How to evaluate execution quality: 7 metrics beyond the spread
A competitive spread gets attention, but it does not tell you the full cost of turning an order into an executed position. Execution quality requires data, context and like-for-like comparisons.
Execution quality is not simply a matter of speed. The European best-execution framework considers price, costs, speed, likelihood of execution and settlement, size and the nature of the order together.[1] For a trader, this leads to a practical question: is the result consistent, measurable and repeatable under the conditions in which I trade?
1. Slippage: examine the distribution, not just the average
Slippage is the difference between the available or requested price when an order is sent and the price at which it is ultimately filled. A monthly average may hide important details. Positive, negative and zero slippage should be separated by instrument, time window, direction, order size and order type.
A serious assessment should also consider the median, percentiles and outliers. Two setups with the same average may behave very differently when volatility rises.
2. Fill rate and rejections
An excellent theoretical price has limited value if a meaningful proportion of orders is not executed. Measure how many orders receive a full fill, how many are partially filled and how many are rejected, including the stated reason.
The result must be interpreted in relation to available liquidity and ticket size. Comparing a small order in a major currency pair with a large order during an illiquid period would lead to misleading conclusions.
3. End-to-end latency
Total time should be measured from order transmission to fill confirmation. Latency may arise from several components, including the client connection, distance to the server, platform, routing infrastructure and execution venue.
A single number is therefore insufficient. Median values and spikes are more useful, particularly when assessing stability over time and during fast markets.
4. Executed price and price improvement
Not all slippage is necessarily negative. A symmetrical process may also deliver fills at a better price than the one observed when the order was sent. The frequency of price improvement can help explain how order flow is handled.
Any comparison requires synchronised timestamps and a consistent price source; otherwise, differences caused by data feeds or system clocks may be mistaken for execution effects.
5. Total transaction cost
Spread, commission, slippage, financing costs and market impact all contribute to the final result. The advertised minimum spread may not represent the spread actually available for a given size or during the hours in which the strategy operates.
6. Behaviour during volatility and limited liquidity
Data collected under normal conditions tells only part of the story. Market openings, macroeconomic releases, rollover periods and gaps can change depth, spreads and execution probability. Identical conditions cannot always be expected, but consistency and transparency can still be assessed.
Trades executed during stress windows should therefore be isolated, comparing slippage, latency and rejection rates with ordinary sessions.
7. Transparency of the execution model
Documentation should explain clearly where and how orders may be executed, which factors influence venue selection and how potential conflicts are managed. Under MiFID II, European investment firms must establish an execution policy, monitor its effectiveness and provide appropriate information to clients.[1]
This does not mean that one venue will always be best for every order. It means that the process should be structured, verifiable and appropriate to the client and instrument. In 2025, ESMA published its final report on the criteria firms should apply when establishing and assessing the effectiveness of their order-execution policies.[2]
A simple comparison method
- Collect a sufficiently large sample of real or test trades.
- Record order, acknowledgement and fill times; requested and executed prices; quantity; order type; instrument; and rejection reasons.
- Segment the data by session, size and volatility.
- Compare like-for-like setups, avoiding different periods and strategies.
- Assess total cost, stability and execution probability together.
The conclusion
The best execution is not necessarily the lowest number in a marketing table. It is the process that delivers the most appropriate overall outcome for the order and remains measurable over time. Before changing broker, platform or infrastructure, define which metrics matter to your strategy and how you will measure them.
Sources
- ESMA — MiFID II Article 27: obligation to execute orders on terms most favourable to the client. Official text covering execution factors, execution policies and monitoring.
- ESMA — Final report on order-execution policies, 10 April 2025. Criteria for establishing policies and assessing their effectiveness.
- EUR-Lex — Directive 2014/65/EU (MiFID II). Full official text.
Raimondo Perfetto
Head of Business Development. I work across liquidity, trading technology, execution and relationships with professional and institutional clients.