The Problems with Prediction Markets, Part 1: Are Consumers Really Protected?

Prediction markets are all the rage, but they are fraught with concerns

It seems like all the talk in gambling circles these days is about prediction markets and why American gamblers seem to love them so much. Sites like Kalshi and Polymarket offer everything from elections, current events to stocks and commodities, economic indicators, weather, cryptocurrency, technology and science and, of course, sports essentially allowing users to wager on the outcome of real-world events. The controversy relating to bets on invasions and war is well documented but for states and Tribes the biggest concern are sports markets. In this two part article I will look at all the issues surrounding prediction markets and why they may not be the panacea that so many are making them out to be.

Responsible Gambling and minimum age concerns

In 41 of the states with commercial casino gambling the minimum age to gamble is 21, whether betting at a casino or wagering on sports. This has been generally accepted as the norm and often tied to the legal drinking age in the state, particularly due to the availability of alcohol in casinos. But users of prediction markets can "invest" at the age of 18 regardless of where they are offered. Prediction markets are regulated by the CFTC and federal law classifies prediction markets as financial derivatives exchanges rather than traditional gambling platforms and the minimum age for investing in exchanges is 18. If one excluded sports betting there may be an argument for the minimum age, but responsible gambling groups claim that sports contracts at prediction sites like Kalshi or Draft Kings Predictions are exactly the same as a sports bets at Draft Kings, so trying to classify them as different in any way is just wrong.

Aside from the fact that it is a state rule often tied into alcohol use, many RG groups claim that age 21 is important because from 18 to 21 youth are more susceptible to addiction. They point to research papers which shows that the part of the human brain that is linked to impulse control and decision making continues to develop until the early 20s. Consequently at the ages of 18, 19 and 20, people are more prone to act irrationally when it comes to gambling and then develop pathological behaviors as they go into adulthood. Studies show that at that age, young adults are more desperate for money and believe that gambling will be a good way to address issues surrounding debt including student debt for university students despite the fact that gamblers will always lose in the end.

But it’s not just youth who are at risk with prediction markets because gambling by its nature can be addictive, which is why sportsbooks are required to put in strict rules to address RG concerns including spotting and stopping betting that appears to be compulsive and also have programs and methods for those who are concerned about their betting to self exclude. This isn’t the case with prediction markets. At last check there was no way for anyone to self-exclude from a place like Kalshi or even DraftKings Predictions and there is no indication anyone has a role at those companies whose job it is to sport compulsive behavior. So if someone is betting erratically on sports at DraftKings they will likely be spotted and contacted, but if someone is betting erratically at DraftKings predictions on sports markets it will be ignored because they are simply investing.

Sports betting disguised as trading

To anyone who bets on sports it is obvious that sports contracts and sports betting are exactly the same thing, but the prediction market sites and the CFTC argue they are swaps and hence fall under the purview of the CFTC. The definition of a swap according to Investopedia is "a derivative contract where two parties exchange cash flows or liabilities of financial instruments, often over-the-counter (OTC) or on SEFs. Interest rate swaps, among the most common forms, typically involve fixed and variable rates between institutions seeking to manage interest rate risk or speculate on market conditions."

So how exactly are sports contracts swaps? The CFTC argues that a contract that pays out based on whether a specific event will happen such as the winner of a game, a point spread or even a player achieving a clearly defined milestone, such as rushing for 100 yards, functions legally as a binary option or event-based derivative and thus is legal per the CFTC rules.

Not surprisingly, gambling commissions, states and Tribes say it’s semantic nonsense and betting on the winner of a football game or a prop is clearly sports betting that should be regulated by states and not the CFTC. To make matters worse prediction markets including those run by traditional sportsbooks like FanDuel or DraftKings often combine bets into same game parlays or other features of sportsbooks but word it in a way where it looks like an event-based derivative instead of a sports bet.

For example, for the September 27th game between the Buffalo Bills and Los Angeles Chargers, a FanDuel bettor could bet on Josh Allen, James Cook and Omarion Hampton to all score a TD at odds of +470. But FanDuel Predict, which only operates where there is no regulated sports betting, can put up a market asking the question “will Josh Allen, James Cook and Omarion Hampton all score TDS on September 27, 2026?” with a Yes option of say 18 and a no option of 82. This works out to odds of +470 for yes and -470 for no since the markets on both sides have to be equal. Considering the vig built into a sports bet the market would probably be offered at odds closer to +550 for yes and -550 for no, but regardless of the odds set it is absolutely clear it’s the same sports bet, just worded in a way that makes it technically legal.

