What “Spread By Session” means
“Spread by session” is a way of describing how the bid–ask spread changes across trading sessions (for example, when different regional markets are active). The spread is the difference between the buy price and the sell price at the moment quotes are available, and it effectively becomes one part of trading costs.
A key point: session-based comparisons are about patterns in liquidity and quoting during time windows. They do not guarantee a fixed cost at any given time, and they don’t remove other cost drivers (execution quality, additional charges, and market conditions).
Common mistakes, their consequences, and neutral checks
1) Confusing a definition with an outcome
Mistake: Treating “spread by session” as if it describes a predictable trade cost for a specific moment. Consequence: You may underestimate the time- and condition-dependence of spreads, then form unrealistic expectations. Neutral check: Write down the exact assumptions behind any estimate (e.g., “assume spread equals quoted spread, ignore other fees”). Then test whether those assumptions could break during low liquidity, volatility, or changing quoting conditions.
2) Using session averages as if they are transferable
Mistake: Relying on historical relationships (like “this session is usually tighter”) as if they apply to the next day or the next week. Consequence: Results can differ when volatility regime shifts or when liquidity changes. Neutral check: Separate “typical historical behavior” from “current expected conditions.” If you cannot justify that conditions match the historical sample, treat the estimate as uncertain.
3) Mixing spread with total transaction cost
Mistake: Assuming spread alone represents costs, while ignoring other cost components such as commissions, financing-related charges, or execution effects. Consequence: Your cost picture can be incomplete, leading to wrong comparisons across times or execution methods. Neutral check: If you compute an “all-in cost,” explicitly list what you include: quoted spread, any commissions/fees, and the difference between quoted prices and execution prices (often discussed as slippage). If a provider’s “spread by session” figure doesn’t include these, don’t combine it implicitly.
4) Ignoring failure modes (when “session” liquidity breaks)
Mistake: Overlooking cases where spreads widen unexpectedly despite being inside a “normally liquid” time window. Consequence: Realized trading costs may spike, and attempts to rely on session timing can fail. Failure modes to consider: temporary illiquidity, aggressive market moves, sudden quote changes, and the possibility of delayed or less favorable execution. Neutral check: Ask what happens during volatility spikes and around high-impact events, and whether the data you’re using reflects those periods.
5) Not distinguishing quote timing from your trade timing
Mistake: Comparing a published “spread by session” statistic to your own execution time without verifying time alignment. Consequence: Even if the statistic is correct for the stated window, your trades might occur during different micro-intervals. Neutral check: Confirm the timing basis used in whatever data you consult (session boundaries, time zone conventions, and sampling method). If those don’t match your trading timestamps, the comparison is unreliable.
Limitations and risks you can’t eliminate
Even with a clear definition, “spread by session” is an observational cost proxy that depends on variable conditions. Historical patterns do not establish future outcomes. Additionally, total cost depends on more than the bid–ask spread, including execution quality and any additional charges.
If you want independent verification, focus on checks rather than certainty: confirm definitions, align time windows, and account for other cost components. When these assumptions don’t hold, expect the realized cost to differ from what “spread by session” alone suggests.
Verification and next questions
If you’re trying to use spread-by-session information safely (without assuming predictable results), ask:
- Which exact cost component is being measured: quoted spread only, or all-in transaction cost?
- What time zone and session boundaries define each “session”?
- Does the data include periods of volatility and illiquidity, or only typical market hours?
- How are execution differences treated (if at all)?
A practical next step is to compare at least two independent views of spreads for the same session window, using clearly stated assumptions, and then evaluate where your estimate becomes uncertain.