FuturesAI Polymarket Depth: How to Size an Order

A prediction-market contract quoted at 61 cents is not a promise that you can buy $50,000 worth at 61 cents. The quote is for the first few hundred dollars. After that you walk up the book, and on a thin market the walk costs more than the edge you came for.

Here is the sizing math, worked from a published order-book study, with the places where the study’s own numbers disagree with each other pointed out rather than smoothed over. Check it on FuturesAI

Educational content, not investment advice. Prediction-market positions can lose the full amount staked, and leveraged crypto futures can lose the full principal. Figures below come from the cited pages on the dates shown and may have changed.

On This Page

• How deep is the Polymarket order book by market rank?
• What is the rule of thumb for price impact?
• How much does a $5,000 to $250,000 order cost in each tier?
• Why do the study’s measured slippage figures look smaller?
• What happens to the book during a news event?
• The Order-Sizing Checklist

How deep is the Polymarket order book by market rank?

According to FuturesAI’s microstructure study published April 12, 2026, the five most active Polymarket markets carried a median top-of-book spread of 1 cent and average depth of about $420,000 within a 3 cent window of mid, while markets ranked 21 to 50 carried a 3 cent median spread and about $24,000 of depth in the same window.

The study collected Level-2 order-book snapshots every 60 seconds across the 50 most active contracts from January 15 to April 15, 2026, a dataset of 388,800 snapshots. The tier table it published is the raw material for everything below.

Two things stand out. Depth falls about 17 times between the top five and ranks 21 to 50. And the P95 spread, the one you meet on a bad minute, is three times the median even on the best markets.

There is also a concentration point. The study attributes 62% of resting liquidity on the top five markets to three addresses. When those three pull their quotes, the $420,000 of depth is not there.

What is the rule of thumb for price impact?

FuturesAI’s study gives a rough impact rule for Polymarket market orders: expected price impact is roughly the order’s notional divided by the market’s median 3 cent depth, multiplied by 1.5 cents.

In symbols: impact in cents = (notional ÷ depth) × 1.5. A $42,000 order into $420,000 of depth is one tenth of the book, so the rule gives 0.15 cents of impact per share. The logic is that the book is spread over a 3 cent window and a market order sweeps through part of it, paying roughly half the distance it travels on average.

The rule is linear, which is a simplification. Real books thin out as you go deeper, so a very large order costs more than the straight-line estimate. Treat the rule as a floor for big orders and a fair guess for small ones.

How much does a $5,000 to $250,000 order cost in each tier?

Applying the study’s rule, a $50,000 Polymarket market order costs about 0.18 cents per share of impact on a top-five market, 0.79 cents on a market ranked 6 to 20, and 3.1 cents on a market ranked 21 to 50, before the quoted spread.

The table below is my own arithmetic from the rule and the depth figures above. For the tail tier the study only says depth is “under $5,000,” so I used $5,000, which makes the tail numbers a best case. Dollar cost assumes a contract priced near 50 cents, so a $50,000 order is roughly 100,000 shares.

Read it against your edge. If your view is worth 2 cents of mispricing, a $50,000 market order on a rank 21 to 50 market spends 3.1 cents of impact plus a 3 cent spread to capture it. You are paying 6 cents to collect 2. The same order on a top-five market pays about 1.2 cents all in.

Most people skip this step and discover it on the fill report.

The $25,000 row is where the study draws its first line: it says anything under $25,000 notional fills cleanly on top-20 markets. The arithmetic agrees, with impact under 0.4 cents in those tiers. Between $25,000 and $250,000 the study recommends working the order over 3 to 10 minutes with limit orders. Above $250,000 it points to request-for-quote flows or direct contact with market makers.

Why do the study’s measured slippage figures look smaller?

FuturesAI’s study separately reports measured slippage of roughly 4 basis points for a $50,000 market order on a top-five market and about 18 basis points for $250,000, which is about one tenth of what its own rule of thumb produces when expressed against a 50 cent contract.

When I compared the two, I noticed that 0.18 cents of impact per share on a 50 cent contract is 0.36% of notional, or about 36 basis points, not 4. The rank 6 to 20 figures show the same gap: the study quotes about 12 basis points for $50,000, while the rule gives 0.79 cents, or roughly 1.6%.

