Aggregate overlapping orders before assessing capacity

The security-window pair is the useful unit

Capacity analysis often begins with a model or account, but the market sees orders in securities over time. Two sleeves can independently pass a participation screen while their combined demand exceeds the same screen. Build a shared view keyed by instrument, side, execution window and accessible route.

Preserve the account and strategy identifiers as well. They explain who owns the instruction and how completed quantities are allocated. Aggregation for risk analysis must not erase ownership or imply that accounts can freely trade with one another. Actual allocation and crossing policies need their own operational and legal review.

Two passing sleeves can produce a failing aggregate

Imagine a hypothetical $50 million combined portfolio. Sleeve A intends a $1.2 million buy and sleeve B a $900,000 buy in the same security during one session. Assumed eligible volume is $15 million. Individually, participation is 8% and 6%; jointly, it is 14%. Under a chosen 10% scenario ceiling, the combined order breaches the ceiling.

With combined fee and half-spread of 3 bps and impact coefficient 20 bps, the aggregate cost rate is 10.483 bps. Applying it to $2.1 million gives approximately $2,201.50. Independently pricing the two orders with their smaller participation inputs would understate the heuristic’s total if both genuinely compete for the same flow.

Offsetting intentions need a separate treatment

Suppose A instead buys $1.2 million while B sells $900,000. Gross intentions remain $2.1 million and net directional intention is a $300,000 buy. A hypothetical perfectly permitted and completed internal transfer might reduce external trading to $300,000, but it is not justified merely by subtracting the two instructions.

For a conservative no-crossing scenario, retain the two external orders and their costs. For a permitted-crossing scenario, separately record the transfer quantity, pricing convention, account approvals and residual external order. Do not present netting as a guaranteed saving, and do not count an assumed internal transfer as exchange volume.

Overlap can be partial

A morning order and an afternoon order may share only part of their execution windows. A daily aggregate can obscure the peak demand. Divide the planned activity into justified time buckets when the available data support it, and aggregate overlapping quantities in each bucket.

A simple illustration has $600,000 from A and $400,000 from B scheduled in a first-hour window with $4 million assumed volume: the bucket participation is 25%. The remaining orders might trade at much lower rates later. A 14% daily figure from the earlier example would miss that first-hour concentration, even though both calculations can be arithmetically correct.

What evidence changes the conclusion

The SEC’s execution description notes that brokers can route to different markets and that execution terms depend on those choices. Different routes may alter access, but they do not prove that two orders reach independent counterparties. Require route and timestamp evidence before treating liquidity pools as independent.

The workbench has no order allocator, matching engine or cross-impact model. Its rows are isolated scenarios unless the user explicitly combines shared demand. A useful review therefore identifies likely overlaps, documents any aggregation assumptions and withholds a portfolio-wide capacity conclusion where ownership or time alignment is unresolved.

Keep in the decision record

• Aggregate by instrument and overlapping window, not only by strategy name.
• Keep gross intentions separate from any authorized net external order.
• Inspect intraday peaks where daily averages hide concentrated demand.

Source notes

1. SEC Investor.gov — Executing an Order
2. FINRA — Rule 5310, Best Execution and Interpositioning

Sources inform the definitions and limitations. Scenario inputs and arithmetic are publication illustrations, not extracted market observations. Access checks are dated in the source inventory.

Continue the analysis