What “Spread By Session” means (and why verification matters)
“Spread By Session” refers to the idea that a forex spread can differ across trading sessions (for example, based on overlap between regional markets). A “spread” is the difference between the quoted buy and sell price at a given moment; when spread is analyzed “by session,” the analysis groups observations into specific time windows and compares the spread statistics across those windows.
Verification matters because session-based comparisons are highly sensitive to how you define sessions, how you collect quotes, and whether you include relevant costs (such as commissions) that may not be part of a displayed spread.
A source hierarchy you can apply before trusting any claim
Use a hierarchy from most stable and reproducible to most interpretation-based:
-
Primary definitions and methodology: Look for documentation that defines the exact session time windows and the calculation method (for example, which timestamps are used, what “spread” includes, and what statistical summary is reported).
-
Data provenance: Prefer information that specifies the quote source and sampling approach (for example, whether observations come from tradeable quotes, indicative quotes, or a specific platform feed).
-
Provider or product documentation: If the claim is tied to a specific provider or platform, verify it using their legal/technical documentation describing how spreads are presented or computed.
-
Independent replication: Treat third-party charts or comparisons as starting points. Verify by recreating the same steps with the same session definitions and consistent assumptions.
-
Historical context only: If the information is only presented as “recent history,” assume it is conditional. Historical relationships do not prove future behavior.
Reproducible verification steps (no real-time data required)
To verify “Spread By Session” information, you can use a reproducible workflow based on assumptions you explicitly state:
-
Lock the session definitions: Decide what counts as each session (including start/end times and time zone). If a source uses a different time zone, results may shift even if the underlying market behavior is unchanged.
-
Choose what “spread” means in your check: Decide whether the spread you verify is “raw bid/ask difference” only, or whether your definition includes commissions or other execution costs. If a source does not clarify this, treat comparisons as non-equivalent.
-
Collect or use a quote series with timestamps: You need bid and ask values (or equivalent) with timestamps. If you cannot get live data, you can still verify methodology by applying the stated calculation to any saved quote dataset.
-
Compute session-grouped statistics the same way: For each session window, compute the summary your source claims (for example, average spread, median spread, or a chosen percentile). Use the same aggregation period (minute-by-minute, tick-by-tick, or sampled intervals) as described.
-
Check sensitivity with controlled variations: Repeat the calculation using small changes to sampling interval or session boundaries. If results flip dramatically, the claim may be fragile.
-
Document assumptions for every step: Your verification should include assumptions such as time zone handling, clock alignment, missing quote handling, and whether spread values are filtered or unfiltered.
Limitations and failure modes to look for
At least one material limitation is usually present in session-based spread claims:
-
Time-window mismatch: Different sources can define sessions differently (or use different time zones). This can create apparent “session effects” that are really boundary artifacts.
-
Sampling and timestamp alignment: If one dataset samples quotes every second and another uses every minute, the computed session averages can differ.
-
What costs are included: A displayed “spread” may exclude commissions, while another source’s “all-in cost” concept includes them. That makes the comparison non-direct.
-
Execution and quote feed differences: Two providers can publish different quote behaviors even for the same instrument. Even if the definition is “spread,” the observed values may reflect different quote generation.
-
Conditional pattern breaks: Session effects can weaken during low liquidity periods or change during volatility spikes. Also, historical patterns do not guarantee future behavior.
Verification or next question to resolve ambiguity
If you want to move from concept to a confident understanding, focus on these next verification questions:
- **Which exact session time windows and time zone were used? **
- **Does the source define spread as bid/ask only, or as all-in cost?