Market Structure, BOS and CHOCH: Objective Rules for Traders

Market structure is the sequence of swing highs and swing lows a trader uses to describe what price has been doing. A break of structure (BOS) is a continuation event: price extends beyond the last relevant swing in the direction of the existing sequence. A change of character (CHOCH), also called a market structure shift (MSS) by many traders, is the first break against that sequence.

Both definitions sound simple and are surprisingly slippery in practice. Whether the same chart shows a BOS, a CHOCH, or nothing at all depends on how you define a swing point, whether you require a candle close, which swing counts as the relevant one, and which timeframe you are looking at. Change any of those and the label changes with it.

This guide focuses on making structure objective enough to test. It is the structural layer of the broader Smart Money Concepts framework, and it sits upstream of entry-zone tools such as order blocks and fair value gaps.

Step one: define a swing point mechanically

Every structure label inherits its ambiguity from the swing definition. “Obvious high” is not a definition. Choose one of the following and record it as part of the strategy.

Fractal or N-bar swing

A swing high is a candle whose high exceeds the highs of the N candles on either side; a swing low is the mirror image. N is typically 2, 3, or 5. Larger N produces fewer, more significant swings and later confirmation. Note the practical consequence: an N-bar swing is only confirmed N bars after it forms, so a “confirmed” swing point is always slightly in the past.

Threshold or ZigZag swing

A new swing is registered only when price retraces by at least a fixed percentage, a fixed tick amount, or a multiple of ATR. This adapts to volatility and filters small noise, but the threshold value becomes a parameter you can accidentally optimize. Fix it before testing and record it.

Structural or “last opposing extreme” swing

The reference swing is the extreme reached before the move that created the current leg—for example, the lowest low before the rally that broke the previous high. This tends to match how traders read charts visually, but it needs an explicit tie-breaking rule when several candidates cluster together.

• Write down N, the threshold value, or the tie-break rule, and treat changing it as creating a new strategy.
• Decide whether swing points are drawn from wicks or from candle bodies. Both are defensible; mixing them is not.
• Decide how many bars of separation two swings need before both count.
• Decide what happens when a single candle makes both the highest high and lowest low of a leg.

What counts as a break of structure?

A BOS is a continuation signal. In an uptrend defined by higher highs and higher lows, a BOS occurs when price breaks above the most recent confirmed swing high. In a downtrend, it occurs when price breaks below the most recent confirmed swing low.

1. Identify the current directional sequence using your swing definition.
2. Identify the single reference swing that must be exceeded.
3. Apply your breach rule: wick beyond, close beyond, or beyond by a minimum distance.
4. Record the bar on which the break was confirmed, not the bar on which it looks obvious in hindsight.
5. Update the structure: the break creates a new reference swing for the next evaluation.

The last step is where discretionary traders quietly cheat. After a break, the reference swing must move—even if the new reference makes your open trade look worse. Freezing the old reference because it suits the current position is the most common way structure rules stop being rules.

What counts as a change of character?

A CHOCH is the first break in the opposite direction to the prevailing sequence. In an uptrend, it is a break below the most recent higher low. In a downtrend, it is a break above the most recent lower high. It is treated as a warning that the sequence may be ending—not as proof that a new trend has begun.

• A CHOCH breaks the swing that was protecting the prior sequence, whereas a BOS extends the sequence.
• A single CHOCH is frequently followed by a resumption of the original trend, especially on lower timeframes.
• Many traders require a CHOCH and then a subsequent BOS in the new direction before treating the trend as changed.
• In a range, CHOCH signals appear repeatedly in both directions and carry very little information.

CHOCH, MSS, and terminology variability

Terminology is not standardized, and this causes real confusion when comparing rules or results. Some educators treat CHOCH and MSS as identical. Others reserve MSS for a break accompanied by displacement, and use CHOCH for any counter-sequence break. Others distinguish internal CHOCH (within a leg) from external or swing CHOCH (against the higher-timeframe sequence).

None of these usages is authoritative. What matters is that your definition is written down, and that when you read someone else’s material you check which definition they are using before assuming their statistics apply to your rules.

Wick versus close: the decision that changes everything

The single largest source of disagreement between two traders labelling the same chart is whether a break requires a candle to close beyond the level or merely to trade beyond it. Both approaches are usable; they produce different strategies with different characteristics.

