Prediction markets are sub-optimal betting vehicles
Prediction markets are being touted as the future of sports betting. I disagree.
The literature in this piece assumes the reader has a fundamental understanding of the structure of prediction markets (and their associated strengths/weaknesses).
If you don’t, check the explanation article and then this piece to see if they work.
The loophole
First and foremost—how are these products legal?
Sports betting is heavily regulated and not available in every U.S. state.
Kalshi bypasses this by not being a ‘sportsbook’, but rather an independent provider offering a resolution system for two counterparties to bet against each other.
They are classified by the CFTC as a contracts market, meaning they can offer events with binary resolutions (e.g. did X team win?). Additionally, because Kalshi is a third-party provider that does not have stake in the event outcome, they are recognized as a peer-to-peer platform and not a bookmaker.
While definitely not the same scope, it does resemble Silk Road. The platform operator (DPR, Ross Ulbricht) was never charged for selling drugs, just providing a marketplace for them.
The performance
The claim is that because there is no VIG, the odds are more beneficial for both parties involved. This is true, however slightly misleading.
While there is no house edge, these platforms do need to charge a fee to turn a profit. Polymarket used to have their fees set at 2% before turning them off indefinitely. For a thought exercise, let’s compare a bet placed on Polymarket at 2% fees versus a sportsbook.
At ‘even’ odds:
Polymarket will be set at 50c for both YES and NO. Assuming zero liquidity constraints, post the 2% fee, bettors have roughly 49c in EV, or 98% of their bet.
Books will set you at −110 on both sides. This means bettors have roughly 47.6c in EV, or ~95.2% of their bet.
The prediction market is offering 2.8% in additional EV post the fee. Without the fee, it is 4.8% additional EV.
The argument this paper makes is that even with zero fees the downside of structured counterparty liquidity outweighs the additional EV offered.
Scalability & Optionality
Prediction market liquidity and optionality are inherently limited in scope compared to a traditional bookmaker.
Since each event requires two participants to agree on a price (probability), matching becomes increasingly difficult the more obscure the event is.
Finding two parties to agree on the probability of the Super Bowl winner is far easier than an in-season match.
This is mirrored in a liquidity curve, where popular markets have the most liquidity, and it continually declines the further down the popularity curve you go.
In these situations, the further down the curve, the more EV a trader loses.
This is because sophisticated actors are only willing to take the other side of a trade where they can take value. The less popular the event, the more value they demand, ultimately reflected in worse execution for their counterparties.
Going back to our earlier comparison of EVs between prediction markets and sportsbooks, the assumption was that you will always have a counterparty willing to match you at the correct probability. This assumption is not correct.
With the casinos, the standardized VIG is ~5% for every event no matter the obscurity (within reason). Sportsbooks don’t want to necessarily offer the more obscure events because they are more prone to sharps exploiting their lines. However, they do anyways in order to ensure maximum optionality for their users. This is because the house is a singular counterparty, and with enough volume the math always works out for them.
With prediction markets, it is not reasonable to assume that the counterparties can offer 5% VIG (or less) for more obscure events because each counterparty is an individual book. There is little reason for an individual book to take on the risk of providing liquidity for these smaller events unless it is extremely favorable.
The secondary effect of this is parlays are practically impossible on prediction markets—the available liquidity is simply nonexistent at a fair probability.
This section brings me to my next point about available liquidity—it is mostly looking to perform arbitrage.
Predatory Liquidity → Arbitrage
The majority of liquidity available on prediction markets is predatory. Because large orders induce slippage, there are ample opportunities for informed users to take a value bet or hedge on a sportsbook.
See below for multiple examples:
Non-informed retail taker flow:
Set orders 5c above the fair probability
Retail flow fills order.
Hedge on a sportsbook (or take value).
Profit.
Static orderflow:
Fair probability shifts after a score change.
Snipe static orders that aren’t automatically adjusted.
Hedge on a sportsbook (or take value).
Profit.
On prediction markets, there is no house so the market is reliant on the available liquidity at each pricing tick. If a large order comes through, the price changes. That is how orderbooks work and has nothing to do with the size of the spread (sure, more liquidity lessens the potential of this happening, but doesn’t negate the chance).
With sportsbooks, the trades are against a single counterparty who sets a limit for each price. If the sportsbook is willing to sell me $100k worth of bets at 55% implied probability, I can fill that $100k at 55%. If I do this, the house may dynamically adjust the odds higher, but it doesn’t affect my fill.
However, because prediction markets are reliant on external LPs, a large order does move the price. If I want to fill $100k of a bet that has $75k of liquidity at the pricing tick offered, the remaining $25k will be filled at a higher price.
Contrary to popular belief, the majority of liquidity above the current pricing tick is not uninformed flow. Instead, it is there to fill retail price movement and hedge on a different market.
There is no downside to market-making markets like this and spoofing orders because you can hedge instantly without price impact (because of the structure in which houses offer exposure).
Conclusion
Prediction markets offer increased EV on select markets, at the expense of limited optionality and liquidity. The reality is also that the majority of available liquidity is predatory. While house liquidity is also predatory, it’s capped at ~5% VIG.
It’s up to the reader to determine which market is best suited for their needs, but most serious gamblers prefer the stable source of liquidity and wide variety of markets.
At this point, it is my belief that sports betting usage for prediction markets will continue seeing traction, but never come close to ‘Ending Vegas’ as so many proclaim.
I mean, sportsbooks can shut down your account if you’re too consistently successful so that seems like a major downside for “serious gamblers”. This is also a way that they’re pretty blatantly more predatory than PMs.
I’m not sure I follow the argument as to why we should expect less liquidity on prediction markets. Assuming 0 fees and similar volumes, why wouldn’t bookmakers (also) offer similar liquidity at the same ~5% VIG on prediction markets? They can even use it to help balance their books. I would personally offer prediction market liquidity at 5% VIG but Polymarket generally has better rates.
Regarding obscure events, I understand the argument to be that they are using profits from their popular events to subsidize these likely unprofitable obscure events. Why couldn’t a prediction market do the same? Polymarket already has an attempt at doing this, liquidity rewards. Of course that requires some sort of fee to fund long term.
Regarding slippage, why would a sportsbook be able to fill $100k but not external LPs? If a sportsbook is willing to fill it, I’m assuming it’s profitable. Is it that you think the sportsbook will accept more risk than a collection of external LPs? The only other argument I see is insufficient volume to balance the trade which would be an issue in either case. Prediction markets can also just limit the bet size (Like sportsbooks do) to whatever is available at the current price to eliminate slippage.
Predatory Liquidity:
1. Feels wrong to me that you frame orders a few cents above fair as ‘predatory’. That’s what VIG is in sportsbooks, except on every order. It’s just generally not the current price in prediction markets because the current price is generally much closer to fair.
2. I agree with your second example, sniping after a score change. But this is also not fundamental. Polymarket could easily clear the book + suspend for a few seconds after score changes if they have the data feeds (Which they seem to).