Supercycle of the prediction market: When everything has Liquidity

Author: Tulip King, Messari Analyst

Compiled by: White55, Mars Finance

The prediction market has not yet been fully understood by the public, but this is your opportunity. We are currently witnessing a complete transformation in the way markets operate, how information is priced, and how futures are formed.

“iPhone Moment”

Every technological revolution goes through a wonderful period during which the old thinking patterns of the public hinder the perception of disruption. In 2007, when Nokia's executives saw the iPhone, they said, “It doesn't even have a keyboard.” They compared the iPhone to mobile phones but overlooked its comparison to computers. The iPhone did not compete with existing mobile phones; it disrupted the concept of single-function devices.

The prediction market is currently the same. When people see Polymarket, they think it is just a strange betting site with poor liquidity. They compare it to sports betting sites like DraftKings or derivatives trading markets like CME, feeling that it falls short. This is similar to the mistake in the Nokia case. Polymarket is not a better betting site; it is replacing the concept of the entire professional financial market. Let's imagine, when stripping away the complexity of each financial instrument, what they actually are:

Options: A bet on whether a future price level will be reached.

Insurance: A bet on whether a disaster will occur.

Credit Default Swap: A Bet on Whether Bankruptcy Will Occur

Sports Betting: A type of wager on the outcome of whether a sports event will win.

A massive industry worth trillions of dollars has been built around essentially binary opposition issues, with each industry having its own infrastructure, regulatory systems, and monopolistic intermediaries profiting from it.

Polymarket boils everything down to a fundamental element: creating markets for any observable event, allowing people to trade, and resolving them when reality determines the outcome. It is no better than DraftKings in sports betting, nor is it more exceptional than the Chicago Mercantile Exchange (CME) in derivatives. What it does is more fundamental: it reduces all markets to basic units and then rebuilds on that foundation. Polymarket is the “iPhone”; everything else is just an “app.”

Multidimensional Trading

When all transactions take place in the same place, it unlocks more possibilities. Imagine if, five years ago, you wanted to build a position expressing the following view: “I believe the Federal Reserve will raise interest rates, but tech stocks will still rise because Trump will express some positive comments about AI.” You would have to open accounts at different institutions, deal with different regulatory frameworks, and different forms of leverage. However, the part of the trade that mentions Trump actually had no market at all.

On Polymarket, you only need to click three times. More importantly, these are not three separate bets, but a coherent worldview expressed through interconnected positions. You can buy “Fed pauses interest rate hikes” as “No”, buy “NASDAQ hits all-time high” as “Yes”, and simultaneously buy “Trump mentions artificial intelligence in next speech” as “Yes”. This connection is the essence of trading itself.

Thinking about the way to predict the market

For those who don't quite understand, here is a real case. Last month on Polymarket, you could build a position like this: buy “Hyperliquid will not airdrop before December 31” for “no” at 67 cents, while buying “Hyperliquid will drop to 20 dollars before 2026” for “yes” at 13 cents. Think about the various outcomes:

Most Profitable: Hyperliquid has neither airdropped nor dropped to $20 this year. Given the current market conditions, this perspective seems reasonable, and its pricing should be higher than the current 8% market rate. You take advantage of this outcome to buy at a low price.

Make a small profit: Hyperliquid is airdropping, the price drops to $20, or the price drops to $20 but there is no airdrop. These scenarios are the most likely outcomes, with a probability of 63%.

No return on investment: Hyperliquid conducts airdrops and prices remain above $20. Given that the market is very concerned about new supply from team unlocks, it seems that introducing more supply through airdrops will receive a very negative reaction from the market. The probability of this outcome occurring should be less than 29%, so you are selling this outcome at a high price.

This sounds like hedging, but it's more than that. It expresses a complex view on how the market handles the supply of new tokens, a perspective you can't convey anywhere else.

