top of page

Wanna Bet? Prediction Markets and the Business of Punting on Everything

Contributors
4 minutes ago
10 min read

Earlier this year, traders on Kalshi, the largest regulated prediction market in the US, learned the value of reading the rules. Kalshi had listed a market titled 'Ali Khamenei out as Supreme Leader?', which attracted approximately USD$54 million in trades. When the question was answered in late February, Kalshi pointed to the fine print: there was a carve-out in the rules for his death that meant the market would not resolve to 'Yes' and those 'traders' (or 'betters', the preferred term for some) would not be paid out.


Traders were, predictably, unimpressed. Many complained on Kalshi's Discord, with one commenting that they were getting 'rugged on a 100% correct prediction'. A class action followed. Kalshi maintained that its rules had been clear all along, then refunded fees and net losses at a cost of about US$2.2 million. It is a rare business that pays millions of dollars to explain its own terms and conditions.

Tens of millions of dollars had been staked on who leads a foreign government. The same apps will happily take your money to speculate on a celebrity's wedding date, whether a particular movie will reach a minimum score on Rotten Tomatoes, or the price gold will reach by the end of the calendar year. Somewhere along the way, 'what happens next?' became an asset class.


Welcome to prediction markets.


How do prediction markets work?


At a high level, the concept is quite simple. The platform asks a yes or no question. You buy a contract on an outcome at a price between $0 and $1, and the price reflects the market's view of the odds. A contract at 20 cents implies a 20% chance. If you are right, it pays $1. If you are wrong, it pays nothing. If you buy 100 of those contracts at 20 cents each, you could collect $100 at a cost of $20.


The platforms are firm that this is not gambling. Users are 'traders' who buy 'shares', and they trade against each other, with no house setting the odds. This is despite Kalshi relying on firms such as Susquehanna International to act as market makers. Several lawsuits have alleged that those firms are a 'house' by another name, but Kalshi disagrees.


The prediction menu is broad - although sport denominates (making up about 90% of Kalshi's trading volume), users can speculate on elections, the weather, or whether a meteor will strike Earth before 2030. There is also a darker side. Kalshi allowed users to wager on whether Palestinians in Gaza would experience mass starvation or whether the Integrated Food Security Phase Classification (IPC) would officially classify Gaza as experiencing a famine. Naturally, this was one of the more controversial of Kalshi's prediction markets. It is, however, not an outlier. Competitor Polymarket took nearly USD$2 million on whether Israel would start a nuclear war with Iran. Hours before the US military captured Venezuela's president, Nicolas Madura, an anonymous Polymarket user bet that he would fall and walked away with approximately USD$410,000.



From niche market to (in)famous


If you think that prediction markets sound a bit strange and perhaps a bit niche, you would have been right a few years ago - although they find their origins in earlier iterations of stock and election markets. The numbers now suggest these platforms are mainstream. Kalshi recently raised USD$1 billion at a USD$40 billion valuation, numbers that are substantial in any industry.


Platforms like Kalshi and Polymarket have entered into partnerships with big industry players. In October this year, the NBA named Kalshi and Polymarket its preferred prediction market exchanges. Yes, preferred exchanges. The deal gives these platforms official league data to settle their markets, the right to use team logos and a place in the ad breaks of the very games people are trading on. In return, the NBA gets a veto over any market it finds objectionable and an information sharing memorandum of understanding with the Commodity Futures Trading Commission (CFTC).


The NBA's reasoning is that fans will bet on its games whether it likes it or not, so it would rather see the data and have a say in what is offered. That is a fair point, and the premier ice hockey and baseball leagues (the NHL and MLB respectively) were already on board before the NBA arrived. However, that point is hard to take at face value when two of the NBA's biggest stars, LeBron James and Giannis Antetokounmpo, have their own endorsement deals with Polymarket and Kalshi.


Media giants have also pulled up a chair at the prediction market table. Kalshi is the official prediction market partner of CNN and CNBC, and CNBC's own report on the NBA deal discloses a commercial relationship with Kalshi that includes a minority investment. Polymarket counts investment from Intercontinental Exchange, the owner of the New York Stock Exchange, among its backers, Google Finance now displays odds from both platforms and the Trump family has invested in the sector. Author Brian Phillips hypothesised what this may mean for prediction markets in his article from The Ringer:


...CNN’s news coverage is in a position to shape public perception. If the network runs 10 stories a day suggesting that the U.S. is about to invade Canada, it will convince many people that this invasion is likely to happen. The prediction markets will shift to reflect this conviction. And CNN, when it puts Kalshi data on-screen as a source of supposed insight, will seem to have evidence that the invasion is in fact imminent, when the “data” is really just a reflection of CNN’s reporting in the first place.

