Kalshi and Polymarket.
Gabby Jones | Bloomberg | Martin Lelievre | Getty Images
Volume has exploded on prediction market platform Polymarket since its launch, with notable surges during the 2024 election and last fall. Yet the majority of closed individual markets on the site never surpassed $10,000 in reported volume.
A CNBC analysis found that about 70% of all closed markets on Polymarket saw under $10,000 in reported volume from 2021 to the end of May this year, according to Polymarket’s Gamma API. The Gamma API records notional volume on both sides of the trade.
Fewer than 10% of all closed markets attracted between $100,000 and $1 million in reported volume.
Over 45,000 markets, or nearly 5% of all closed markets, had no reported volume whatsoever.
Kalshi, Polymarket’s competitor, also had a large number of shallow markets, according to an analysis done on the on-chain platform Dune. Unlike Polymarket’s Gamma API, Kalshi’s notional volume on Dune only counts one side of the trade.
Low volume markets are not ideal for prediction market traders. For starters, prices can fluctuate widely, said Constantin Bürgi, a professor of economics at University College Dublin.
“Thin markets by nature imply that small investments can result in large market movements and are typically more volatile,” Bürgi told CNBC.
New traders can also be left vulnerable in low-volume markets because spreads between buying and selling can blow out, making trades more expensive, said Eric Zitzewitz, professor of economics at Dartmouth College.
Less appealing
Thin markets are also less appealing to seasoned traders.
“I like higher volume, short term [markets],” said 26-year-old Logan Sudeith, a former financial risk analyst based in Atlanta, who started trading full-time on prediction markets last fall. “It’s more capital efficient.”
Markets lasting up to a week had the highest number of contracts with at least $1 million in reported volume on Polymarket. These week-long markets had contracts related to the war on Iran, U.S. President Donald Trump or Elon Musk.
Short-term markets can be a sweet spot for traders, as are those with a large of number of participants.
“People like to trade things close to being resolved,” Zitzewitz, the Dartmouth academic said last month, noting that participants are “more likely to trade in a market [when] lots of people are there.”
Bots dominate shallow markets
Within Polymarket, over 80% of volume in markets under $10,000 comes from bots, said Joshua Della Vedova, a business professor at the University of San Diego.
Della Vedova identified wallets, or digital accounts, as bots if they made more than 50 trades per day or more than 1,000 total trades.
Using this definition, Della Vedova found bots made roughly $1.2 million in shallow markets and roughly $35.1 million in markets that had more $10 million in volume from November 2022 to February 2026.
“They are making money across all markets,” Della Vedova said, using Polymarket’s on-chain data. That contrasts with retail traders, who have faced losses in shallow or heavy markets.
While bots dominate the volume in shallow markets, they don’t push prices away from fair value since there is a risk of high losses. Bots on Polymarket also prefer heavier markets over thinly traded ones because the ultimate goal is earning a profit, the UC San Diego professor added.
“[Bots] make money per transaction, and therefore they prefer to trade in these larger markets, but they will trade across the whole spectrum,” Della Vedova said.
Accuracy on thin markets
Experts are mixed on whether thin markets are also accurate.
Evercore ISI strategists found high-volume markets have more reliable probabilities than thin markets after analyzing five years of completed markets on both Polymarket and Kalshi.
After finding only 8% of markets touched $1 million in volume on both platforms, the strategists said “most quoted probabilities sit in the thinly traded tail – where calibration is weakest.”
Other researchers said the relationship between market size and accuracy is non-linear. For Theis Ingerslev Jensen, a Yale University professor of finance, accuracy is driven by who is trading rather than how much trading is occurring on a given market.
Jensen and researchers at the London Business School, found skilled or informed traders drove the majority of the accuracy on Polymarket.
“Thin markets are not automatically inaccurate, but they are less reliable,” Jensen told CNBC. “The key question is whether skilled traders still have enough incentive and ability to trade.”
Influence remains
The abundance of shallow markets on both platforms are unlikely to affect how prediction markets operate for the general public and Wall Street, said Harry Crane, professor of statistics at Rutgers University.
“The volumes traded on those markets should be taken into consideration,” Crane said, but “the lack of liquidity, on its own, does not discredit a market’s signal or make the market economically useless.”
Polymarket declined to comment and Kalshi did not respond to requests on CNBC’s findings.
As prediction market volume continues to grow at a breakneck pace, Crane expects bigger markets will expand while low-volume markets stay shallow. What matters is that traders are aware of the risks.
“Protect yourself at all times,” Crane added. “Each individual entity needs to address them on their own.”
Methodology:
CNBC pulled all closed market data via Polymarket’s Gamma API from 2021 to the end of May 2026. The Gamma API counts notional volume on both sides of a trade. For this reason, CNBC wrote “reported” volume coming from the Gamma API.
This analysis on Polymarket was reviewed by Joshua Della Vedova, a business professor at the University of San Diego. He cross-checked our findings with an independent, on-chain trade dataset of 222 million resolved on-chain Polymarket trades. Our findings aligned with Della Vedova’s.