Traditional markets force you to compress a complex worldview into rough directional bets. You might have a complicated argument that Nvidia's earnings will exceed expectations, but you would still sell the stock due to market expectations (good news already priced in). In the options market, you can only choose between call options and put options, or perhaps construct some less appropriate spread combinations to barely express your opinion, but you have to pay a high premium for that. On the Polymarket platform, you can accurately express your views by simply purchasing the two “yes” options of “Nvidia earnings exceed expectations” and “Nvidia stock price falls 5% after earnings report.”

The real brilliance lies in your ability to contemplate those intermarket correlations that shouldn't exist but indeed do. Florida is about to be hit by a hurricane, while the Tampa Bay Buccaneers are set to play in Detroit. Traditional thinking suggests that these two events are unrelated. However, you have this hypothesis: if the hurricane really wreaks havoc in Florida, the NFL referees will undoubtedly favor the Buccaneers to create an inspiring narrative. Thus, you bet 60 cents that the Buccaneers won't win, while betting 20 cents that a hurricane will hit Florida. You are not betting on a specific outcome; you are betting on a structure of correlation. You profit from your understanding of how narratives affect referee decisions.

This is why prediction markets do not compete with existing markets: they operate on completely different levels of abstraction. All other markets only give you a single lever to pull. In contrast, Polymarket provides you with countless levers, and more importantly, it allows you to pull specific combinations of levers that align with your view of how the world actually works.

Why were savvy investors wrong again?

The first criticism you always hear about prediction markets is related to liquidity. “Large-scale trading is impossible”; “The spreads are too wide”; “These are just gamblers wagering their lunch money.” This is not a bug, but your opportunity.

Individuals do not dislike Kel, but the liquidity issue will ultimately be resolved by the existing market incentive mechanisms.

It might be worthwhile to think about why liquidity will explode from a mechanical perspective. Traditional market making is actually relatively simple: you are usually making a market for things that have a clear correlation with other items. Stock options are related to stock prices, futures are related to spot prices. Everything has hedging methods, correlations, and reliable models to rely on. This is why a handful of companies like Citadel and Jane Street are able to make markets for thousands of instruments.

Prediction markets are more difficult. Each type of market requires a specialized intelligent system:

The sports market needs models that can be updated with every score, every injury report, and every weather update.

The political market needs natural language processing to analyze public opinion polls, speeches, and social media sentiment.

The event market needs a machine learning system that can calculate the benchmark rate from historical data.

The mentioned market (the market being referred to) requires a language model trained on thousands of text records.

You cannot allow a market maker to monopolize all markets, as each market requires completely different expertise.

In the long run, this is actually beneficial for liquidity. It will no longer be a few giants controlling all market-making activities, but rather a large number of specialized market makers will emerge. Some quantitative analysts will become the world's leading experts in pricing for the Mention market. Other teams will dominate weather-related events, while another group will focus on celebrity behavior. The fragmentation that seems like a disadvantage will actually create resilience and depth.

In the next five years, a brand new type of financial company will emerge: prediction market experts. They will not make markets in stocks or bonds, but rather in the realm of reality itself. The company that first succeeds on a large scale in doing this will become a fortress in the new financial system.

Unsettling facts about the truth

Prediction markets are not really about predictions; they are about creating an economic incentive mechanism for the truth. We are currently in a strange moment where everyone has an opinion on everything, but no one has a real stake. Your favorite analyst has been wrong twice in a row on recession predictions. The number of accounts ruined by those commentators on CNBC is countless. Yet, they manage to keep their platforms, retain their audiences, and continue to be wrong without any consequences.

The trust in the media has declined, but that's all.

This situation is unsustainable, and everyone knows it deep down. Today's society has constructed an information ecosystem that rewards engagement rather than accuracy. Being loud is more important than having the correct viewpoint, and social media has exacerbated this issue. Nowadays, the most popular opinions can become the “correct” opinions regardless of their accuracy. Those with the most followers have become the experts, and so-called experts are simply the ones who can garner the most likes.

