Break of structure (BOS) and CHoCH: what ICT traders mean, and what is proven

Smart Money Concepts By Alphaex Capital Updated

A quick-reference summary before the detail.

Key takeaways

  • A break of structure (BOS) and a change of character (CHoCH) are terms from the ICT methodology of Michael Huddleston, and ICT traders use them to mark the moments a trend either continues or starts to reverse.
  • In ICT terms, a BOS is a close beyond a prior swing point in the direction of the prevailing trend, which the methodology reads as confirmation that the trend is continuing, while a CHoCH is the first close against the trend, which it reads as an early reversal warning.
  • The structure these terms sit on, the sequence of higher highs and higher lows, is not ICT's invention. It comes from Dow Theory, systematised a century ago by William Peter Hamilton and Robert Rhea, and that scaffolding is externally defensible.
  • The part that has external support is the broader behaviour: trends do persist and price levels do cluster around prior swings. Carol Osler's Federal Reserve Bank of New York research documents how stop orders cluster near round numbers, which is the closest honest anchor for why structure breaks matter.
  • No peer-reviewed or tier-1 study tests ICT's specific BOS or CHoCH rules and publishes a win rate. Every "70 percent win rate" claim you read online is unsourced lore, and the only honest way to size the edge is to test it on your own data.

What a break of structure actually means, in ICT terms

In ICT methodology, a break of structure is price closing beyond a prior swing point in the direction of the prevailing trend, and ICT traders read that close as confirmation that the trend is continuing. In an uptrend built from higher highs and higher lows, a bullish break of structure is a close above the most recent higher high.

In a downtrend built from lower highs and lower lows, a bearish break is a close below the most recent lower low (The Inner Circle Trader, ICT methodology).

ICT stands for the Inner Circle Trader, the public name of trader Michael Huddleston, who taught this framework through his YouTube channel and the 2022 Mentorship Model. The terms are his, taught informally and repeatedly rather than derived from a textbook, and the recurring practitioner phrasing is that a BOS confirms continuation.

I keep the phrase "in ICT methodology" attached to the definition because the framing is the whole point. A break of structure is a labelled price event inside a community's method, not a fact about what institutions did on the tape, and whether the labelled break reliably precedes continuation is an empirical question each time.

Change of character: the first break against the trend

A change of character is the first break of a swing point against the prevailing trend, and ICT traders treat it as the earliest warning that a reversal may be starting. In an uptrend, a bearish change of character is a close below the most recent higher low.

In a downtrend, a bullish change of character is a close above the most recent lower high (The Inner Circle Trader, ICT methodology).

The distinction from a BOS is direction alone. A break of structure breaks a swing point in the trend's own direction and reads as continuation; a change of character breaks a swing point against the trend and reads as the first sign of a shift.

The practitioner phrasing is that a CHoCH is the first sign of a shift, not confirmation of one.

I stress "first" because that word does most of the work. A change of character is a warning, not a verdict, and many CHoCHs fail while the original trend resumes on the next timeframe up.

Reading it as a guaranteed reversal is the most common way traders lose money with the concept.

Break of structure versus change of character

The comparison between BOS and CHoCH is the single most-searched structure question, and most answers online blur it. The clean rule is directional: a BOS breaks a swing point in the trend's direction and confirms continuation, while a CHoCH breaks a swing point against the trend and warns of reversal (The Inner Circle Trader, ICT methodology).

AspectBreak of structure (BOS)Change of character (CHoCH)
Direction of the breakIn the direction of the prevailing trendAgainst the prevailing trend
What ICT traders read it asContinuation confirmedFirst warning of a possible reversal
Example in an uptrendClose above the most recent higher highClose below the most recent higher low
Where it fits in a reversalComes second, confirming the new directionComes first, flagging that the old trend may be done
How reliably ICT traders treat itA continuation read, still able to fail as a false breakAn early warning, not a confirmed reversal

I lay the two side by side because the difference is one directional choice, not two unrelated patterns. In a reversal sequence the change of character fires first, then a break of structure in the new direction confirms the shift, which is why ICT traders wait for both rather than acting on the CHoCH alone.

The structure ICT is labelling, and where it really comes from

The scaffolding underneath both terms is the sequence of swing highs and lows. An uptrend is a run of higher highs and higher lows, a downtrend is a run of lower highs and lower lows, and a range is the absence of a clean progression in either direction.

