Open a Polymarket page and the first number most traders look at is the one with the dollar sign next to it. "$125,745 Vol." sits under the title like a verdict on how serious the market is. The internal value behind that display, verified on Polymarket's own rendered page, is 125744.818466. A child market on the same page prints "$36,601" against an internal value of 36601.299831. The figures match. The dollar sign does not mean what most people read it to mean.
That gap matters because volume is the number traders use to decide whether a market is worth their time. If you think the headline reads as cash that changed hands, you will overestimate how deep a market is, how much money it would take to move the price, and how much skin is actually in the game. The truth is more interesting, and it falls cleanly out of how a binary prediction market is built.
Deposit $20, get $50 to trade
Welcome bonus on the world's largest prediction market.
18+ · New users only · Availability and terms vary by region · Trading involves risk of loss · BeGambleAware.org
The headline is a share count, not a cash figure
Here is the mechanic. On Polymarket, shares in an outcome trade between 0.00 and 1.00. Each pair of opposing shares (one YES, one NO) is fully collateralised by $1.00 USDC, and shares are created when the two sides' offers add up to that dollar. Winning shares pay out $1.00 each at resolution. That much comes straight from Polymarket's help centre, which lays out the mechanic in its "What is Polymarket" and "How Are Prices Calculated?" pages.
So one share is worth, at most, one dollar. The venue reports volume as the count of shares traded, printed with a dollar sign because a share's face value is a dollar. It is a convention, and iPredicta follows it too.
But face value and cash paid are different things.
What our own check found
iPredicta pulled the full trade history of 10 Polymarket markets on 29 August 2026, priced across the full range from 2.1 cents to 98.2 cents, and summed the numbers two ways. The first sum: the displayed volume figure divided by the number of shares traded. That ratio came in between 0.919 and 1.000, with nine of the ten markets at 0.992 or above and one matching to four decimal places. In plain English, the headline really is counting shares, each worth close to a dollar of face value.
The second sum is where the story sits. We divided the displayed volume figure by the actual cash paid by traders. The ratio ran from 1.03 to 3.36. Broken out by price:
On the three markets priced between 2.1 and 4.6 cents, the ratios were 1.03, 2.82 and 3.36. On the five markets priced between 48.5 and 61.5 cents (the mid-price band), the ratios ran from 1.71 to 2.29, clustering around 2. On the two markets priced between 95.3 and 98.2 cents, the ratios were 2.12 and 3.18.
So the displayed figure ran from roughly equal to the cash on one market, to about three and a third times the cash on another. The gap is not a clean function of price. You cannot read it off the midpoint. On these 10 markets, the only way to know the real cash figure was to reconstruct the trade history.
Why both sides of a trade put money in
This is the part most explainers skip, and it is the whole reason the share count and the cash come apart. A trade on Polymarket is not one trader giving another trader a few cents. It is both sides contributing collateral to back a new share pair.
Take a market where YES is priced at 3 cents. A trader who wants to buy 1,000 YES shares pays $30. The trader on the other side of that trade is buying 1,000 NO shares at 97 cents, and they pay $970. Between them, $1,000 goes into the collateral that will eventually pay out the winning side. Reading the activity as "1,000 shares at 3 cents, so $30 of trading" misses $970 of real money. The buyer of the long-shot YES contributed almost nothing in cash terms; the buyer of the near-certain NO did almost all of the work.
At 50 cents, both sides pay $500 to create 1,000 shares each way, so the symmetry is clean. At the extremes, one side carries almost all the cash. And which shares actually trade (YES or NO, at what prices, at what sizes) determines the cash number. A single share count cannot tell you any of that.
Why the convention is defensible
It would be easy to call this misleading and move on. It is more useful to see why the venue does it this way. A share in the correct outcome pays $1.00 USDC at resolution. A share in the wrong outcome pays nothing. So a share has a face value of exactly one dollar, bounded and known. Quoting activity in face-value dollars is the cleanest, most consistent unit the venue has: it does not depend on the current price, does not require reconstructing historical cash flows, and settles unambiguously at resolution.
