PRO RESEARCH / EXECUTION
A strategy can be directionally right and economically untradeable. Trading desks judge a model by what can be executed at a given size and time, not by midpoint returns alone. This guide explains the cost stack, where it becomes nonlinear and how to audit fills.
Research status: Educational framework only. No verified live quote, broker fill, investment recommendation or commercial endorsement.
Build a consistent round-trip cost stack
For a one-unit illustrative trade, suppose gross expected price capture is $0.26, round-trip spread impact $0.08, commissions $0.05, adverse slippage $0.04 and financing $0.01. Net modeled capture is $0.08 per unit. Omitting the cost components exaggerates expected capture by $0.18.
The exact allocation of spread depends on whether the backtest starts from mid, bid, ask or actual fill. Do not double-count the spread if the simulator already fills at executable bid and ask. Always identify whether fees are one-way or round-trip.
Why large orders have nonlinear costs
Bid-ask spread is only the visible minimum friction. Available top-of-book quantity can be smaller than the intended order. Sweeping multiple levels creates price impact; queuing as a maker creates adverse-selection and non-fill risk. Crypto venues may additionally use maker/taker tiers, rebates and liquidation mechanics.
Execution quality should therefore be analyzed by order size, time of day, venue, volatility regime and whether a signal remains valid while the order sits unfilled. The best theoretical entry may have the worst practical fill probability.
Measure what was actually executed
- Record decision timestamp, order submission, acknowledgement and complete fill times separately in UTC.
- Capture mid, bid/ask, executable depth and contract size at the decision time.
- Separate forecast price movement from implementation shortfall and commissions.
- Store partial fills, cancellations and venue failures rather than deleting rejected orders.
- Compare median and tail costs by session, not just an overall average.
Pro stress scenario
Suppose a strategy’s gross average outcome is +0.18R. Ordinary all-in costs are 0.09R, leaving +0.09R. In a high-volatility event window, effective costs double to 0.18R and the modeled net edge becomes zero. If the spread also widens enough to prevent entry, the missing trade must be represented in any empirical study.
A high win rate does not rescue negative net expectancy. Use the execution-quality analysis tool only as a model of entered assumptions, not as a claim that a broker will fill at a particular price.
Cross-check your assumptions
- Trade cost model — Model the full cost stack
- Trading fees explained — Read the cost components
- Provider evaluation — Verify venue and broker documentation
Editorial framework · Version 2026-11-05 · Hypothetical examples only · Financial markets involve material risk. Inputs and data rights must be verified independently.
ILLUSTRATIVE RESEARCH GRAPHIC · NOT LIVE DATA
Gross edge vs execution drag
Charts here explain a method. They do not use a licensed market feed, a live quote, realized investment returns or an audited performance history.

