What “fixed spread” means in measurable terms
“Fixed spread” is commonly used to describe a pricing arrangement where the spread is intended to remain constant over some conditions. To measure it, you need a definition that turns the concept into a number you can record. A practical measurement definition is:
- Fixed spread (measured spread) = ask price − bid price at a specific moment.
Because quotes and fills happen at different times, you should also define whether you are measuring quote-time spread (based on a displayed bid/ask) or fill-time spread (based on the actual execution prices). These are often different, especially when markets move or execution takes time.
Mechanism: what you record to measure it
Use the same “measurement fields” every time.
- Instrument: the exact currency pair or symbol (and contract specification, if your platform distinguishes them).
- Venue / data source: where the bid and ask come from (for example, your platform quote feed). Measurement is only comparable if the source is the same.
- Timestamp definition: decide what “time” means. Options include:
- Quote timestamp: the time shown or logged when bid/ask were captured.
- Request timestamp: when you request an order.
- Fill timestamp: when the order is filled.
- Bid and ask values: record both prices, not only the spread, so you can recompute spread = ask − bid.
- Unit and normalization: spreads may be shown in points, pips, or absolute price terms. Pick one representation and convert consistently using the instrument’s price format.
Two comparable measurements
To explain the idea clearly, you can report two numbers:
- Quote-time measured spread: spread computed from bid/ask captured at the chosen timestamp.
- Fill-time realized spread: spread implied by the fill prices (using the actual execution prices for bid and ask sides).
If “fixed” is meaningful in practice, quote-time measured spread should be stable across the observation window, and fill-time realized spread should not deviate materially from that stability.
Evidence and example: comparing results across time windows
Assume you want to check whether the spread you observe is stable.
Example setup (no live data required):
- Pick one instrument.
- Define observation windows, such as 10 recordings spread across a period.
- At each recording, capture bid, ask, and the timestamp.
- Compute spread = ask − bid for each recording.
Then compare:
- Stability of quote-time measured spread: are the computed spreads identical (or within a very small rounding threshold) across recordings?
- Consistency with fill-time realized spread: if you also observe execution outcomes, compare fill-time realized spread to quote-time measured spread.
Material limitation: quote-to-fill differences
Even if the pricing arrangement is described as “fixed spread,” the realized result can differ because the market can move and your execution can occur after the quote you captured. This can show up as:
- a change between quote-time and fill-time spread,
- differences due to order handling (for example, delays or re-pricing),
- additional costs that are not part of the bid/ask spread.
Therefore, your measurement should explicitly state which time you used and whether you included only the bid/ask component or also other trading costs.
Limitations and risks: what can break the measurement
Several limitations affect measurability, even when the term “fixed spread” is used.
-
Stable mechanics vs variable conditions A pricing label may describe the intended spread behavior, but your observable spread can still vary with conditions not fully controlled by the quote display (timing, execution latency, and market movement).
-
Fees and commissions vs spread A “spread” measurement based on ask − bid does not automatically include separate commissions or any other trading costs. Two pricing setups can have the same bid/ask spread but different total costs.
-
Rounding and display vs internal values Platforms may display prices with limited decimals. If internal pricing uses finer precision, the displayed spread may look “fixed” while the underlying values differ slightly.
-
Failure mode to look for A key failure mode is spread widening between the time you record a quote and the time you fill an order. Your measurement method should be able to detect that by comparing quote-time and fill-time computations.
Verification and next question
To independently verify fixed spread behavior, use a repeatable recording method: