Execution isn’t just technology. It’s incentives. Your broker, venue, and order type decide who gets filled first, who gets re-quoted, and who gets “last looked.”
This pillar is about designing rules that survive imperfect fills. If your system requires perfect execution, it’s not a system. It’s a fragile wish.
Queueing: why “my level” didn’t fill
In many markets, especially in fast conditions, price is not a single point. It’s a queue of available liquidity. When you place a limit, you join the queue. When you place a market order, you cross the queue and pay for immediacy.
- Limit: price control, uncertain fill.
- Market: fill certainty, uncertain price.
- Stop: conditional market order when triggered.
In stress, the queue changes faster than your screen. That’s why “it touched my level” is not a guarantee.
Last look, re-quotes, and rejection in plain English
Some execution models allow liquidity providers to reject your trade if the market moved during a short window. This is often called “last look.” Retail traders experience it as:
- re-quotes,
- fills that are worse than expected,
- orders that “fail” during fast moves,
- asymmetric outcomes (you don’t always get positive slippage, but you often get negative).
The correct response is not outrage. It’s adaptation: change when you trade, how you trade, or how much you trade.
Partial fills and the discontinuous risk problem
Thin liquidity can produce partial fills. That creates a dangerous state: you believe you are hedged or exited, but you are partially exposed. Discontinuous exposure is how “one small mistake” becomes a large loss.
Microstructure realities that show up in retail platforms
- Indicative prices: what you see may not be firm in fast markets.
- Minimum stop distance: brokers can enforce buffers that change your intended risk.
- Order type restrictions: certain orders may be disabled around events.
- Price source discretion: different feeds can be used for execution vs display.
Designing an execution-aware strategy
1) Build a spread gate
Define a threshold above which you will not trade (or will trade smaller). This is a rule that turns “bad conditions” into a non-event.
2) Build a speed gate
Fast tape is expensive tape. If you don’t have infrastructure to compete, don’t pretend you do. Your system can explicitly avoid the fastest minutes.
3) Use size as the universal adapter
Size is the only lever that works in every scenario. You can’t negotiate with liquidity, but you can reduce your footprint.
4) Separate entry logic from execution logic
Your signal can be correct and your entry can still be poor. Treat execution as its own subsystem: how you stage orders, how you scale in, and how you exit when the market is fast.
Practical adaptations (small, but powerful)
- Stagger entries: split size into two tranches to reduce queue pain.
- Use “no-trade” windows: rollover, opens, major releases, thin holidays.
- Prefer robust setups: avoid strategies that require 2-pip precision in a market that breathes 10 pips.
- Assume worst-case on stops: stops are for survival, not for precise exits in a spike.
Measurement: stop arguing with your memory
Track execution like a professional:
- expected entry/exit vs actual,
- spread at the time,
- session/time,
- notes on tape speed.
After a few dozen trades, you’ll know when your broker and your strategy are aligned, and when you’re paying invisible rent.
Order types in practice (and what they really buy you)
Market orders
Markets buy certainty at the cost of price. They are appropriate when the edge comes from being in the move, not from being filled at a specific tick. If your edge is fragile to 1–2 pips, market orders will eventually expose that fragility.
Limit orders
Limits buy price at the cost of certainty. In fast markets, the “best” price is often the one that never fills. Limits are excellent for mean-reversion entries and for scaling, but dangerous when you need guaranteed exit.
Stop orders
Stops buy participation in momentum but surrender price control. When triggered, stops usually become market orders. In gaps, they execute at first available prices. Treat them as catastrophe tools, not as precision tools.
Stop-limits
Stop-limits try to blend both: trigger like a stop, then limit the worst price. The tradeoff is brutal: you can avoid a terrible fill, but you can also fail to exit. That means stop-limits are not “safer.” They are different.
Event-window protocol (what to do when the market is discontinuous)
When a decision or release is imminent, you face a choice:
- Hold risk: accept gap/volatility risk and size for it.
- Reduce risk: cut size so a worst-case fill is survivable.
- Step aside: let liquidity return, then trade the post-event structure.
The mistake is holding full size while expecting brochure execution. In discontinuous windows, your execution model becomes the primary risk factor.
Exit priority: the order of survival
- First: prevent forced liquidation (margin buffer).
- Second: prevent thesis drift (don’t widen stops to avoid being wrong).
- Third: accept imperfect fills as the cost of certainty.
Professionals don’t demand perfect exits. They demand survivable exits.
Mobile reality (yes, it matters)
If you trade from a phone, you must assume you are slower than the market at the worst times. That’s not a moral judgement. It’s physics. Your rules should compensate:
- avoid the fastest windows,
- trade smaller,
- use wider tolerances,
- prefer setups that don’t require split-second management.
In FXRISK Manual, execution literacy is not optional. It’s how you avoid learning microstructure through losses.
When the correct fix is a different venue
If your strategy’s edge comes from tight stops, rapid entries, or trading around releases, then broker quality and execution model are part of the strategy. A broker that is “fine” for swing trading can be lethal for event execution. The clean rule is:
- If your results depend on microstructure, you must measure microstructure (fills, rejects, spreads by regime).
- If the measurements are consistently bad, you either change the strategy to fit the venue, or change the venue to fit the strategy.
Trying to force precision out of a venue that doesn’t offer it is like trying to do surgery with boxing gloves.
Note: Execution-aware risk notes. Not signals. Not advice. Assumes you can lose everything.