tickdistill-learn

Big Orders in Order Flow: Aggregated vs Separated (and Why the Difference Matters)

By TickDistill — order-flow microstructure signals. Educational content, not financial advice.

What is a “big order” in order flow?

In a market’s trade tape, most prints are small. Occasionally a much larger quantity hits the book at once — a big order. Traders watch these because a large, aggressive execution can mark where a serious participant is willing to commit capital.

But “a big print appeared” is, on its own, almost meaningless. The useful question is who sent it and what it implies. That is where the distinction between aggregated and separated big orders comes in.

Aggregated big orders: large, but ambiguous

An aggregated big order is what most tools show by default: many orders in the same direction hitting the same price level (or sweeping several levels) in the same instant, merged into one large figure.

Aggregated activity is still useful — but for a different purpose (see sweep maps below).

Separated big orders: smaller, but high-conviction

A separated big order is a single order from a single participant — one “mouse click,” one decision.

Intuition: aggregated answers “how much hit the tape?”; separated answers “how many distinct actors committed?”.

Why one big order still isn’t a signal — but density is

A lone big order, separated or not, rarely means anything unless it is genuinely large, at a notable level, with other confluences. The far more interesting pattern is density: many committed orders accumulating over a short window.

The high-value setup combines three things at once:

  1. Density — many separated big orders cluster in a short time window.
  2. Containment — price barely moves while this happens: the accumulation stays inside a tight range.
  3. Directional extreme — the buy/sell split is unusually one-sided.

That third point is the key, and it is about rarity, not raw percentage. A 60/40 long-vs-short split is normal noise and says nothing. An 80/20 split is rare — it falls outside what you’d normally see — and that is the information: many different committed actors quietly leaning the same way, at the same level, without moving price.

A concrete example

Imagine that in the hour before a market opens, a steady stream of separated big orders comes through, roughly 90% on the short side, while price stays inside a narrow band. That is many distinct, committed participants positioning short quietly — without pushing the market. It is a setup a trader might read as: when the session opens, downside pressure is already loaded. (Educational illustration only — not a recommendation.)

The mirror case uses aggregated density: mapping the price levels where large aggressive sweeps clustered helps locate where stop-losses likely fired or liquidity was taken — zones that can later flip into support or resistance (“liquidity hunt” reasoning).

An honest note on data

Cleanly separating “one actor” from “many” is easiest with order-by-order (MBO) data. From public aggregated-trade feeds you work with taker-side prints: a single large taker print is a strong proxy for a separated order, and a burst of prints is a proxy for aggregated activity. It is a robust approximation — not a claim of perfect order identity. Good signals are explicit about this; we are.

How TickDistill packages this

All thresholds are normalized per market so the same logic reads sensibly across instruments, and recurring mechanical windows (e.g. an index’s market-on-close auction, a perp’s funding settlements) are handled so they don’t distort the statistics.


TickDistill sells clean, computed order-flow inputs — not trading advice or guaranteed alpha. You combine them into your own strategy. Backtests are illustrative and not a promise of future results.