WhitmanTrading

What Is an Intermarket Sweep Order?

Intermarket sweep order is an order type that executes immediately on one venue without waiting to route to a better price displayed elsewhere, because the sender simultaneously sends orders to take those other quotes. It is a declaration that the sender has handled the routing obligation themselves.

A share trades on many venues at once, and a rule requires you to take the best displayed price wherever it sits. The intermarket sweep order is what you use when obeying that rule the slow way would cost more than it saves.

How it works

A price series with an order executing on one venue immediately.
A sweep order executes without routing away. Illustrative chart - not real market data.

Ordinarily an order must respect better quotes elsewhere. If another venue displays a better price, your order is routed there before it can execute where it is.

A steady series where routing to a better venue takes time.
Normally a better quote elsewhere must be honoured first. Illustrative chart - not real market data.

That routing takes time, and in that time the better quote can vanish — so the protection meant to get you a good price can leave you chasing a price that has already gone.

A rising series where a sender assumes the routing duty.
The sender takes that obligation on themselves. Illustrative chart - not real market data.

A sweep order is marked to say “I have dealt with this.” The venue may execute it immediately without checking or routing, because the sender has asserted they are handling the other quotes.

A falling series where simultaneous orders hit several venues.
By firing orders at every venue at once. Illustrative chart - not real market data.

What the sender is actually promising

A choppy series with a single decision printing across venues.
Which is why a large trade prints across all of them. Illustrative chart - not real market data.

They must simultaneously send orders to take every better-priced quote they are stepping over. Not afterwards, not optionally — as part of the same action.

A slow series where venue fragmentation persists.
And different again over a long horizon. Illustrative chart - not real market data.

So the obligation is not waived, it is relocated. The rule still gets satisfied; it is satisfied in parallel rather than in sequence.

A calm series where a single venue holds most of the liquidity.
A quiet stretch hides what it measures. Illustrative chart - not real market data.

Which is also why it is not a retail tool. Using it correctly requires live feeds from every venue and the infrastructure to fire at all of them at once.

A worked example

A stock is quoted on four venues. Venue A shows 100 shares at 10.00, venue B shows 500 at 10.01, and venues C and D each show 1,000 at 10.02.

A buyer wants 2,000 shares now. Routing sequentially — A, then B, then C, then D — means each leg takes time, and the later quotes can move before the order reaches them.

A falling series with a stop level marked.
A stop fills where the market is. Illustrative chart - not real market data.

Instead they fire four sweep orders at once, one per venue, sized to the displayed quantity at each. All four execute in parallel, and the full 2,000 is filled at the displayed prices.

The trade prints on four venues in the same instant. One decision, four executions, and a tape that shows what looks like four separate buyers.

Why it exists at all

Because a market is many markets. The same share trades simultaneously in a dozen places, and none of them is the market — the “price” is an aggregate of separate books that are always slightly out of step.

And because best-execution rules are per-quote, not per-market. A rule protecting the best displayed price has to be enforced against every venue displaying one, which means either sequential routing or parallel sweeping. There is no third option.

Sequential routing was the original answer and it created its own abuse. An order visibly working its way from venue to venue announces itself, and the quotes ahead of it can move before it arrives.

The sweep exists to close that window. Whether it fully closes it is a live argument in market structure, and the honest answer is that it narrows the window rather than eliminating it — the orders still leave at slightly different times and arrive at slightly different speeds.

The original data

On this site’s shared series a round trip costs 0.0098, about 2% of the median bar range of 0.493. The ninetieth percentile bar is 1.101 and the largest is 2.338.

A sweep is an attempt to keep an execution inside that ordinary cost rather than paying the much larger cost of a price that moved while the order was in transit.

A candlestick chart annotated with the cost of a round trip.
A round trip costs a share of a bar. Illustrative chart - not real market data.

And direction runs average 2.01 bars with a longest of 11. On a bar chart those timescales look generous; a sweep operates several orders of magnitude below the smallest bar on any chart a person looks at, which is why none of this is visible in ordinary price data.

A price series with volume shown beneath.
Volume and price measure different things. Illustrative chart - not real market data.

What it looks like on the tape

A burst of prints at several prices within a fraction of a second. That pattern is often one order, not many participants agreeing about something.

Which matters for anybody reading volume. A sweep can look like a surge of independent interest and be a single decision, so treating the print count as a count of participants overstates what happened.

And it interacts with payment for order flow. Retail orders are frequently internalised rather than sent to a venue at all, so the displayed book a sweep interacts with is not the whole market either.

The practical takeaway for a private trader is interpretive, not operational. You will not send one of these. You will see their effects constantly in the tape, and knowing what produced the pattern stops you reading intent into it that was never there.

When it fails

The characteristic failure is reading a sweep as conviction. A cluster of aggressive prints across several venues looks like urgent, broad demand, and a trader takes it as a signal that informed money is buying. It may be one algorithm completing a routine order in the only way the rules permit. The pattern that looks like many participants agreeing is, structurally, the signature of a single order obeying a best-execution obligation — and no amount of staring at the tape distinguishes the two.

A candlestick series with a gap through a level.
A gap skips the level entirely. Illustrative chart - not real market data.

A second failure is assuming your broker offers it. Most retail platforms do not, and do not need to.

A third is believing it exempts anybody from best execution. It relocates the obligation; it does not remove it.

A fourth is treating the displayed book as the whole market. Much volume never reaches a lit venue at all.

A declining series cut short at a decision point.
Four venues printed at once. One buyer or four? Illustrative chart - not real market data.

And a fifth is thinking any of this is optional. Fragmentation is a fact of the venue structure, and every large order has to deal with it somehow.

Order types covers the wider set of instructions and what each gives up. Order book covers the per-venue stack this order hits in several places at once. And payment for order flow covers why much retail volume never reaches those books.

What I actually do

This is the order type that explains why market structure looks the way it does. A rule that says you must take the best displayed price is obviously right and creates a timing problem, and the sweep order is the machinery built to live with both.

— Michael Whitman

This page is educational, not financial advice. Test every idea on your own charts before risking money.