It seems that in the US, the whole fight has come down to a deceptively simple question: is an event contract a swap or a bet? The platforms and the CFTC say these are swaps, which are regulated federally and therefore override state gambling laws. Many states take the view that a bet (or 'speculative position') on a football game is, in fact, a bet on a football game. Courts have split on who is right, and the Supreme Court is widely expected to have the final word.

We wrote in January about Tennessee sending cease and desist letters to Kalshi, Polymarket and Crypto.com. Kalshi sued and later won a preliminary injunction, with the court finding that its sports contracts were likely swaps and beyond the state's reach. Utah went the other way. In August, a federal judge there ruled that federal law does not shield Kalshi from Utah's gambling laws, which is a significant problem in a state that bans gambling altogether.

Utah then pressed the point on 28 October, filing criminal charges against Kalshi and its co-founders, Tarek Mansour and Luana Lopes Lara. The Deseret News describes a felony count of pattern of unlawful activity, 12 felony counts of online gambling promotion and a misdemeanour, while other reports describe 14 counts, 13 of them felonies. The case was built old-school: a state investigator opened an undercover account in September and placed 12 sports bets, including a USD$5 wager on the spread in a Utah college football game. Across all 12 bets he risked $114.87 and finished $25.10 down, which must make it one of the more modest sting operations on record.

Utah's Attorney General summed up the state's position:

"Calling it a 'prediction market' doesn't change what it is."

Kalshi called the charges "a crazy overreach" and pointed to Arizona, which charged Kalshi first in March only for a federal judge to block the prosecution. The federal government has also picked a side. A day before Utah filed its charges, the CFTC sued Utah to stop it enforcing its gambling laws against federally regulated exchanges. Meanwhile, a bipartisan group in Congress has introduced a bill called the Prediction Markets Are Gambling Act, a title that leaves very little to the imagination.


ASIC: we have seen this film before


Australia's regulators have been watching all of this with polite scepticism. On 30 October, ASIC Chair Sarah Court said that ASIC "is not currently convinced that prediction markets for retail clients are in the public interest", adding that the contracts "have all the features of a bet". ASIC also flagged the integrity risks that tend to follow money wherever it goes, including the misuse of confidential information, market manipulation and disputes over how outcomes are decided. Its warnings were aimed in two directions: (1) Australians who trade offshore were told they have no protection under Australian law, and (2) Australian financial services licensees that facilitate access to offshore prediction markets were reminded that all their licence obligations still apply. Some licensees already offer that access, so the reminder was not just hypothetical.

In its statement, ASIC said its approach to prediction markets is consistent with its approach to binary options, products it has already removed from the retail market.

A binary option is an over-the-counter derivative that pays out depending on whether a specified event happens within a set timeframe, such as a share price rising, a market index moving or a central bank changing interest rates. The client either receives a fixed payout or loses the entire stake, with nothing in between. ASIC's concerns centred on that all or nothing payoff, along with very short contract durations and negative expected returns. ASIC banned the sale of these instruments to retail clients from 3 May 2021, and the ban was later extended to 1 October 2031. The numbers behind that decision were stark. Earlier ASIC reviews found that about 80% of retail clients lost money, and in the 13 months before the ban took effect, 74 to 77% of active retail clients lost money, with aggregate net losses of $14 million.


The actual order is narrower than 'anything with a yes or no payout'. It covers certain over-the-counter derivatives tied to the value or amount of an asset, rate, index or commodity, issued by someone carrying on a business of issuing those derivatives. An election result is not automatically caught just because the payout is $1 or nothing. The order also excludes derivatives able to be traded on a licensed market, meaning one authorised under an Australian market licence. A US licence is not enough.

Having users open the contracts does not, by itself, get you out of the net. Section 761E of the Corporations Act can treat a broker or market operator as the legal issuer even where users trade against each other. The order also prohibits arranging the issue or acquisition of covered binary options. The contract terms and legal structure matter more than who clicks 'create market'.

