomial against that pooled rate. They treat the 365 breakers as independent, which overlapping windows make only roughly true, so read the p-values as approximate. Every break and its block is in the SPY breaker block file.

When it fails

The failure is drawing breakers without a break definition, and every losing trade becomes a new setup. An area is entered and price goes through it. Rather than taking the stop, the area is relabelled a breaker and traded from the other side. No structural break was specified or checked; all that actually happened is that the first idea failed. Two round trips are paid, and the second trade exists only because the first one lost. With a written break rule the relabelling is impossible, because the block either followed a qualifying break or it did not.

The second failure is a vague swing rule. Two readers of the same chart then find different breaks.

A third is drawing the area after the return. Everything then works backwards.

A fourth is no expiry. This page gives each block 20 sessions; without a limit, failed levels accumulate indefinitely.

A fifth is reading a stall as a rejection. On SPY the block held price inside more often than it turned it back.

And a sixth is reading structure on a timeframe you do not trade. The two disagree by design.

Breaker block covers the area and the candle rule. Break of structure covers the event that creates it. And change of character covers the break that goes against the run.

The practical check

You cannot have the second without the first, and yet most explanations of breakers skip straight to drawing the box. Write your break rule down first, then count what your breakers actually did on your own instrument before you trust one.

— Michael Whitman

This page is educational, not financial advice. Test every idea on your own charts before risking money.