Table of Contents
At the heart of the Kalshi/Polymarket/etc “is this CFTC regulated or gambling?” debate is the line between hedging and gambling. As someone that worked on a trading floor for a long time, the question of where that line sits is not new.
When smart people spend all day managing risk and laying off risk and generally assessing the odds in myriad ways, it should come as no surprise a lot of betting on random unknowns in the world takes place.
So what we are going to do here is propose a different framework in which to consider this question. Why an entirely new framework? Because we think any approach that requires a definition for gambling, or a litmus test for whether something is gambling, or principles under which things should be regulated as gambling, is destined to fail. And we think it is going to fail because gambling is as hard to define as pornography. And nobody seems to have a watertight definition for either. So building up all the rigamarole of tests and criteria and whatnot needed to assemble a workable regime from a "what is gambling?" starting point is just much harder than needed to address the real issues at play here.
Strange Gambling
Movies and books about finance often include anecdotes about betting on how much someone can eat or drink. Or other frankly childish bets which amount to little more than hazing and cheap entertainment for stressed-out, bored and possibly-a-bit-on-the-spectrum people that sit near each other all day. Fair enough.
But when your job is actual risk management and hedging and you see bets flying back and forth on sports, food, drink, weather, anything really: you do get some perspective on what is gambling and what is managing “real” risk.
And this looks, pun intended, a lot like pornography. Obviously betting slips do not look like pornography; they look like betting slips. It is, however, famously difficult to define pornography. US Supreme Court justice Potter Stewart famously acknowledged this challenge while asserting "I know it when I see it." And for most people that feels about right. That case was about a French film in the 1950s from a first-rate filmmaker, with a first-rate cast, that won an award at the Venice Film Festival. Stewart took the position the film in question was legitimate art. If this incident tells us anything it tells us that context matters and sometimes defining terms is hard.
The same can be said about gambling: most people have a sense of when they think a transaction is a legitimate hedge or financial activity and when they think it is gambling. If you are in an unmarked, unlicensed, illegal basement bar's back room there is a pretty good chance you are gambling. And the financial services equivalent of a Venice Film Festival award is approximately an article in the FT or WSJ that describes what you are doing as "investing." For this prediction markets stuff notice the mainstream financial press – whatever you think of them remember that context matters – is openly grappling with this question. Empirically, it is not easy to tell.
If we cannot articulate specific rules we are left with something like a broad social consensus. Deciding to screen, and then award, a film at one of the world's most prestigious film festivals demonstrates a large measure of social consensus. At the same time financial TV openly asking if prediction market platforms are just gambling shows the opposite. Not making an accusation just asking a question. But if you are questioning something then you do not have a consensus on what it is!
A cynical person might point out that financial media have a long-clear bias for promoting profitable things even if they look a bit sketchy. Everyone has heard snide comments about Forbes 30 Under 30 prison pipeline and such. If they are questioning legitimacy then, yes, Potter Stewart's remarks are relevant.
Look Around
Spend a bit more time with a huge stream of wagers flowing among your colleagues on a trading floor and it all becomes fuzzy. So let’s start simple. Slot machines and roulette are gambling if anything is. The odds are fixed and known in advance. The play is quick. The rules are simple. And there is zero skill involved. Think a bit more, though, and you realize being good at cheating could count as a skill. That probably sounds like a joke but we will come back to it shortly. It is not a joke.
At the other end of the spectrum we have activities like trading physical commodity futures or forwards or options. Those feel like “real” risk management and not gambling. Most everyone accepts that buying stocks and bonds is investing not gambling.
But push a bit more on these feelings and it all sort of falls apart. For the vast majority of people oil or wheat futures do not offset any risk in their real lives. Even if you argue that oil prices are linked to some kind of expense there is no case to make that the spread between oil in Dubai and Singapore offsets any risk for someone that lives in Montreal. And many retail stock traders are prone to ill-researched risky bets that might at least plausibly be characterized as gambling.
Cheating
What about cheating then? Seriously. Thinking about what cheating is in a gambling context helps to understand how gambling differs from “legitimate” financial risk taking.
First there is the question of odds. In roulette, you are able to work out the odds of every outcome in advance without knowing anything beyond the rules of the game. Or at least you are supposed to be able to. If a casino consciously tilts the wheel and does not tell anyone we call that illegal cheating. But what if you work out that a supposedly fair wheel is not in fact fair and bet using this information? That is not cheating. It is research. Tilting the wheel is illegal. Noticing it is tilted is not. Though even this is a bit grey in that you are supposed to call out the crooked casino and not effectively collude with the casino to take advantage of the other players.
Cheating by a player is influencing the odds somehow. This can be influencing the wheel with some absurd magnet or laser scheme from a movie. Or it can be violating the rules by, say, somehow changing your bet once the wheel starts spinning.