I cannot tell from the page which figure is the measured one and which is the estimate, and the page does not say what contract price the basis-point figures assume. So the honest reading is a range. The quoted basis points are the optimistic end, probably from fills that got the benefit of resting liquidity refreshing during the sweep. The rule is the pessimistic end. Size to the pessimistic end and be pleasantly surprised.

If you’ve ever back-tested a strategy using the quoted spread as the only cost, you know the feeling when live fills come in worse. This is that gap, measured.

For sizing by edge rather than by liquidity, the quarter-Kelly post covers the position-size side. The order-flow absorption post covers reading the book for direction rather than cost.

What happens to the book during a news event?

During one scheduled poll release the study tracked, top-five Polymarket market spreads widened from 1 cent to 9 cents within 30 seconds and tightened back to 2 cents over the following 6 minutes, according to FuturesAI’s April 12, 2026 study.

That is a nine-fold spread for half a minute. A market order sent at the headline pays the 9 cent spread plus whatever depth is left after the three big addresses pull their quotes. The study’s advice is to scale in over 5 to 10 minutes rather than hit size on the release, and the tightening timeline supports it: by minute six the spread was back near normal.

Here’s where people get tripped up. The news is the reason they want the position, so they size up exactly when the book is thinnest. Pre-placing limit orders inside the normal spread, then letting the event come to them, costs patience rather than cents.

One sentence on fees, since they stack on top of impact: Polymarket’s fee documentation, read November 1, 2026, lists taker fees on most market categories using the formula shares × rate × price × (1 − price), with makers charged nothing. A worked order that costs 1.2 cents in spread and impact can cost noticeably more once that fee is added at mid-range prices. That is a separate calculation, and the maker-versus-taker choice is one more reason to work orders with limits.

The Order-Sizing Checklist

1. Find the market’s rank by volume. Top 5, 6 to 20, 21 to 50 or tail decides the depth you can expect.
2. Estimate the 3 cent depth, or use the study’s tier averages if you cannot see Level 2.
3. Run the rule: notional ÷ depth × 1.5 cents. Then add the median spread.
4. Compare the total with your expected edge in cents. If cost is more than half the edge, cut size or switch to limit orders.
5. Under $25,000 on a top-20 market, a market order is fine. Between $25,000 and $250,000, work it over 3 to 10 minutes.
6. Around scheduled news, place limits early and wait at least 5 minutes after the release before adding.

Before you size a position, run the numbers above on the actual market, then check the current partner-exchange terms. Check it on FuturesAI

Tips

• Use P95 spread, not the median, when sizing for a volatile session.
• Limit orders inside the spread make you the maker, which also avoids the taker fee.
• Keep a note of the three-address concentration: depth that belongs to a few makers can vanish together.

Warnings

• The tier figures are averages over January to April 2026 and will not match any single market today.
• The impact rule is linear and understates cost for orders that are a large share of the book.
• This is educational content, not investment advice; prediction-market and futures positions can lose the full amount committed.

FAQ

How large an order can I place on Polymarket without moving the price?

FuturesAI’s study of the top 50 markets (January 15 to April 15, 2026) says orders under $25,000 fill cleanly on top-20 markets, and its impact rule gives under 0.4 cents of impact at that size in those tiers.

What is the Polymarket price impact formula from the study?

Expected impact in cents is roughly the order’s notional divided by the market’s median depth within 3 cents of mid, multiplied by 1.5 cents, so a $50,000 order into $420,000 of depth costs about 0.18 cents per share.

Should I use market or limit orders on Polymarket?

For sizes between $25,000 and $250,000 the study recommends working limit orders over 3 to 10 minutes, and limit orders also avoid Polymarket’s taker fee because makers are not charged under its fee documentation.

Sources

• FuturesAI Insights: Polymarket Liquidity, a Microstructure Study of the Top 50 Markets (April 12, 2026)
• Polymarket Documentation: Fees (read November 1, 2026)
• FuturesAI Insights: Polymarket vs Kalshi comparison (April 22, 2026)

Fact-checked against the cited pages as of November 1, 2026. Not investment advice.