WhitmanTrading

What Is an Iceberg Order?

Iceberg order is a large order that displays only a small portion of its size in the public book, with the remainder hidden and released in further slices as each visible portion fills. The displayed quantity understates the real size, which is the reason the order type exists.

An iceberg order is a large order wearing a small one’s clothes. Understanding it is the fastest way to stop over-reading the depth displayed on a chart.

How it works

A price series with partially displayed resting size.
An iceberg order shows only part of itself. Illustrative chart - not real market data.

A large order is entered with a display quantity. Only that portion appears in the public book; the rest is held by the venue and is not shown.

A steady series where a level refills repeatedly.
The rest refills as each slice fills. Illustrative chart - not real market data.

When the visible slice fills, another appears. The level seems to replenish, which is what gives the order type its name - most of it is below the surface.

A rising series where displayed depth understates reality.
So the book understates what is there. Illustrative chart - not real market data.

The hidden portion usually loses queue priority. Most venues place a refreshed slice behind orders that were displayed in full, which is the cost of concealment.

A falling series where concealment is the purpose.
Which is the whole point of using one. Illustrative chart - not real market data.

The problem it solves

A choppy series where large resting size would move price.
It looks different in a choppy market. Illustrative chart - not real market data.

Showing a very large order moves the market. A displayed order for a hundred thousand shares tells everybody what is coming, and price adjusts before the order is filled.

A slow series where a position is built over months.
And different again over a long horizon. Illustrative chart - not real market data.

So the order is broken up. Showing a thousand at a time lets it sit in the book without announcing its full size, and the execution proceeds without the impact.

A calm series where the level is untested.
A quiet stretch hides what it measures. Illustrative chart - not real market data.

It is the same problem a dark pool addresses, solved differently. One hides the order on a separate venue; this one hides most of it on the public book.

A worked example

An institution wants to buy 200,000 shares. It enters an iceberg with a display quantity of 1,000 at a fixed limit price.

The book shows 1,000 bid at that level. Somebody sells into it, the level fills, and a moment later 1,000 appears again at the same price.

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

Repeat that two hundred times. The level absorbs every seller who arrives, and from outside it looks like a price that simply will not break.

It is not support in any analytical sense. It is one participant with a fixed limit and a lot of size, and when the 200,000 is done, the level offers nothing at all.

What you can and cannot infer

A level refilling repeatedly is a hint. Size appearing again and again at the same price is consistent with hidden quantity behind it.

It is not proof. Several separate participants replacing orders produce the same pattern, and the book does not label which is which.

And the hint expires without warning. The hidden quantity is finite, and nothing signals its exhaustion - the level simply stops refilling.

Which is why “the seller is done” is a conclusion only available afterwards. Before that, a level holding and a level about to give way look identical from outside.

The trade-off the user accepts

Concealment costs queue position. On most venues a refreshed slice joins the back of the queue at that price, behind orders that were displayed in full.

So the order fills more slowly than a fully displayed one would. That is the price of not announcing the size, and it is paid on every refill.

The display quantity is the dial. A larger visible slice fills faster and reveals more; a smaller one conceals better and waits longer.

Which makes it a judgement about urgency. Somebody who needs to be done today shows more; somebody with a week shows less - and neither setting is correct in general.

The original data

On this site’s shared series: median bar range 0.493, ninetieth percentile 1.101, largest bar 2.338. Direction runs average 2.01 bars with a longest of 11. A round trip costs 0.0098, about 2% of the median bar range.

The 0.0098 round trip is what makes hidden size expensive to probe. Testing whether a level has more behind it means trading, and each test costs about 2% of a median bar whether the answer is useful or not.

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 a 2.01-bar average direction run limits how long an inference survives. Whatever you conclude about a level from watching it refill, the market’s direction has usually changed before the conclusion is old.

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

What this means for reading depth

Displayed size is a lower bound, not a measurement. There can always be more behind a price and never less, which makes the book asymmetric as evidence.

So a thin-looking level may be deep. Absence of displayed size is not absence of interest, and acting as though it were is the mirror image of the same error.

Retail platforms usually offer the order type too. The display quantity is a field on the order ticket at many brokers, so it is not exclusively an institutional tool.

And the useful conclusion is modest. Depth tells you what is shown; it cannot tell you what is available, and any level-two reading that forgets this is reading a number somebody chose to publish.

When it fails

The characteristic failure is trading against a wall that is not there. A large displayed order looks like a barrier, and positions get taken on the assumption that price cannot pass it.

Displayed orders can be cancelled instantly. The wall disappears the moment it is approached, and the trade was built on a number that carried no commitment - the opposite error from missing hidden size, and produced by the same mistake of treating the book as a measurement.

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 a refilling level is one participant. Several separate orders look the same from outside.

A third is expecting the level to hold indefinitely. Hidden size is finite and gives no notice when it runs out.

A fourth is inferring intent from size. An index fund rebalancing and somebody with a strong view produce identical icebergs.

A declining series cut short at a decision point.
The reading is clear. What does it leave out? Illustrative chart - not real market data.

And a fifth is building a system on displayed depth. The input is a number participants choose, which means the system is measuring a decision about disclosure rather than a fact about supply.

Order book covers the displayed depth this order type conceals. Limit order covers the order type an iceberg is a variant of. And liquidity covers the difference between displayed and available size.

What I actually do

This is the single best reason not to trust the depth on your screen. The book shows what people are willing to display, which is a different thing from what they are willing to trade. Every conclusion drawn from displayed size - a wall of sellers, a support level with real buying behind it - is drawn from a number that anybody can choose to understate.

— Michael Whitman

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