The broader Australian legal landscape


Australian law does not deal with prediction markets by name. If someone wanted to operate something akin to a prediction market in Australia, that operator would need to bring its product within one of two existing frameworks: (1) financial services regulation or (2) gambling regulation.

If prediction market contracts are derivatives, they are financial products under the Corporations Act 2001 (Cth). Operating a platform on which they are traded may require an Australian market licence, and under section 798A of the Corporations Act ASIC must have regard to whether granting that licence is in the public interest. ASIC has now indicated, before receiving any application, that it is not currently convinced retail prediction markets meet that test, and no prediction market operator has applied for such a licence. The existing binary options order covers binary options and does not automatically extend to prediction markets. ASIC does, however, retain its product intervention power under Part 7.9A of the Corporations Act, which allows it to act where a product has resulted in, or is likely to result in, significant detriment to retail clients. That is the same power it used to ban binary options. Licensees that facilitate access to offshore prediction markets also remain subject to all of their licence obligations, a point ASIC made expressly in its statement.

The alternative is to treat prediction market contracts as wagering. That brings in state and territory licensing and the Interactive Gambling Act 2001 (Cth), which prohibits unlicensed interactive gambling services from being provided to customers in Australia. The Australian Communications and Media Authority (ACMA) enforces that prohibition and blocked Polymarket in August 2025 as an illegal gambling service, whilst Kalshi lists Australia as a restricted jurisdiction. Neither the Act nor the state regimes were designed with exchange-based event contracts in mind, and it remains unclear whether a licensed Australian wagering operator could offer such a product on events such as elections or the weather. There has yet to be an Australian contender willing to test this point.

Whilst some may view these regimes as obstacles, they are also the source of the protections Australians give up when they trade offshore: licensing and conduct obligations on the financial services side, and self-exclusion through BetStop and other harm minimisation measures on the gambling side. ASIC's warning that Australians using offshore prediction markets do not have protections under Australian law is a reminder of what sits on the other side of a VPN.


Australia's new gambling rules


Even the grey zone is getting smaller. The Interactive Gambling Amendment (Gambling Reform) Act 2026 became law in August, and its principal reforms start on 1 January 2027. Most of the attention has gone to wagering ads, but the more interesting provision for this market is new section 15J. This section requires banks and payment system participants to prevent, so far as reasonably practicable, payments to designated interactive gambling services, which in practice means illegal operators. The maximum civil penalty is $364,000, and ACMA will issue guidance on what "reasonably practicable" actually means. "Funds" includes digital currency, although the duty does not extend to winnings flowing back to Australian customers, and internet service providers, DNS providers and app stores pick up similar obligations of their own.


That does not automatically capture every virtual asset service provider (VASP). The question is whether the provider falls within those banking or payment system definitions. The payment system definitions are broad, and ACMA's new guidance expressly identifies cryptocurrency service providers among the entities that may be caught, depending on their role. Including digital currency in 'funds' is not, by itself, enough to impose the duty on every crypto business.

Put together, the practical effect is that an offshore prediction market treated as an illegal gambling service may find some Australian funding routes blocked. That could include a crypto on-ramp where the provider and transaction fall within section 15J, but the provision does not establish a universal block on crypto transfers. Banks will first need to work out which merchants are illegal gambling operators, which is easier legislated than done. And for good measure, licensed wagering providers will generally no longer be able to use athletes, celebrities or influencers in their ads, so an Australian answer to LeBron's endorsement deal is off the table.

Where this leaves us

Prediction markets can be genuinely useful at pulling together scattered information and putting a price on it. They can also turn almost any question, from a movie's Oscar success to an election result, into something you can trade from your phone, with all the incentives for insiders and spin that come with it. In the US, the question of what these products actually are will be answered by the courts. In Australia, the regulators have seemingly offered their answer before anyone has even asked - at least until some operator brave enough comes by to test the waters.

For brokers, exchanges, payment providers and crypto platforms, the more pressing question is whether an existing product already touches this space, through offshore access, payment flows or on-ramps. ASIC and ACMA have made it clear they are watching, and it would be wise to have an answer ready before they come asking.


Written by Steven Pettigrove and Luke Higgins

© Michael Bacina and Steven Pettigrove. All rights reserved

  • White LinkedIn Icon
bottom of page