That is the structure a BOS or CHoCH breaks.

This part is not ICT's invention. The higher-highs and higher-lows definition of a trend comes from Dow Theory, systematised a century ago in William Peter Hamilton's "The Stock Market Barometer" (1922) and Robert Rhea's "The Dow Theory" (1932), built on Charles Dow's original Wall Street Journal editorials.

ICT relabelled a real, externally defensible framework rather than discovering it.

Even the swing point itself has older roots. Bill Williams defined a fractal swing as a candle with two lower candles on each side, while ICT traders, like me when I chart, more often mark swings structurally by eye, which introduces subjectivity the community rarely acknowledges.

The broader market structure page covers the validated trend-structure scaffolding in more depth, and it is the honest anchor for everything ICT layers on top.

Strong versus weak breaks, and the close-versus-wick question

ICT traders split breaks into strong and weak, and the split matters more than the SERP admits. A strong break of structure closes beyond the swing point with displacement, meaning a large body, a forceful move and minimal overlap, which the methodology reads as institutional commitment.

A weak break pips beyond the level with a small body or only a wick, which the methodology reads as a probable stop run rather than genuine continuation (The Inner Circle Trader, ICT methodology).

The displacement that marks a strong break is the same displacement that leaves a fair value gap behind, which is why the two concepts are taught together. A break with no gap and no body is the methodology's own signal to distrust the break.

The confirmation rule that pages quietly skip is whether a break needs a candle close. Requiring a body close beyond the level produces fewer, higher-quality signals, while accepting a mere wick produces many false breaks.

The choice changes how often the method fires, and I have not seen a ranking page state plainly that this one decision drives most of the disagreement between ICT educators.

Market structure shift, and the naming mess around it

A third term complicates the picture. A market structure shift, or MSS, is used inconsistently across the ICT community.

A subset of educators treat MSS as a straight synonym for a change of character, while ICT purists reserve it for a fuller, confirmed shift across timeframes, with the CHoCH as the earlier, lower-timeframe warning that precedes it (The Inner Circle Trader, ICT methodology).

I flag the conflation because the SERP uses the two interchangeably and then assumes a distinction that may not exist in the source material. When two educators argue about whether a move was a CHoCH or an MSS, they are often working from different definitions of the same label rather than reading the chart differently.

The practical response is to fix your own definition before you trade. Pick one meaning for a market structure shift, write it down, and apply it the same way every time, because consistency is what makes a labelled method testable at all.

How ICT traders use a break in a trade

The break is not the entry. In ICT's own logic, a BOS or CHoCH identifies directional bias, and the entry comes later, at the order block or fair value gap formed during the displacement leg, after price retraces to it (The Inner Circle Trader, ICT methodology).

The full sequence layers the pieces. A higher-timeframe bias sets the direction, a break of structure or change of character confirms the leg, price retraces into a discount or premium zone, and an order block with a fair value gap inside it marks the specific level.

The swing-trading application of SMC is where those pieces assemble into a complete setup.

I present the sequence as a method rather than a recommendation. The same structure can be described perfectly and still lose if the zone fails, so the discipline of waiting for the retrace and the confluence is the method's genuine contribution, separate from whether any single labelled break pays.

What is actually proven about structure breaks

The honest evidence sits one level up from ICT. Carol Osler's research at the Federal Reserve Bank of New York, "Support for Resistance," found that currency stop-loss orders cluster predictably around round numbers and prior swing points, which generates the support and resistance behaviour traders observe.

That is real, peer-reviewed, and it explains why breaks beyond a level feel meaningful.

The observation that trends persist is also documented. Time-series momentum research, including work by Tobias Moskowitz, Yao Hua Ooi and Lasse Heje Pedersen, finds that assets that moved up or down over a lookback tend to continue, which is the population-level fact that a continuation break encodes.

None of that validates ICT's specific rules, only the broader tendency.

I pair that with the honest caveat that breakout-rule evidence is contested. Early studies such as Brock, Lakonishok and LeBaron (1992) found technical breakout rules had predictive power in historical data, but Sullivan, Timmermann and White (1999) showed similar results can arise from data-snooping once the full universe of rules is tested.

The legitimate order-flow toolkit lives in volume profile and order-book imbalance, taught in exchange education such as CME Group's, which is what reading institutional flow means when done with data instead of a candle label.

