What “execution quality” means for raw spread brokers
Execution quality describes how closely an actual trade price matches the price implied at the moment an order is sent or quoted, after accounting for all costs that affect the realized result. For raw spread brokers, “raw” typically means the broker passes through a market-derived bid/ask spread and may add its own commission and execution-related effects. In practice, you judge execution quality using outcomes you can measure: how often orders fill, at what prices they fill, and what total cost you effectively pay.
How to evaluate execution quality: observable factors
1) Fill rate and fill certainty
Start with whether orders are filled as intended. Compare intended order type (for example, market vs. limit), order size, and time-in-force with what actually happens: partial fills, missed fills, or fills that occur much later than expected. A lower fill rate can still produce good average prices, but it indicates a controllable operational weakness.
2) Slippage and price improvement
Measure realized trade price versus a reference price. Common references include the last quoted bid/ask at order entry time, or the quote at the moment the broker accepts the order. Slippage can be positive (price improvement) or negative (worse than reference). Evaluate both the average and the distribution (for example, how often trades fall beyond a tolerance band). This separates “usually fine” from “sometimes bad.”
3) Latency and timing consistency
Execution quality is affected by how quickly an order reaches the matching or dealing process and how consistently the system performs under similar conditions. Without real-time data feeds, you can still assess timing using platform timestamps and the event sequence visible in order/trade history. The goal is not to prove causation, but to identify whether execution delays and variability are material.
4) Total cost: spread + commission + execution effects
Raw spread wording can obscure total cost. To evaluate cost, compute an all-in effective spread or cost per trade using the realized entry and exit prices plus commissions/fees shown in statements. This requires explicit assumptions for every calculation (for example, whether you include overnight financing, taxes, or only execution-side charges).
5) Order handling behavior
Different systems handle market and limit orders differently, especially during fast market moves. Look for patterns such as frequent partial fills, repeated requotes or deviations, or systematic differences between order sizes. These are mechanical behaviors you can detect from your own order history.
Evidence and example approach (with assumptions)
Assume you place a batch of N orders of the same instrument and similar size within comparable market conditions. For each order i, record:
- Order entry timestamp (as shown on the platform)
- Reference price at entry (using the platform’s displayed quote)
- Realized execution price
- Commission or fee charged for that order
Then define two metrics:
- Realized slippage = (execution price − reference price) adjusted for whether the trade is buy or sell
- Effective cost per trade = slippage impact plus commission per unit
Because market conditions change, you must control assumptions: N orders should be grouped into windows with similar volatility/liquidity (as best as you can infer from your records), and you should not mix very different periods. Even with careful grouping, you are measuring realized behavior, not a guaranteed property.
Limitations and common failure modes
Market microstructure effects
Execution quality can worsen when spreads widen, liquidity thins, or price moves rapidly. Even if a broker’s mechanics are stable in calm conditions, realized outcomes can degrade during stress. This is a material limitation: stable execution performance in one regime does not ensure performance in another.
Reference-price mismatch
If the reference price is not synchronized with when the order was actually tradable, slippage measurements can be misleading. Platform-displayed quotes may lag behind what the matching system sees, so “slippage vs. displayed quote” is a measurable proxy, not a perfect benchmark.
Incomplete visibility into internal routing
You may not see how orders are routed, whether they are queued, or what venues are used. As a result, you can verify realized outcomes (fills, timing, realized prices) but not fully verify internal causes.
Historical performance is not predictive
Past fill quality and slippage distributions do not establish future results.