Direct answer
A liquidity gap is a market-structure area that can form when price moves quickly from one level to another while resting buy/sell interest is relatively thin or temporarily missing. A worked example is a hypothetical sequence with stated assumptions about order placement, the absence of nearby resting quotes, and how price transitions when traders hit the remaining available liquidity.
Mechanism or definition (what we mean by “gap”)
A common way to describe liquidity gaps is:
- Resting liquidity: orders already placed and waiting to be filled.
- Thin liquidity area: a price range where there are relatively few resting orders.
- Gap (in price movement): when price travels from one liquidity “island” to another with little trading activity in between.
Important: liquidity gaps are not guaranteed to appear in every fast move, and the exact edges of a “gap” depend on measurement choices (for example, what you treat as “thin enough” and what data granularity you use).
Worked example (scenario with all assumptions)
We will use a simplified order-book and price path. No real-time prices are used.
Assumptions (state everything):
- We observe an FX market with an internal “price” that moves in steps.
- There are two main levels of resting liquidity:
- A sell-liquidity cluster at 1.2000 (lots of sell orders waiting).
- A buy-liquidity cluster at 1.2040 (lots of buy orders waiting).
- Between 1.2005 and 1.2035, resting liquidity is insufficient (think: few orders remain or they get consumed quickly).
- A sudden buy impulse begins at time T0.
- A “gap” is defined here as the range from the last trade near the first liquidity cluster to the first trade near the next liquidity cluster, using our assumed boundaries.
- Trading spreads, slippage, and execution rules are ignored in the numeric example; they will be discussed later as limitations.
Scenario:
- Before T0, price is held near 1.1998–1.2000 because sell orders at 1.2000 provide resistance.
- At T0, an incoming buy order (or a coordinated demand shock) consumes the nearby sell liquidity at 1.2000 and pushes price upward.
- Once the price passes 1.2005, it reaches the assumed thin zone 1.2005–1.2035. Because there are few resting orders there, the price does not “pause” and does not have many natural stopping points.
- Price therefore jumps toward the next area with meaningful resting buy liquidity, and the first noticeable stabilization occurs near 1.2040.
Compute the “gap” boundaries (based on our definition):
- Last trade near the first cluster: 1.2000.
- First trade near the next cluster: 1.2040.
- Liquidity gap range (by our working definition): 1.2000 to 1.2040, which is a 0.0040 (40 “pips” in a 5-decimal style, but we are only using the arithmetic difference).
What to take from the number: The point is not that every gap equals exactly 0.0040. The point is that with stated assumptions—two liquidity islands and a thin middle—price can traverse quickly and leave a relatively “empty” zone.
Evidence or example comparison (stable mechanics vs variable conditions)
The stable mechanics in the scenario are:
- Price needs to consume nearby resting liquidity.
- If the middle range has too little resting liquidity, price can move quickly to the next place with liquidity.
What is variable and can change outcomes in real conditions:
- Market participation: how many market makers or participants are quoting in the middle range.
- Execution and microstructure: whether orders are refreshed quickly or pulled.
- Costs and constraints: spreads, slippage, and order size limits can change the path price takes.
- Measurement choice: the “edges” of a gap depend on what data frequency you use and how you label thinness.
Limitations and risks (failure modes)
Even if the definition is clear, there are material limitations and failure modes:
- Ambiguous boundaries: If you define the gap differently (for example, using different thresholds for “thin”), you can get a different gap range. 2. Not every fast move is a liquidity gap: A move can be rapid for reasons that do not imply persistently thin resting orders in the middle. 3. Data limitations: With coarser data, you may see a “jump” that is really just a sampling artifact. 4.