Now imagine you work at an oil company and you know about a huge oil find in one region. Or you know a port will close before much of the market because it is next to your office and you can see a fire. Trading oil in these circumstances is not cheating because there is no rule against it. Trading companies know more than end users all the time. Their business is to know the products they trade and use that information to make money. Everyone expects that, for example, giant oil companies know more about oil prices than retail buyers at a petrol station.
The key difference here is the game either allows, or does not allow, you to go off and find more information. In gambling things that would otherwise be considered “smart” or legitimate research are cheating because the rules explicitly call them cheating by defining the world of the game to be narrow and classifying any deviations as cheats. And there we get something like a candidate rule: is research, beyond reading the rules and making inferences only from those rules, allowed and able to legally yield an edge? Studying basic games of chance yields nothing once you compute all the odds. The research road is short. Commodity markets have endless opportunities for learning more.
Stress Testing the Framework
This sounds reasonable. Now apply it to horse racing. Knowing the weather and maybe knowing which horses do better in wet or dry or hot to cold or whatever could make some kind of difference. The rules allow checking the forecast. How could anybody ban that? And events are influenced by weather because they take place outdoors. This is known to everyone. So is betting on horse races not gambling?
Maybe this feels wrong and you want to add a carve-out because there is no “legitimate” underlying economic interest in a horse winning a race. But a win will impact the value of the horse and the horse may have many owners who trade their ownership stakes and on and on. Now maybe you think that is a bit, well, small.
Consider a Formula 1 team. The budgets are huge. The prize money is huge. And teams often have borrowed money or owe suppliers for parts or somehow have liabilities. Winning gets you more prize money and more sponsorship. For a struggling team with debts trading below par an unexpected win could cause the debt to reprice.
Now imagine you are owed money by a marginal team. A complex bet that they will do poorly over the next few races might legitimately hedge your financial risk. Maybe you think betting on wins is fine here. But what about betting on the cumulative time behind the winner the team’s best car finishes? Or how many goals a football player scores in a season? That might matter if you’ve loaned that player a lot of money to buy, well, whatever football players spend money on.
This argument posits a lot of sports wagers are as legitimate on the “I’m managing risk” front as commodity futures. Betting if your colleague can eat 3 kilos of guacamole in 15 minutes probably cannot qualify here but…maybe someone can have a go at an argument? Certainly betting how many cigarettes someone smokes in a year could be linked to their healthcare costs or something. Aggregate cigarette consumption across a company will impact that company’s healthcare costs and thence the stock price. These things are not roulette! Studying has some marginal value.
And?
So we think approaching the entire “what does the CFTC regulate?” question from the gambling perspective is not useful. Mainly because a lot of participants in financial markets are gambling and yet we think it is fine because those markets also have a large investing purpose. Writing rules to classify products is hard and perhaps impossible.
Most of what is colloquially called gambling has historically been regulated in the US by the states and the US Constitution does not give the federal government power to regulate gambling. Trying to parse out what is and is not gambling is incredibly difficult as we just saw. And the CFTC was not intended to be a federal gambling regulator. How do we know that? Well the name for one. And second because the federal government does not have the power to regulate gambling within the states.
Now to put a fine point on the issue: we think the entire approach being bandied about today is not going to lead anywhere useful. Are prediction markets gambling? This is the wrong question. The rationale behind the existence of the CFTC is the federal government’s power to regulate interstate commerce and the fact that “commodity” markets are pretty well linked across states. Gold prices, oil prices, wheat prices and these sorts of “commodity” prices are linked even for physical products that sit within different states because they are physical goods one can ship from place to place easily.
If you see prices getting out of line you can make deals and ship product. People do this and there is a large social good in having liquid markets for common commodities so businesses can lay off risk and scale.
What about wagers on basketball? You cannot ship a basketball game. But you can enter into offsetting wagers in different states. And to do so you will need to have whatever licenses those states require. A reasonable number of states have chosen to issue zero or very few such licenses. This points the way on resolving the question of what the CFTC regulates and what states regulate.
No state bans oil trading. In fact there are essentially no licenses or regulatory requirements to carry out interstate “commodity” trading beyond the bare framework of contract law and things in the general vicinity of fraud and money laundering laws. What is different is not the nature of the wagers – sports or oil – but whether the states have reached a consensus the underlying activity is not (heavily) regulated. If states have light-to-no rules then federal preemption in the name of interstate commerce and a level playing fields makes sense. If states have wildly different rules then it does not.
Is a sports wager a “commodity swap?” Stop looking at the swap part and ask if the states have reached a consensus on whether the underlying is a lightly regulated commodity or a highly regulated other non-commoditized thing. Of course what precisely counts as a lightly regulated commodity is non-trivial. This approach does not remove all the complexity from what is, at heart, a complex set of questions. But this makes a lot more sense than trying to define something that has proven remarkably difficult to nail down for so long.
