Are Prediction Markets an Unserious Venture?

We're doing something different this month. Keeyan and I are taking opposite sides on the introduction of prediction markets in Canada, and we want to hear your thoughts and takes in the comments.

Last month, ABC News reported that the man who had been running Donald Trump's teleprompter since 2016 spent the winter betting on what the president was about to say. Gabriel Perez won around $100,000 over more than a dozen speeches, using Kalshi's Mentions markets, which pay out when a particular word gets said on stage. Kalshi's surveillance team caught him, froze most of the money and handed the file over to the CFTC.

In April, prosecutors brought the first U.S criminal insider trading case these markets have produced, against an Army sergeant accused of using classified details of the Maduro operation to turn $33,000 of Polymarket positions into more than $409,000. A Google engineer was charged in May with making over $1 million betting on which search terms would trend, using data from inside the company. And Maria Corina Machado's Nobel odds went from under 4 percent to over 70 in the hours before last October's announcement.

Canada made its own move in March, when the CIRO set out a narrow path for event contracts sold through registered dealers, and in June Wealthsimple Predict became the first of these products most Canadians could actually buy. Interestingly, Wealthsimple holds both a banking license and this exemption for prediction markets, simultaneously (most financial institutions restrict these kinds of businesses as clients). However, regulations have remained tight, allowing predictions only on economic data, financial markets and climate, with sports and elections left out. And despite their presence at a recent Blue Jays game (MLB), Polymarket is still banned in Ontario after the OSC pushed it out last year.

So, Canada has opened up something the rest of the world is still fighting about, in the same year it produced its first real insider trading cases. Keeyan thinks that was a mistake. I think we've got a shot at writing rules nobody else has managed yet.

Prediction Markets: What George Carlin Would Call “Gambling”

Much of my worldview is informed by libertarian principles, so I recognize the irony in taking a restrictive stance on people’s rights to freely transact. However, I see two fundamental issues with prediction markets, which transcend borders. Canada’s approach is just the perfect example of how regulation sometimes creates more issues, instead of solving them.

The main problems are the trustworthiness and centralization of oracles (who “resolves” a market?), and malicious engineering of outcomes. Recently, a market on Polymarket which allowed users to predict whether Strategy, Inc (formerly MicroStrategy, the company run by Michael Saylor) would sell Bitcoin between May 27-31 resolved to “No”, despite Strategy’s reported sale of 32 BTC on an SEC filing. Since Polymarket uses UMA, a decentralized truth-verification protocol, one would assume that markets always resolve correctly – especially given that Polymarket itself has no stake in the outcome (they make their money on fees). UMA tokenholders vote to resolve markets in dispute, with each vote carrying a weight proportionate to the amount of their holdings. Nothing is stopping prediction position holders from simultaneously holding UMA tokens. The result of this kind of unjust market resolution was a lawsuit, in this case, and could lead to many other similar ones in the future.

The obvious but misguided way to fix the issue with unreliable or corrupt oracles would be to adopt a centralized resolutions model, as Kalshi (the U.S.’s largest prediction market operator) has done, but that comes with obvious drawbacks too. Notably, the involvement of a market operator in the decision-making process around who makes, and who loses money, could lead to clear conflicts of interest. This is the same issue that insurance companies have been accused of perpetuating, when they adjudicate claims in-house, instead of using independent adjusters, and exactly the kinds of thing that decentralized  finance has the potential to fix.

While insider trading is nothing new, the damage it could cause was mostly limited to financial crimes and undue losses caused to others. Using insider information to buy a stock before earnings would not, for example, crash a plane, or derail a train. But prediction markets are now expanding to cover flight delays, bankruptcies or seizures of train companies, and even the ousting of foreign leaders. This can create perverse incentives for financially motivated individuals to practically influence the outcomes of certain predictions (for example, by booking a ticket, flying on and disrupting a flight). Of course, the ill-motivated actions would likely be illegal, and so would the fraudulent gain if the contract resolved in such a person’s favor because of their meddling. And certain kinds of prediction markets are already disallowed by federal law, with more restrictions proposed (Congress has been debating this topic since 2003). But as long as prediction markets operate and offer betting opportunities on public infrastructure and foreign policy outcomes, people will find ways to turn the odds in their favor, moral or not.

Dan made great points in his take, and I agree that Canada (and the United States, for that matter) should take initiative and write a regulatory playbook for prediction markets imminently. As Dan points out, Canada could benefit from taking a leadership role in this, because it is often losing “regulatory capture” when its citizens resort to using offshore or US platforms to transact in ways which their own country doesn’t sanction. I do think, however, that Dan is missing one important point: the regulation of prediction markets should have as much or even more to do with demanding market operators to be transparent and honest (never trust the casino!) as it does with regulating consumers’ access to them.

Daniel Says: “Write the Rulebook Here”

Let’s start with what all four of those insider-trading stories have in common, which is that we know about them at all. The exchange's own surveillance desk flagged Perez and froze most of the money. The sergeant's positions were reconstructed from the blockchain, and the Nobel leak showed up on a price chart while it was still happening. Each situation was identified and controlled within weeks, meanwhile the insider trading of equities usually surfaces years later, if at all.

Keeyan sees a product that attracts insiders, and he's right that it does. What I want to know is how fast the venue catches them, and doing so within three months is quick in the standard of business.

There's also the question of what a ban gets us. Single game sports betting was a criminal offence here for decades, which did nothing to stop Canadians wagering billions offshore with books that answered to nobody, until Parliament gave up and legalized it in 2021. Crypto went the same way and cost people more, because when there was nowhere legitimate to trade, the volume went to whoever would take it, and some of that was on FTX. Polymarket is banned in Ontario right now and somebody was handing out flyers for it at a Blue Jays game in the spring. A ban mostly decides whose rules people trade under, and it usually isn't ours.

Keeyan’s claims are legitimate: Thin markets can be moved by one patient wallet, and a lot of contracts have no business even existing. The offshore platforms have listed their fair share of those, including bets on whether particular people would live out the year. It’s no question that this is gambling, all of which a regulator should get to decide.

By treating these as derivatives sold through registered dealers rather than licensing them as a casino product, Canada has made the right move in upping the scrutiny and obligations these markets receive.  The obvious missing piece is that people who already know the answer shouldn’t be allowed to bet on it. Perez knew the words before the president said them. Clear insider-trading, but the Americans are still arguing about whether he broke any law at all, which will take years. Canada should be writing our laws down right now.

There's a second reason this should matter to our dear readers. The best evidence in the Maduro case came off a blockchain, because that's what these venues settle on, and a public ledger tied to a verified account is about the strongest audit trail in finance. A regulated market here would also give institutions something they can't get cleanly today. A fund wanting to hedge a rate decision or a tariff ruling has no simple venue for it, and this would be one.

Canada has traditionally favored a “follow the herd” mentality, but it’s also no stranger to being the first mover. The world's first spot Bitcoin ETF was approved here in February 2021 while the Americans spent three more years litigating, and when they came around, they borrowed the structure. Prediction markets will exist at scale whether we take part or not, and there will be more Perez’s either way. I'd rather the rules everyone copies were ours.

So, who has it right? Tell us in the comments which of us you're with or make the case we both missed.