What “Broker Pricing” and “execution quality” mean
Broker pricing is the price information and execution path a broker offers when you place an order. Execution quality describes how closely the outcome of your order execution matches the execution you would expect from transparent inputs like quoted prices, order rules, and known costs.
A useful way to think about it is separation:
- Stable mechanics: the order handling logic (how quotes become fills), latency behavior, and how costs are applied.
- Variable conditions: market movement, liquidity changes, and temporary order-book gaps.
Because you can’t control market variability, assessment relies on measuring differences (expected vs realized) while stating assumptions.
Measurable factors to evaluate
You can assess execution quality by tracking several observable components, each tied to an assumption.
1) Spread and effective transaction price
Quoted spread is not always the same as the effective spread you experience.
- If you expect a fill near the mid/quote, compute the realized difference using your execution timestamps and the relevant reference quote.
- Assumption: you use a defined reference (e.g., at decision time) and a consistent rule for converting buys/sells.
What to measure:
- Difference between reference price at submission and average fill price.
- Include all explicit costs you pay (commissions, fees), not only the displayed spread.
2) Slippage under defined conditions
Slippage is the gap between an expected execution price and the realized fill price.
- Assumption: “expected” is based on a specific reference time and price definition.
- Compare slippage distributions across multiple runs.
You should look for patterns such as:
- Systematic slippage away from the reference (not just one-off events).
- Larger slippage during periods with low liquidity or rapid quote changes.
3) Latency and timing consistency
Execution quality is affected by how quickly the system receives your order and processes it relative to quote updates. What to measure:
- Time from order submission to fill confirmation (using logs you can record).
- Consistency: whether delays vary widely across order size, time of day, or order type.
4) Order type behavior (market, limit, stop)
Different order types transform quotes into fills differently.
- For limit orders, execution quality includes how often you get filled, and at what price relative to your limit.
- For market orders, execution quality includes how the broker selects liquidity and how costs and spread translate into the final fill.
Assumption: you interpret each order type using its standard rule set (e.g., limit won’t trade beyond your limit for the intended direction, but market microstructure still matters).
Evidence and examples you can run without relying on promises
A practical test approach is to create controlled observations:
Example: “expected vs realized” calculation (with stated assumptions)
Assume you place an order at time T when a reference quote shows bid/ask. You submit a buy at T and receive fills at prices p1…pn.
- Define a reference rule: for a buy, expected reference = ask at time T.
- Compute realized average fill = average of p1…pn (weighted by fill sizes).
- Compute price error = realized average fill − expected reference.
Assumption: your reference quote corresponds to the market state at T under a consistent definition.
Then repeat across multiple runs and record:
- Mean and spread of the price error.
- Whether errors correlate with fast-moving conditions (proxy: quote update frequency, without assuming causality).
Evidence limitations you must account for
Even with careful calculations, mismatches can occur:
- Your reference quote may not be the same as the venue-level best price used for execution.
- Timestamps may be delayed, rounded, or measured differently across systems.
So the goal is not certainty about “who caused what,” but a grounded estimate of how outcomes deviate from your defined expectation.
Limitations, risks, and common failure modes
Execution quality can fail in multiple ways; at least one of these should be on your checklist.
1) Slippage and adverse selection
When liquidity disappears or the market moves between submission and fill, slippage increases. This can happen even if the broker is behaving consistently.
2) Partial fills, requotes, and non-intuitive outcomes
Some orders may fill in parts, reject, or behave differently than a naive “one price becomes one fill” model. Failure mode examples:
- Partial fill: you end up with multiple prices and need correct averaging.