More than $1.6 million in trading volume has already been wagered on the outcome of North Carolina’s US Senate race this fall on Kalshi alone. The market development comes after traders generated $10,982 in trading volume on who would replace Rep. Chuck Edwards as the Republican nominee in North Carolina’s 11th Congressional District earlier this month.
The activity follows North Carolina’s decision to give prediction markets a more formal role through this year’s state budget, which created a new tax framework for the industry.
As markets expand further into North Carolina politics, they are raising questions about how they differ from traditional sports betting and what happens when the people with the best information about an outcome may also be driving a prediction market.
Trading on NC politics
On predictive market platforms, like Kalshi and Polymarket, traders can buy and sell contracts tied to everything from economic indicators and weather to sporting events and elections.
Prediction markets generally use contracts tied to whether an event or outcome will occur. For example, Former Democratic Gov. Roy Cooper trading at 93 cents on Aug. 20 can be interpreted as the market assigning roughly a 93% probability that he will win the November election. Those prices can change continuously as traders buy and sell contracts ahead of the Nov. 3 election.
If the contract resolves in the trader’s favor, it ultimately pays $1; if not, it pays nothing. Traders can also sell their positions before an event has concluded.
The Commodity Futures Trading Commission describes event contracts as derivatives, typically structured as swaps, whose value is tied to the outcome of an underlying event. The agency states that prediction markets can be used both to speculate on events and to aggregate information about the likelihood of future outcomes.
That structure is central to the argument that prediction markets should be treated differently from sportsbooks, even when both allow customers to put money behind predictions about future events.
Jacki McGavick, a spokesperson for Kalshi, told Carolina Journal the distinction begins with who is on the other side of the transaction.
“A sportsbook is a bookie: it is the counterparty to every customer bet, and it sets the odds,” McGavick said. “Like the house at a casino, it wins when the bettor loses and vice versa. That gives it an incentive to stack the odds against the customer and embed margins that make losses more likely.”
McGavick added its exchange model removes that direct conflict by separating the platform’s revenue from the side that ultimately wins a trade.
“Kalshi is a financial exchange that has no house,” McGavick said. “We don’t set the lines — traders set prices, the same way prices on the NYSE are set by market participants, not the exchange itself. Every trade is visible on the orderbook, unlike a sportsbook, where customers never see that price movement or market depth. And like a traditional financial market, traders can exit a position at any time.”
McGavick further stressed that Kalshi generates revenue through transaction fees rather than from the side that wins a contract, and that it allows traders to buy and sell positions through an order book, more closely resembling a traditional financial exchange.
Critics, however, have argued that the difference is less meaningful when prediction markets offer contracts tied to the same sporting events available on gambling platforms.
Mick Mulvaney, executive director of Gambling is Not Investing and former White House chief of staff, said in a press release earlier this year that “prediction markets are unlicensed sports gambling apps, full stop,” arguing they should face the same state regulatory structure as sportsbooks.
Regulations in North Carolina
North Carolina lawmakers formally waded into that debate through the state budget signed by Gov. Josh Stein in July.
The budget increased the tax on licensed sports wagering operators from 18% to 23% of gross wagering revenue. Sportsbooks operating in the state are licensed and regulated by the North Carolina State Lottery Commission.
Prediction markets received a different framework.
The budget created a 6% tax on a prediction market operator’s net trading-fee revenue attributable to North Carolina, beginning Jan. 1, 2027. Unlike the sportsbook tax, the prediction market levy is applied to the fees collected rather than to an operator’s gross wagering revenue.
The law also says a prediction market registered with the CFTC may lawfully operate in North Carolina based on its federal registration and compliance with federal commodities law. It specifically states that the state tax does not impose any additional “license, registration, or other regulatory requirements” on prediction market operators.
The North Carolina Fiscal Research Division projected $2.16 billion in trading volume during 2026, with the industry expected to grow substantially over the remainder of the decade. The new tax is predicted to generate about $1 million during the 2026-27 fiscal year.
McGavick said that Kalshi views North Carolina’s new, overall approach favorably. “The state’s explicit recognition adds legal clarity and could serve as a model other states look to as they consider their own approach to prediction markets.”