Sports contracts in states without regulated sports betting

Many will likely ask why that’s a problem, but the answer is that these prediction markets on sports are forcing states like Texas, Georgia, South Carolina, etc. that have not legalized sports betting to require their residents be allowed to bet on sports anyways, and  in states like Hawaii, Utah and Alaska that have no legal gambling at all, residents there can wager on sports because they are "investing". In Utah, gambling even goes against their constitution. As mentioned earlier all but four states are challenging the legality of the sports markets but until they are ruled illegal there is nothing stopping people from wagering on sports where they shouldn’t be. Naturally FanDuel, DraftKings and Fanatics moved into markets where they can not offer traditional sports betting with prediction sites to stay competitive with Kalshi and Polymarket. But if they were all on an equal playing field, there is no doubt those sites would be only too happy to close the prediction sites and set up sports betting sites instead.

Prediction markets sites pay no taxes

The other issue is that prediction market sites don’t pay taxes. States rely on revenue and licensing fees for healthcare, infrastructure and education and for some states like New York, Pennsylvania and Illinois, among others, the taxes are extremely high. So when sportsbooks in New York netted $2.5 billion in revenue in 2025 from sports betting, the state of New York received $1.3 billion of that revenue. But if prediction sports markets generated $2.5 billion in revenue from users in Georgia, California or Texas, those states would receive squat, even though their residents were betting on sports. Err, sorry, "investing."

Insider trading

In January of this year the United States invaded Venezuela and arrested Nicolás Maduro, One bettor, Gannon Ken Van Dyke, a U.S. soldier who undoubtedly had insider information on the invasion placed a wager on Polymarket that Maduro would be removed from power winning $410,000. Van Dyke was arrested in April by the U.S. Department of Justice for what they said was "unlawful use of confidential government information for personal gain." The timing of the bet and the amount made it clear that he was using non-public information so making the arrest was easy and Van Dyke will be held accountable for the violation, right? Not so fast. It seems that proving insider trading on prediction markets is far more difficult than trading on the stock market and van Dyke’s attorney is arguing that it can’t be absolutely proven guilty that his client used non-public information when placing the bet, which would be required for criminal charges.

"There's only one person who could have ordered the operation -- that one person is the president. They are never, ever going to get the president to divulge when and how and what, so this is just an exercise in futility," Mark Geragos, van Dyke’s attorney stated. And it is unclear if Todd Blanche would want to get involved since it could open a pandora’s box about the extent of insider trading on prediction markets. So it’s very possible Van Dyke may indeed walk away with both his winnings and no jail time. The trial date is set for December and van Dyke is out on a $250,000 bond.

But that is not the only incident of insider trading and it seems for every case known there are probably many others that have gone undetected. It’s only the high-profile incidents that come to light.

In February of this year George Santos won $17,500 at Kalshi betting that he would attend the State of a Union. Obviously, Santos knew that he would attend, so it was the ultimate form of using insider information for gain, but other than giving up his winnings and agreeing to a ban on Kalshi, Santos did not serve any jail time. Similarly, Gabriel Perez was fined and had to give up profits and agree to a ban from Kalshi for trading on what President Donald Trump would say at press conferences. The problem is that Perez was in charge of the teleprompter so he knew well in advance what the president would say.

There are many other similar instances of this type of use of insider information to profit including charges against a Google employee who allegedly made over $1 million on Polymarket betting on what users would search the most for on Google,  but it appears the lengths the CFTC can use to charge individuals is limited. Unlike the SEC which has power to arrest individuals for insider trading, the CFTC’s hands seem to be tied or they simply don’t care. Martha Stewart went to jail for 5 months after the SEC tagged her for selling stock based on a tip from a friend, Raj Rajaratnam was sentenced to 11 years in prison for securities fraud after the SEC found he used information from insiders at companies like IBM and Intel to make stock trades, and Christopher Collins went to jail for 26 months for trading pharmaceutical stock after he was informed of negative drug trial results long before that was made public. So it appears the CFTC either has no power to jail individuals involved with insider trading or they don’t want to because it would dissuade usage of prediction markets.