What volume is still good for
If the displayed figure is not cash, what is it useful for? Comparing activity. A market showing $5m of face-value volume has had more shares change hands than one showing $500k, so it has seen more trading. What volume does not tell you is how much you could trade at the current price: that is in the order book, in the orders resting at each price. For the question of how much a market has traded, volume works as intended.
It just is not a cash figure. If you want to know how much money actually went through a market, you need the trade history, and on these 10 markets the cash number was smaller than the headline every single time. Our guide to how prediction market odds work covers the pricing side of this; the volume side is the other half of the same mechanic.
Trade 400+ coins
The premier crypto platform, trusted by 100M+ users.
18+ · Eligibility and terms apply · Crypto is volatile; capital at risk
How to read a Polymarket contract properly
The practical upshot is modest. When you see a figure like "$125,745 Vol." at the top of a market, read it as a count of shares traded, each with a face value of a dollar. It is not the cash that changed hands. On every one of the 10 markets we reconstructed, the cash was smaller than the displayed figure, by an amount you could not estimate from the price alone, because it depends on which side of each trade carried the weight.
Our own figures show why no rule of thumb works. On the five mid-price markets, the displayed figure ran from 1.71 to 2.29 times the cash. On the three long-shots it ran from 1.03 to 3.36, and on the two near-certainties it was 2.12 and 3.18. Treat the headline as the liquidity signal it is, not as the cash figure it looks like.
Our Kalshi vs Polymarket page compares the two venues. This piece covers Polymarket's volume figure only.
iPredicta tracks Polymarket markets and surfaces the ones worth watching, with the mechanics explained so traders can read a contract page for what it actually says. The volume number is one of those details that looks obvious and is not. It rewards a second look.
Frequently asked questions
Does the dollar sign on Polymarket's volume figure mean dollars of cash?
No. The dollar sign reflects face value, not cash paid. The figure is a count of shares that have traded, each of which has a face value of up to $1.00 USDC because that is what a winning share pays at resolution. When iPredicta checked 10 Polymarket markets on 29 August 2026, the displayed volume figure was larger than the cash that actually changed hands on every single one. The ratios ranged from 1.03 to 3.36 times the cash. Reading the headline as dollars-paid overstates how much real money moved through the market, sometimes by a little, sometimes by more than three times.
Why is there a gap between the displayed volume and the cash that was paid?
Because both sides of every trade put money in, and the headline counts shares, not the cash each side paid. On Polymarket, each pair of opposing YES and NO shares is fully collateralised by $1.00. If YES trades at 3 cents, a buyer of 1,000 YES shares pays $30 and the counterparty buying 1,000 NO shares at 97 cents pays $970; between them, $1,000 of real cash goes in. A share count does not show how that cash was split between the two sides. Which side carried the weight on which trade is what determines the true cash figure, and the volume display does not record that.
Can I work out the cash figure by dividing volume by the current price?
No, and this is the trap. The gap between displayed volume and cash is not a function of the current price. On the 10 markets iPredicta reconstructed, the ratio of displayed volume to cash ran from about 1 to about 3.4 times, and sat around 2 at mid prices, but with wide spread at every price band. Three markets priced between 2.1 and 4.6 cents showed ratios of 1.03, 2.82 and 3.36. The only reliable way to know the cash figure for a given market is to pull its trade history and sum the actual payments. There is no shortcut multiplier.
Is Polymarket's volume figure wrong or misleading?
It is a convention, not an error. Every share on Polymarket has a face value of $1.00 USDC because that is what a winning share pays at resolution, so quoting activity in face-value dollars is internally consistent and settles unambiguously. iPredicta follows the same convention. It is defensible. It just does not answer the question most traders think it answers. Treating it as a cash figure overstates how much real money moved through a market, by an amount that cannot be read off the price and that varies widely from one market to another.
If volume is not a cash figure, what is it useful for?
Comparing activity across markets. A contract showing $5m of displayed volume has had far more shares change hands than one showing $500k, so it has seen more trading. How much you could trade at the current price is a different question, answered by the order book, not by volume. Volume is still the headline activity number for a reason. The caveat is only that it answers 'how much activity has this market seen' rather than 'how much cash has gone through this market'. For the first question, it works. For the second, you need the trade history.