She added that the 6% rate better reflects its business model because the tax applies to net trading fees rather than sportsbook-style wagering revenue.
Not everyone agrees that the framework strikes the right balance. Former New Jersey Gov. Chris Christie, an adviser to the American Gaming Association, argued in a July Charlotte Observer op-ed that North Carolina gave prediction markets an unfair advantage over state-regulated sportsbooks by subjecting them to a lower tax rate and no comparable state licensing requirements.
“North Carolina has set a horrible precedent: rewarding bad behavior while undermining the state’s authority and putting North Carolina consumers at risk,” Christie wrote.
When the market gets smaller
While a statewide Senate election provides traders with months of polling, fundraising reports, and other public information, the recent fight over the NC-11 Republican nomination demonstrates how prediction markets can operate very differently when the pool of decision-makers becomes much smaller.
Edwards withdrew from his re-election campaign following a House Ethics Committee report finding violations of House rules related to sexual harassment and a hostile work environment. Members of the 11th Congressional District Republican Executive Committee were tasked with selecting his replacement.
Kalshi opened a market on who the committee would choose.
The market initially favored state Sen. Tim Moffitt, R-Henderson. Kalshi price history data show Moffitt at about 51% on Aug. 9, while state Rep. Jennifer Balkcom, R-Henderson, stood at about 11%. By Aug. 10, after Moffitt publicly declined to seek the nomination, Balkcom climbed above 40% while Moffitt fell below 4%.
Balkcom ultimately won the nomination on the first ballot over five other candidates at the Aug. 10 meeting. By the time the Kalshi market settled, it had generated a total trading volume of $10,982.
There is no indication that a candidate, committee member, or anyone else involved in the NC-11 selection improperly traded on the market.
But the contest illustrates the market-integrity questions that can arise when relatively few people control an event’s outcome and may possess information that ordinary traders lack.
Candidates and party officials, for example, could know about private vote commitments, candidate withdrawals, or internal discussions before those developments become widely known.
Kalshi said people who have direct influence over a contract’s outcome are prohibited from trading on that contract and can face discipline or regulatory enforcement if they do.
To enforce this, Kalshi uses political data to identify candidates, campaign staff, election officials, poll workers, and others who may have conflicts with specific markets.
The company also said all customers undergo identity verification and that trading activity is monitored for potential red flags, including unusual timing, win rates, coordinated activity, and patterns linked to employment, relationships, social media, or geolocation.
What counts as inside information?
Despite these safeguards, determining when a trader crosses the line from being better informed to improperly using confidential information can be complicated.
Jeanette Doran, senior legal counsel at the John Locke Foundation, cautioned against simply importing the securities-law concept of “insider trading” into prediction markets.
Generally, simply possessing better information — or even information that is not widely known — is not necessarily prohibited, Doran said.
Determining what information is actually “public” can create another challenge.
Information does not necessarily have to appear in a newspaper or press release to be considered public, Doran said. It could be publicly available but obscure. At the same time, information does not automatically become public simply because it is circulating within a relatively small political or professional circle.
“The relevant questions include how the trader obtained the information, whether it was available outside that circle, and whether it was understood to be confidential,” Doran said.
That distinction also makes suspicious-looking trades difficult to evaluate from price data alone.
A well-timed or highly profitable trade may raise questions, but it does not itself establish that someone improperly used confidential information, Doran added. Regulators would generally look for evidence linking a trader to the information, including communications, relationships, access, timing, and trading patterns.
“In the NC-11 race, for example, a price movement before the committee announced its choice could be consistent with trading on confidential information, but it could also reflect traders making good predictions from public signals,” Doran said. “The key factual question is what the trader actually knew, how they learned it, and whether the trader’s use of it breached any applicable duty.”
As prediction markets expand further into North Carolina politics, the difference between the statewide Senate market and the NC-11 selection demonstrates both the appeal and the challenge of the model.
Prediction markets are designed to reward traders for finding information that the rest of the market has missed. As the events being traded become smaller and more specific, the harder question may increasingly be whether that advantage comes from better analysis — or from information the trader was never supposed to use.