Insiders betting on sports

Of course, all of the insider trading mentioned was on current events or what the prediction markets like to call markets of public interest, but what about sporting events?

Sportsbooks have many employees whose job is devoted to looking for unusual betting patterns and spot cheating. And for the most time it has been successful. It happened in the Jontay Porter incident when the league and sportsbooks discovered that Porter was purposely leaving games injured so that bets on him would go under the props totals that sportsbooks put out for him and it happened with Emmanuel Clase and Luis Ortiz when sportsbooks discovered they were purposely throwing balls so that bets on the next pitch being a ball or strike would pay out as a ball each time.  But it doesn’t seem Kalshi or Polymarket would have that same expertise or desire to scrutinize patterns of strange betting, since unlike sportsbooks, prediction markets are all Person-to-Person (P2P) betting so the companies themselves are not losing anything due to cheating, unlike a traditional sportsbook like BetMGM or Bet365 who are the house for all wagers.

It’s clearly a big concern, especially since FanDuel, DraftKings and Fanatics have launched prediction markets in states where they aren’t permitted to offer sports betting. But until clear rules are put in place or a case generates significant attention, as happened with van Dyke, investigations or charges for insider trading on sports are unlikely.

"I guarantee you the same type of time and effort launched against Emmanuel Clase and Luis Ortiz that has resulted in charges against them for personally throwing bad pitches to help other bettors win would never have been launched if those bets were made at a place like Polymarket or even DraftKings Prediction," the analyst said to me. "What should be a big deal is now met with a shrug as cheating is just becoming he norm. And  people won’t stop betting on prediction markets just because it could be fixed. They’ll just view it like betting on wrestling, jai alai or three card Monty. You know it’s fixed but you hope you are on the right side of the sucker’s bet."

Market manipulation

Ask any sportsbook manager and they will tell you one of their biggest frustration and concerns is collusion and market manipulation. This usually isn’t an issue with money lines or point spreads on professional sports, or even college games involving major teams. But in lower-volume markets, such as lesser sports or Division II college games, just a few large bets can move the line and create an opportunity for a sizable scalp when two or more bettors are colluding. This is especially true when the bettors working together use multiple books to manipulate the market. Sportsbooks generally deal with this by getting to know their bettors and understanding when a player is not betting as they normally do. In addition, sportsbooks also put in lower limits on sports like darts, table tennis rugby or an NCAA basketball game between say Gardner Webb and Canisius. Many books also create MOUs with the leagues and gaming commissions to spot strange behavior and betting patterns and they institute integrity monitoring tools to flag large bets or obvious collusion.

A prime example of this occurred for a tennis match in 2007 when Betfair halted trading on a tennis match after it received ten times the normal volume for a small match in Poland when bets kept coming in on 87th ranked Martin Vassallo Arguello against 5th ranked Nikolay Davydenko, even when Davydenko won the first set. Davydenko subsequently lost the 2nd set and retired in the third set with an apparent injury. Betfair voided all wagers and said it was clearly cheating even though both Davydenko and Vassallo were cleared by the ATP.

The big problem with prediction markets compared to sportsbooks is that it is far easier to manipulate markets since there is often no precedent on unique bets and there are no limits. So if someone put up a market on a player like Vassallo and there is $100,000 matched on the event there is no integrity tools at a prediction market site that will flag it as anything out of the ordinary, like was the case with Betfair. And with Polymarket, Robinhood and Crytpo.com, since all bets are made via cryptocurrency there is nothing stopping someone from setting up an account with fake identification, placing wagers and receiving payouts. KYC requirements that are common with sportsbooks like taking a selfie with a piece of identification is not performed  on prediction markets and both deposits and withdrawal to crypto is more or less anonymous. There are also no effective anti-money laundering rules with these sites as there is with traditional sportsbooks overseen by gaming commissions that may spot market manipulation.