The win-rate claim nobody can source

You will read that a break of structure has a 70 percent win rate, or 80, or some other precise figure. I have not found a peer-reviewed or tier-1 study that tests ICT breaks of structure and publishes a win rate, and neither has any page I have seen quote one with a citation.

Every specific percentage circulating on forums and educator blogs traces back to an unsourced claim repeated until it reads like data. The honest statement is that no verified backtest exists in the public literature, and anyone quoting a number should link the study or retract it.

I would rather tell you the evidence is missing than hand you a reassuring statistic. The absence of a clean win-rate study is itself the finding, and it means the only honest way to size the edge is to test the method yourself on your own data and treat your result as a sample, not a certainty.

Where structure breaks fail

Most breaks do nothing useful, and I want that stated first. In any trend there are dozens of structure-breaking candles, the methodology's own users ignore most of them, and survivorship bias colours every chart example you see online.

The breaks that worked get screenshotted and the breaks that failed do not.

Breaks fail outright as false moves. Price pips beyond a swing high to sweep buy-side liquidity, the resting stop-buys above it, then reverses sharply, so what looked like a bullish BOS becomes a bearish CHoCH on the next leg.

The strong-versus-weak lens exists precisely to filter these stop runs out.

The deeper problem is hindsight rationalisation. When a level breaks and holds, the method calls it a genuine BOS; when the same level breaks and fails, the method relabels the move as a liquidity sweep, so the framework can appear to have been right either way after the fact.

The disciplined response is to define your break, your close rule and your invalidation in advance, trade only the small subset with real confluence, and keep a record of failures alongside winners.

FAQ

What is a break of structure in trading?

In ICT methodology, a break of structure is a close beyond a prior swing point in the direction of the prevailing trend. ICT traders read it as confirmation that the trend is continuing, for example a close above the most recent higher high in an uptrend.

It is a labelled event within a methodology, not a fact about institutional order flow (The Inner Circle Trader, ICT methodology).

What is a change of character (CHoCH)?

A change of character is the first break of a swing point against the prevailing trend. ICT traders treat it as the earliest warning of a possible reversal, such as a close below the most recent higher low in an uptrend.

It is a warning rather than a confirmed reversal, and many changes of character fail (The Inner Circle Trader, ICT methodology).

What is the difference between BOS and CHoCH?

The difference is direction. A break of structure breaks a swing point in the trend's own direction and reads as continuation, while a change of character breaks a swing point against the trend and reads as the first reversal warning.

In a reversal sequence the change of character comes first, then a break of structure in the new direction confirms it (The Inner Circle Trader, ICT methodology).

Does a break of structure mean the trend will continue?

Not reliably. A break of structure is the methodology's continuation read, but a break can be a false move that pips beyond a level to sweep liquidity and then reverses into a change of character.

ICT traders filter for strong breaks with displacement and a body close, and even then the read is probabilistic, not guaranteed.

Does a break of structure need a candle close?

ICT traders disagree. Requiring a body close beyond the swing point produces fewer, higher-quality signals, while accepting only a wick produces many false breaks.

The choice materially changes how often the method fires, and it is the single biggest source of disagreement between ICT educators.

What is a market structure shift (MSS)?

The term is used inconsistently. Some ICT educators treat a market structure shift as a synonym for a change of character, while ICT purists reserve it for a fuller, confirmed shift across timeframes, with the change of character as the earlier, lower-timeframe warning.

The conflation is worth knowing because the two terms are often used to mean the same thing (The Inner Circle Trader, ICT methodology).

Do ICT breaks of structure actually work?

There is no peer-reviewed or tier-1 study proving ICT breaks of structure produce a reliable edge, and no verified win-rate statistic exists in the public literature. Trends do persist and stop orders do cluster near levels, documented in research such as Carol Osler's Federal Reserve work, but whether the specific ICT rules are profitable is an untested empirical question you should evaluate on your own data.

Who created the break of structure concept?

The term and its mechanical rules come from Michael Huddleston, publicly known as the Inner Circle Trader (ICT), taught through his YouTube channel. The underlying trend structure of higher highs and lower lows is older, coming from Dow Theory as systematised by William Peter Hamilton and Robert Rhea a century ago.

Continue Learning

Explore more guides and build on what you just read.