What Data Is Needed to Assess Liquidity Definition?

Liquidity definition: what it means before using it

Liquidity generally describes how easily an instrument can be bought or sold with limited price impact. In forex research, “liquidity definition” typically means the specific, measurable concept you choose (for example, ease of execution, depth, trading activity, or price responsiveness) and the exact way you translate that concept into data and thresholds.

To assess a liquidity definition accurately, you need to separate (1) the stable mechanics of how liquidity is defined and measured from (2) the variable conditions that affect market microstructure and trading costs.

What data is needed (inputs)

Assessing a liquidity definition starts with identifying the data elements used to operationalize it:

Price and order/quote observations

Mid-price or reference price at regular times.
Bid/ask quotes (to represent execution frictions via spread).
For models that require microstructure depth, you may also need order book snapshots (or a proxy for depth).

Execution-related measures (if your definition is execution-focused)

Trade sizes (or executed volume at the time of trades).
The relationship between trade size and subsequent price movement.
Measurement horizon (the time window after a trade or quote observation).

Activity and participation (if your definition is activity-focused)

Number of trades and traded notional over time.
Concentration measures (for example, whether activity is dispersed or dominated by a few large trades), if available.

Transaction costs and constraints

Spread statistics (average, median, and variability).
Commission/fees and other execution costs, if you are assessing real execution, not just market quotes.
Margin or leverage constraints can affect what participants can execute, so note any execution constraints relevant to your analysis—without assuming they apply everywhere.

Jurisdiction and venue assumptions (metadata you attach to the definition)

The market venue or data feed the observations represent.
The time zone, trading session coverage, and whether data excludes off-market periods.

Provenance, timeliness, and quality checks (how to verify your inputs)

A liquidity definition is only as reliable as the data provenance, timing alignment, and validation you apply.

Provenance checks

Source identity: Document where the data comes from (for example, quote feeds vs trade data vs computed vendor aggregates).
Coverage and completeness: Confirm whether missing intervals exist and how they are handled.
Reproducibility: Ensure someone else can repeat the same transformations using the same raw inputs.

Timeliness checks

Update frequency: Quote-based measures require consistent sampling; trade-based measures require correct event timestamps.
Synchronization: If you combine multiple data streams (e.g., quotes with trades), verify time alignment.
Staleness awareness: Definitions can change in meaning if the data is outdated relative to the market regime you claim to assess.

Quality and calculation checks

Explicit assumptions: State assumptions for any calculation (e.g., how you define “liquid” threshold, how you treat outliers, and the measurement horizon).
Sensitivity analysis: Test whether results change materially when you vary sampling intervals, horizons, or threshold choices.
Internal consistency: If your definition implies low price impact, verify that your observed measures are consistent with that implication.

Evidence or example: how definitions are operationalized

A common assessment approach is to define liquidity using a measurable output, such as:

Spread-based liquidity: Use bid/ask spread statistics as the friction proxy.
Depth/impact-based liquidity: Use order book depth (or an impact proxy) to estimate how prices move when trade size increases.

In either case, the evidence needed is the dataset that supports the operational measure: the relevant quote/trade observations plus the transformation steps that map observations to the liquidity metric.

Limitations and failure modes (what can go wrong)

At least one material limitation is usually unavoidable:

Variable market conditions Liquidity can differ across sessions, volatility regimes, and major news periods. A definition assessed on one period may not generalize.
Provider and venue differences Different data sources can compute or aggregate quotes/trades differently. Even when two sources are both “market data,” their operational definitions may diverge.
Execution vs observation mismatch Quote-based liquidity does not guarantee that an order can be executed at those quotes. If your definition is execution-focused, ensure your data truly represents tradable outcomes.
Costs and constraints omitted If you ignore commissions, swap/financing effects (where applicable), or execution constraints, the practical meaning of “liquidity” may be misrepresented.