How do you define market structure?

When traders say “structure,” they usually mean something like: “the market is making higher highs and higher lows,” or “structure broke,” or “we’re ranging.”

That sounds intuitive—until you try to make it precise. How high is a “high”? Which low matters? When is a break confirmed? Is a wick enough?

A practical definition that works for both discretionary thinking and quant implementation is:

Market structure is the ordered sequence of significant swing highs and swing lows, plus the rules for how those pivots get confirmed and invalidated.

Structure is basically the grammar of price action. Swings are the letters; structure is how the letters form readable words.

1) Structure starts with pivots: without pivots, “higher high” is meaningless

To talk about structure you first need a method to extract pivots (swing highs/lows). That’s why “how to define a swing” and “how to define structure” are inseparable.

Once you have pivots, you can label relationships:

• Higher High (HH)
• Higher Low (HL)
• Lower High (LH)
• Lower Low (LL)

The pattern of HH/HL vs LH/LL is the simplest structural description of trend vs downtrend.

2) The second ingredient: what counts as a “break”?

A structure break is a claim that “the previous reference point no longer holds.” In practice you need rules like:

• Do we require a close beyond the pivot, or is a wick enough?
• Do we need follow-through (e.g., a second bar confirmation)?
• How far beyond the level counts (ticks, %, ATR)?

This matters because markets often probe levels and snap back. Without a break rule, you’ll call every probe a break.

3) “BOS” and “ChoCH” are just labels for two different events

Different trading communities use different terms, but two structural events show up everywhere:

Break of Structure (BOS)

In an uptrend, BOS often refers to price taking out the prior swing high (continuation). In a downtrend, taking out the prior swing low.

Plain idea: “the trend is still doing trend things.”

Change of Character (ChoCH)

ChoCH often refers to breaking the opposite side that the trend should defend. For example, in an uptrend, price breaks a prior swing low.

Plain idea: “the market stopped behaving like the previous trend.”

You don’t have to use those words, but the logic is useful: continuation breaks vs potential reversal breaks.

4) Ranges are structure too: structure isn’t only trends

A mistake is to treat “structure” as something that exists only when the market trends. Ranging markets have structure, it’s just different:

• repeated failed breaks,
• oscillation between boundaries,
• compression (lower volatility) and then expansion.

From a quant perspective, you can treat “range” as a state where directional drift is low and mean-reversion signals work better.

5) A quant-friendly way to formalize structure: a state machine

Once you have pivots and break rules, you can implement structure as a simple state machine:

• State: Uptrend / Downtrend / Range
• Inputs: new pivot confirmed, pivot broken, volatility regime, etc.
• Transitions: Uptrend → Range (failed continuation), Uptrend → Downtrend (ChoCH + confirmation), etc.

This is not just theoretical. It forces you to answer ambiguity with explicit rules.

6) What structure is not

Structure is not:

• one trendline you draw by eye,
• one indicator signal,
• a story (“buyers are strong”).

Those might correlate with structure, but structure itself is the consistent mapping from price history to pivots and breaks.

7) Practical takeaway: define structure in a way that matches your execution horizon

If you trade short-term, your pivots must be sensitive enough to update quickly, otherwise your structure will lag. If you trade long-term, overly sensitive pivots will make you flip states too often.

So the best structure definition is the one that:

• produces stable labels in backtests,
• aligns with your holding period,
• and doesn’t rely on hindsight.

Once you have that, “structure” stops being a vague word and becomes a concrete, testable framework for analysis and strategy.