Definition and what “spread by session” means
Spread by session refers to comparing bid–ask spreads at different times (often called sessions), using a consistent method. A bid is the highest price a buyer is willing to pay, and an ask is the lowest price a seller is willing to accept. The spread is the difference between ask and bid for the same instrument at a specific moment.
When someone says “worked example of spread by session,” they usually mean: pick a way to label time windows (for example, Session A, B, C), set clear assumptions, then calculate the spread for each window so you can see how the comparison would be done.
A worked numeric scenario (with every assumption stated)
Assume the following fixed setup:
- The instrument is the same across all sessions.
- “Session spread” is the spread observed at one sampled moment per session (not an average).
- We use the same formula each time: spread = ask − bid.
- We assume bid and ask values (because no real-time market data is used).
Now define three sessions:
- Session A: a time window you label as “A.”
- Session B: a different time window labeled “B.”
- Session C: a different time window labeled “C.”
Assumed prices:
- Session A: bid = 1.1000, ask = 1.1004
- Session B: bid = 1.1002, ask = 1.1006
- Session C: bid = 1.0999, ask = 1.1011
Calculate spread in each session using ask − bid:
- Session A spread = 1.1004 − 1.1000 = 0.0004
- Session B spread = 1.1006 − 1.1002 = 0.0004
- Session C spread = 1.1011 − 1.0999 = 0.0012
Interpretation using only what the numbers show:
- In this scenario, Sessions A and B have the same spread (0.0004).
- Session C has a wider spread (0.0012), meaning the bid–ask gap is larger in that labeled time window.
A common alternative definition is an average spread per session (for example, average of multiple samples within the session). If you use that, you must state how many samples you took and at what frequency, because those choices can change the result even under identical market conditions.
How the mechanics affect the outcome (and what you can verify)
To independently verify the calculation method, check these mechanics:
- Same instrument and same quote convention: spreads are not comparable if one session uses a different quoting format.
- Same calculation: always use ask − bid for the chosen sample definition (single sample vs average).
- Consistent session boundaries: if Session C starts/ends differently, you may sample different conditions.
Two “variable” factors you should treat as uncertain unless you measure them:
- Market liquidity and volatility vary by time. That can change bid–ask spreads across sessions.
- Provider pricing and execution vary. Different providers can show different spreads at the same time.
Therefore, “spread by session” is not a single universal number; it is a result of your definition (session windows) and measurement method (single snapshot or averaged samples).
Limitations and failure modes
Material limitations in a worked example like the one above:
- Single-sample limitation: using one snapshot per session can misrepresent a session’s typical conditions.
- Sampling bias: choosing times “that look different” can overstate session effects.
- Changing conditions: spreads can shift quickly; historical relationships do not guarantee future similarity.
- Execution and costs mismatch: the displayed spread may not reflect the effective cost after slippage, commissions, or other provider-specific fees (if any exist).
A practical failure mode is confusing spread (ask − bid at a moment) with overall trading cost (which can include multiple components). If you treat them as the same, your comparison may be wrong.
Verification or next question to make it self-checking
If you want your own self-contained verification, you can replicate the same arithmetic with your own assumed or observed bid/ask samples:
- Pick session boundaries and state them.
- Decide whether you use one sample or an average.
- Record bid and ask at each chosen sample time.
- Compute spread = ask − bid, then compare across sessions.
If your next question is, “How can I choose session windows responsibly?