Influencers can manipulate markets

But aside from cheating or colluding to affect markets, there is also a risk of people with significant influence saying something on social media to move the lines on markets. This is illegal with stocks and influencers or commentators can be charged with security fraud if they make misleading statements that causes a stock price to move. But there is no similar rule with commodities. In fact, Trump Media & Technology Group launched a paid data feed called Truth API that gives advanced access to Truth Social posts from prominent accounts, including Donald Trump’s, for up to $100,000 a month. It’s clear as day that the only reason anyone would be willing to pay that type of money for access to these posts a day before they are posted is because they feel they can make more by setting up a prediction market and taking action on it before the post is actually made. If this was done for a stock (e.g. Trump giving advanced notice that the DOJ will be charging a major public company with a crime that will make the price go down), then he could be charged with securities fraud. But giving advanced notice that he will be introducing tariffs, invading another country, or doing something that can only be relevant for those trading on prediction markets, is not necessarily illegal and certainly wouldn’t be investigated against him by the CFTC that he oversees.

Fake narratives on who is making money

If you read the narratives of winners on prediction markets both as part of news articles and social media one is led to believe that it is an easy way to make money as long as you do your homework. A June 2026 article WSJ showed that just isn’t true.

The company looked at 1,105 videos created by 10 internet personalities that promoted Polymarket and in 778 of those videos it appeared that the social media influencers were placing trades and winning. But upon further scrutiny the WSJ reporters said that the 778 videos were all just simulated or fake versions of websites rather than the actual betting on the platform and the number of wins far exceeded what was true, since like most gambling ads it fakes large wins.

Reporters said that 118 videos showed the users making $900,000 off their trades on Polymarket, so the WSJ reporters decided to see what would have resulted if those trades were really made on the Polymarket platform instead of via the fake simulations. They identified the real odds at the actual time of the bets and what the outcomes were and found out that 69 of those trades would have lost. Instead of winning $900,000 the bettors would have lost over $166,000. Moreover, the social media videos often showed bets made on markets that never actually existed but would catch a person’s attention. The report raised a lot of concerns and has led to calls by a couple of U.S. Senators to launch an investigation into Polymarket and their deceptive promotional practices.

Who is actually winning?

So how many people actually win on prediction markets? The report along with an independent study of 2.4 million accounts show that only 30% win on the product and that 77% of profits went to the top 1% of players and 67% of profits went to 0.1% of the users. It also showed that the bottom 10% of traders on Kalshi lost an average of $4000 indicating that it’s a true sucker’s bet. Kalshi even acknowledged this saying that for every winner there are 2.9 losers, which obviously contradicts the social media suggestions that playing prediction markets is an easy way to win money.  

In fact. the WSJ report indicated who was making money and it wasn’t the general public. The winners were:

 - Quantitative trading firms with data infrastructure such as Susquehanna International Group
 - Algorithmic retail traders with professional-grade setups such as professional poker player Michael Boss
 - Speed-focused startup trading firms that pay hundreds of thousands each year to acquire live data feeds
 - Market makers who set up the markets and constantly buy and sell only when they are guaranteed a profit, similar to sports bettors who only scalp bets

It is quite laughable that in the early days of online sports betting many sportsbooks were outraged at Billy Walters and the Computer Group for daring to develop to create technology to make money. In comparison, that is peanuts to what is happening today.

Fake betting figures

And when it comes to sports markets, bettors are saying that the volume and wins cited by Kalshi are inflated. A poster on X called retard mode indicated that 61% of sports bets on Kalshi are actually parlays instead of single games or props but they count each part of a parlay as one bet. So for September 19th the poster showed a chart that said that even though Kalshi said they had $1.91 billion bet on sports contracts for that day, they only actually traded $136 million and it was a shell game. $1 placed on a 14-team parlay was counted as $14 in volume and 48% of the parlay slips contain 11 or more legs, so the poster says that Kalshi’s reporting is bogus. It is obvious that Kalshi felt that by inflating the numbers they would attract investors but calling a $1 10-team parlay a $10 bet is not only disingenuous, it is laughable.

These concerns highlight just how many unanswered questions remain as prediction markets continue to blur the line between investing and gambling. Age restrictions, responsible gambling, insider trading,  fake narratives, and market manipulation are serious issues that regulators can no longer afford to overlook.

Coming in Part 2: When Prediction Markets Cross the Line

In Part 2, we’ll look at some of the more controversial issues surrounding prediction markets, including sports betting disguised as trading, profiting from tragedies and other questionable markets, misleading claims about who is actually making money, and the potential for corruption in other forms.