Institutional Execution Considerations

At retail size, execution is a rounding error. At institutional size it competes with the signal for control of the outcome. The variables below are the ones that change most between a backtest and a fill.

Liquidity is fragmented and time-varying

Forex has no single venue. Liquidity is distributed across banks, ECNs, and aggregators, and the depth available at a price is not the depth advertised. Practically:

• Depth concentrates during session overlaps and thins at session boundaries.
• The seconds around scheduled releases can see liquidity withdrawn entirely.
• Quoted spread and effective spread at your size diverge as size grows.

The platform publishes the current spread alongside each quote and flags market sessions precisely because these conditions are observable rather than assumable.

Spread behaviour across sessions

Spreads follow a predictable daily shape — tighter during major overlaps, wider at the rollover boundary and in thin holiday conditions. A strategy whose average edge is smaller than the spread differential between session peak and trough is really a strategy about when it trades.

Slippage and market impact

Two distinct costs:

• Slippage — the difference between decision price and fill price, driven by latency and momentary depth.
• Market impact — the price movement your own order causes, which grows super-linearly with size relative to available depth.

Impact is why large orders are worked over time rather than sent at once, and why a strategy that is profitable at one size can be unprofitable at ten times that size with identical signals.

Implications for evaluating a strategy

When reviewing published performance, ask which assumptions about execution are embedded. Results measured in pips from mid-price do not include spread, and results at small size do not extrapolate to large size. This is not a criticism of any particular record — it is the standard adjustment any institutional allocator applies.

Key takeaways

• Liquidity is fragmented, time-varying, and thinner than quoted at size.
• Session timing changes cost structure enough to change strategy viability.
• Impact grows faster than size; results do not scale linearly.