How can Liquidity Gaps be measured?

Direct answer

Liquidity Gaps can be measured by defining what “gap” means in your dataset, then calculating whether tradable prices jump or whether order-book depth (or its proxy) shows missing liquidity across a chosen time or price range. In practice, measurement is not one number: it is a set of fields (gap size, location, duration) computed using time-stamped observations and explicit assumptions.

Mechanism or definition

A Liquidity Gap is a region where buy and sell liquidity is insufficient or absent so that price can move across it with relatively little trading at intermediate levels. To measure this in a self-contained way, define the gap using observable inputs, such as:

A price discontinuity definition: For a time series of quotes or trades, define a “jump” event when the price moves from one level to another with no prints (or no meaningful updates) inside an interval.

Gap location: the boundary levels before and after the jump.
Gap size: the difference between those boundary levels (e.g., in points or percent).
Gap timestamp: the time of the first observation after the jump and the last observation before it.
Gap duration proxy: the time between the last “pre-gap” observation and the first “post-gap” observation.

A depth/absence definition (order-book proxy): If you have depth data, define a “missing liquidity” zone as a price band where displayed depth falls below a threshold. Then measure:

Gap width in price: upper minus lower bound of the low-depth band.
Low-depth thickness: how far above or below a threshold the depth stays.
Time persistence: how long the band remains under the threshold.

Both approaches require assumptions: what data feed you use, whether you rely on quotes or executed trades, and the threshold for “no liquidity” (for example, “no prints” or “depth below X”).

Evidence or example

Consider a simple, fully specified example based only on time-stamped trades (no real-time claims):

Observation window: one-minute intervals.
Gap rule: a liquidity gap occurs if consecutive trades have a price difference larger than a chosen threshold and there are no trades within an intermediate price range.
Inputs: trade timestamps and executed prices.
Outputs for each detected gap: (a) pre-gap price = last trade before the jump, (b) post-gap price = first trade after the jump, (c) gap size = post-gap − pre-gap, (d) gap time = time(post) − time(pre).

To compare across different days or systems, you must keep the same windowing and thresholds, and you should record whether the dataset includes the same liquidity information (quotes vs trades). Without that, “bigger gaps” can simply reflect different reporting or sampling, not a real change in market structure.

Limitations and risks

At least four material failure modes affect measurement:

Proxy mismatch: If you use quote gaps, they can differ from liquidity available to execute trades. A quote may update even when executable liquidity is limited.
Threshold sensitivity: The choice of “jump size” or “depth below X” changes results. Two analysts using different thresholds can reach different gap counts.
Provider and venue effects: Different data sources can show different timestamps, spreads, or continuity of prints, which can create or hide apparent discontinuities.
Execution and costs: Even if an observed gap exists, real execution may involve slippage, partial fills, or routing behavior that changes the observed path of prices. That means historical gap metrics do not automatically translate into predictable outcomes.

Also note the general limitation: measurement relies on assumptions about the observation feed, so historical relationships do not establish future results.

Verification or next question

To independently verify your measurements, publish (in your own notes) the gap definition and the exact fields computed: price boundaries, gap size, and timestamps. Then test stability by changing one parameter at a time (for example, the threshold or time window) and observe how much the gap counts and sizes change. If results swing dramatically, the measurement is likely capturing data artifacts rather than a consistent liquidity discontinuity.

If you want a deeper comparison, a useful next question is how your chosen definition behaves during volatile periods and whether your dataset reflects the same trading activity you intend to measure.