Definition and what you are actually assessing
Execution price is the effective price at which an order is filled. In practice, you often care about it as what you paid/received for the traded quantity, after accounting for the market quote convention and any execution-related costs that are determined by the fill.
To assess execution price accurately without assuming any real-time data, you need a clear set of inputs, their provenance (where they came from), and how timely or stale they may be.
Inputs you need (the core “data checklist”)
- Order mechanics (stable inputs)
- Instrument identification: the currency pair (e.g., EUR/USD) so the quote convention is unambiguous.
- Order side and type: buy vs sell, and whether the order is intended to execute at the market’s prevailing quote or under specific constraints.
- Requested quantity: the amount that the order attempts to trade, because partial fills can change the effective outcome.
- Expected direction of pricing: for FX, execution outcomes depend on whether you pay the ask (buy) or receive the bid (sell). Define this before combining any numbers.
- Market reference data (variable inputs)
- Bid/ask quotes used for execution reasoning: either the quote stream you will treat as the reference, or the broker/platform “execution report” fields that describe what quote the system used.
- Time reference: the timestamp associated with the fill (or with the quote you model). Execution price is time-dependent, so the exact reference moment matters.
- Liquidity and volatility context (as inputs, not predictions): you do not need to forecast, but you need to know that the relevant period may have had wider spreads or less depth.
- Execution-related costs (often misunderstood components)
- Spread component: the difference between bid and ask at the relevant moment.
- Commissions/fees: any fixed or per-trade costs that affect the effective price you should compare against your reference.
- Financing/adjustments (only if applicable to your comparison): if you are assessing “effective” economic price over a holding period, document the rule set you are using; otherwise focus on the fill price itself.
Provenance: where each input came from
For each data item above, record provenance:
- Source of the order fields: the order ticket/order request log vs a summary view.
- Source of bid/ask or execution quote: whether it came from a market-data feed you control, or from execution reports provided by a provider/platform.
- Source of timestamps: confirm the time zone/format and whether timestamps reflect “when the quote was generated” or “when the fill was recorded.”
If you mix sources with different conventions (for example, one source timestamps in UTC and another in local time), you can still compute something, but it may not correspond to the true execution moment.
Timeliness and quality checks (how to verify inputs)
At minimum, apply these checks:
- Completeness check: you must have (a) instrument, (b) side, (c) fill quantity, (d) the quote basis (bid or ask), and (e) the relevant timestamp.
- Consistency check: verify that buy/sell logic matches bid/ask usage. If side is “buy,” ensure your reference uses ask; if side is “sell,” ensure bid.
- Staleness check: if quotes are taken from an earlier time than the fill timestamp, treat that as an assumption and label the resulting execution price as an estimate based on stale data.
- Aggregation check: if there are multiple fills (partial fills), compute execution price as a quantity-weighted outcome, and state whether your comparison uses total filled quantity.
Evidence or example (with explicit assumptions)
Assume a simplified scenario with two fills for the same instrument and order:
- Fill 1: buy quantity 1,000 units at an ask reference of 1.0950.
- Fill 2: buy quantity 2,000 units at an ask reference of 1.0960.
- Assume no commissions and define execution price as the weighted average of the filled quantities using the ask for buys.
Then the quantity-weighted execution price is:
- (1,000 × 1.0950 + 2,000 × 1.0960) / 3,000
- = (1.0950 + 2 × 1.0960) / 3
- = (1.0950 + 2.1920) / 3
- = 3.2870 / 3
- = 1.095666…
Key point: this example is only valid under the stated assumptions about bid/ask convention, fee handling, and using the quote levels associated with each fill moment.