Wick-based breaks

• Signals appear earlier, which can mean better entry prices.
• Far more signals overall, including many that reverse immediately.
• Highly sensitive to a single spike, a thin-liquidity print, or a news wick.
• Interacts awkwardly with sweeps: the same wick may be a break under this rule and a liquidity sweep under another.

Close-based breaks

• Fewer signals, each with more evidence behind it.
• Later confirmation, which usually means a worse entry price and a wider stop.
• Depends on the timeframe you close on: a 5-minute close and a 1-hour close are very different filters.
• Depends on the session boundary and data feed, especially in 24-hour markets where candle open times differ between providers.

Adding a displacement or distance requirement

A middle option requires the break to exceed the level by a minimum amount—a tick count, a percentage, a fraction of ATR, or a candle body larger than a rolling average. This filters marginal breaks at the cost of missing the ones that only just qualified. Whatever you choose, express it as a number, not as “convincing.”

Timeframe hierarchy

Structure exists independently on every timeframe, and the timeframes routinely disagree. A 5-minute CHOCH can occur inside an untouched 4-hour uptrend. Neither label is wrong; they answer different questions. The problem arises when a trader browses timeframes until one agrees with the position they want to take.

1. Choose exactly two or three timeframes and name their roles: bias, setup, and execution.
2. Fix them for the entire test. Changing the pair mid-sample creates a new strategy.
3. Write the alignment rule explicitly: for example, “only take long setups while the bias timeframe has not printed a close-based CHOCH below its last higher low.”
4. Define what happens when the bias timeframe is ranging rather than trending—most often, no trade.
5. Record the bias state at the moment of entry, not as remembered afterwards.

A useful discipline is to write the bias down before the session starts and refuse to revise it intraday. If it turns out to be wrong, that is data. Revising it in real time removes your ability to ever measure whether the bias rule adds anything.

Internal versus external structure

Internal structure refers to the smaller swings inside a larger leg; external or swing structure refers to the major swings that define the leg itself. Traders use internal breaks for timing and external breaks for direction.

• An internal CHOCH inside a strong external trend is common and often resolves as a pullback rather than a reversal.
• An external CHOCH is rarer and is what most traders mean by “the trend may be changing.”
• Minor highs and lows that look like obvious targets are sometimes labelled “inducement”—but the label is an assumption about other participants, not an observation.
• Distinguishing internal from external requires the swing definition to be applied on two named timeframes, otherwise the distinction is purely visual.

Why structure signals fail

Structure rules generate a large number of signals that go nowhere. Understanding the recurring causes helps you decide which filters to add and which to leave alone.

• Ranges: inside a range, price alternately breaks minor highs and lows, producing continuous BOS and CHOCH labels with no directional information.
• Single-wick breaks: a spike beyond a level with an immediate reversal satisfies wick-based rules and little else.
• News and scheduled events: a release can break structure in one candle and fully reverse in the next.
• Low-liquidity periods: session gaps, rollover, holidays, and weekend crypto trading produce breaks on very little participation.
• Timeframe mismatch: an execution-timeframe break that contradicts the bias timeframe often resolves against the smaller timeframe.
• Late entries: entering several bars after the break means the stop must sit far away, damaging reward-to-risk even when the direction is right.
• Ambiguous swing clusters: when several swings sit close together, small differences in the reference swing flip the label.

A false signal is not a broken rule. Every rule set produces losing signals; the question is whether the winners are large enough relative to the losers. That is an expectancy question, answered by a logged sample rather than by adding another filter after each loss.

Three rule templates

These are educational templates for defining testable variants, not trading recommendations. Adjust every threshold to your market and timeframe and test before risking capital. Test them separately—they are three strategies, not one.

Template A — Strict close-based continuation

1. Swing definition: 3-bar fractal, wick-based, on the execution timeframe.
2. Bias: the higher timeframe sequence must be making higher highs and higher lows for longs.
3. Signal: an execution-timeframe candle closes above the most recent confirmed swing high.
4. Entry: on the open of the next candle, or on a limit at the broken swing level within a fixed number of bars.
5. Invalidation: below the swing low that preceded the break, plus a buffer of 0.25 ATR.
6. Target: the next external swing high, or a fixed 2R, chosen in advance and never mixed.
7. Time stop: exit if the trade has not reached 1R within a defined bar count.

Template B — CHOCH reversal with confirmation

1. Context: the bias timeframe has reached a pre-marked level, such as a prior-day extreme.
2. Signal 1: an execution-timeframe close beyond the last protecting swing, against the prior sequence.
3. Signal 2: a pullback that fails to make a new extreme in the old direction.
4. Entry: on the pullback, at a predefined reference price with a maximum wait in bars.
5. Invalidation: beyond the extreme that formed before the CHOCH, plus a buffer.
6. Target: the first opposing external swing, with an optional partial exit at a fixed R.
7. Filter: no entry within a defined window around scheduled high-impact news.

Template C — Displacement-filtered break

1. Signal: a structure break where the breaking candle body is at least 1.5× the 20-period average body.
2. Additional requirement: the break must occur inside a defined session window.
3. Entry: at a predefined retracement level of the displacement leg, with a maximum wait in bars.
4. Invalidation: beyond the origin of the displacement leg.
5. Target: fixed R, or the next external swing, chosen before entry.
6. No-trade: if the required reward-to-risk is not available at the entry reference, skip the trade and log it as skipped.

How to test structure rules

1. Write the swing definition, breach rule, timeframes, and alignment rule before opening any chart.
2. Use bar-by-bar replay so the outcome of the current leg is hidden while you label.
3. Label structure prospectively and leave incorrect labels visible instead of erasing them.
4. Log every qualifying signal, including the ones you would have skipped emotionally.
5. Apply realistic spread, commission, and slippage; structure setups often have tight stops where costs matter.
6. Keep wick-based and close-based variants in separate samples.
7. Collect enough trades before judging—see how many trades are needed to test a strategy.
8. Validate the finished rules on a period or instrument you did not use while designing them, following the backtesting process.

What to journal for structure trades

• Instrument, date, session, and the exact bias and execution timeframes.
• Structure event: BOS or CHOCH, and internal or external.
• Breach type: wick, close, or displacement-filtered, with the measured breach size.
• Bias-timeframe state at entry: trending, ranging, or unclear.
• Reference swing used, and whether alternatives were nearby.
• Entry model, planned stop, planned target, and reward-to-risk at entry.
• Realized R, costs, and slippage against the planned stop.
• Screenshots at entry and exit, with the structure labels visible.
• Rule adherence and one sentence on what would have changed the outcome.

With that tag set in place, setup analytics can answer specific questions: do close-based breaks outperform wick-based ones after costs, do CHOCH reversals only work when the bias timeframe is trending, and does the displacement filter improve reward-to-risk enough to justify the trades it removes? Report results in R-multiples so a slipped stop does not disappear into a currency figure.

Structure across markets and sessions

Structure rules are not market-neutral. Session boundaries determine which candles exist, and candle boundaries determine which closes count. In FX, the London open often resolves the Asian range, so a break during that window may behave differently from the same break at other times; the trade-offs of session filters are discussed in the killzones and session timing guide. Record the session and spread context in a forex trading journal.

In crypto, continuous trading means daily candle boundaries are a convention rather than a market event, and venues may differ; keep venue and instrument type separate in a crypto trading journal. In equities, overnight gaps can jump straight past a structural level, and the opening auction concentrates activity; a stock trading journal should record whether the break occurred at the open, midday, or into the close.

Common market structure mistakes

• Labelling swings by eye and calling the result a system.
• Switching between wick-based and close-based breaks depending on the trade.
• Re-drawing structure after the outcome is known so every loss becomes an invalid setup.
• Adding timeframes until one supports the desired direction.
• Trading structure signals inside an obvious range and then blaming the rule.
• Placing the stop just beyond the broken level with no buffer for normal noise.
• Entering late, accepting a much wider stop, and keeping the original target.
• Ignoring costs on tight-stop setups where spread is a significant fraction of risk.
• Comparing your results to someone else’s statistics without checking their definitions.

How Traderizz helps

Traderizz lets you save each structure variant as a separate strategy, tag the break type, structure event, bias state, and session, attach annotated screenshots, and compare realized R and expectancy per tag. Because variants stay separated, you can see whether the close-based rule is carrying the results while the wick-based version quietly loses. A regular weekly review turns that comparison into a single decision each week: keep, tighten, or drop the rule.

Market structure is a language for describing what price has already done. It becomes useful when the description is precise enough to be repeated, and misleading when it is flexible enough to justify anything. Precision is what makes the difference measurable.