The prediction market completely overturns this model. Suddenly, being right has value, and being wrong has a cost. The market doesn't care whether you graduated from Harvard, whether you have a blue V certification, or whether you've written a book about the market. It only cares whether you are correct. When you create a system that only rewards accuracy, wonderful things happen: accurate people suddenly have a reason to speak up, while wrong people finally fall silent.

But this is not just about transferring wealth from the wrong hands to the right hands. Prediction markets are building a parallel information system that operates in a way that is completely different from the existing media ecosystem.

In the old world: information spreads through social networks, depending on its virality.

In prediction markets: information is priced based on real-world conditions.

These are entirely different selection mechanisms that will produce entirely different results.

Polymarket is still in its infancy, but its accuracy is already astonishing. The political market is more predictive than any opinion poll aggregator. The Federal Reserve market has already seen fluctuations before any economists updated their predictions. This is not because the traders in the prediction market are smarter, but because the incentive mechanism is to ensure predictions are correct, rather than to pursue entertainment.

Reward market, not prediction market

The public now refers to “prediction markets,” just as it refers to Bitcoin as “digital gold”—technically not wrong, but completely overlooking a more important point. True prediction markets are passive observers. They price probabilities but do not intervene in outcomes. However, Polymarket is not passive, and that is the crux of the matter: any market with human participants inherently carries incentives to alter outcomes.

Here is a detailed explanation. When there is a market prediction on whether someone will throw green items during a WNBA game, someone made the following calculation:

Buy $10,000 worth of YES stock at a price of 15 cents per share.

Throwing objects by oneself

When the market is determined to be “YES”, a fee of $66,000 will be charged.

After deducting legal fees and the cost of a lifetime ban, the net profit is about $50,000.

In theory, the income and expenditure of this WNBA disruption incident should be: Compensation = Criminal charges + Social humiliation + Lifetime ban + Efforts required.

Compensation that is too high will attract imitators, while compensation that is too low will go unnoticed. The market will find the accurate equilibrium price.

Charlie Munger: “Show me the incentive and I will show you the outcome.”

But the “prediction market” has turned into a bounty market - it is not predicting whether someone will throw an item, but rather offering a specific amount as a reward for someone to facilitate this event. This is neither a loophole nor manipulation, but rather the most important feature of the prediction market, yet no one mentions it.

Imagine a thought experiment like this. I decide to run for mayor of New York City. The odds offered by the market are quite reasonable, only 0.5%. With such odds, I can buy 20 shares, each worth $5,000, totaling $100,000. If I win, each share can bring in $1 million. When I put these shares in escrow as compensation for the campaign team, things get interesting. I hired 20 hitmen, promising them that if successful, each would receive $1 million.

The aforementioned experiment created something that should not exist: a politically funded campaign backed by the market, where the stakes increase as the odds decrease. The market is essentially saying, “This outcome is too unlikely; we give you 20 to 1 odds to make it happen.” Prediction has turned into a bounty. The market is not just observing reality; it is also funding specific futures.

Some events are completely unaffected by the reward mechanism. The value of becoming president is so high that no prediction market reward can significantly enhance its incentive effect. But for the thousands of other events: from corporate decisions to cultural phenomena to sports event outcomes, the reward mechanism is real and active.

Currently, we have not achieved the kind of future prediction market guided by wisdom as envisioned by Robin Hanson (the core designer of modern prediction market mechanisms). Instead, we are witnessing some stranger phenomena: the market pays those who facilitate a specific future.

Conclusion

My personal prediction is that within ten years, the Polymarket model will consume most areas of traditional finance. This is not only because it is a lower-cost and more efficient trading platform, but also because it integrates all markets into a single raw platform and then rebuilds on that basis, which is more efficient than maintaining thousands of specialized market structures.

The dominoes will fall in the following order:

DraftKings - Sports betting is essentially a prediction market with worse odds.

Chicago Options Exchange - Options are simply complex binary bets on price levels.

Insurance - merely a prediction market for unilateral transactions.

Credit Market - Bankruptcy Prediction Plus Additional Steps

Every vertical field will resist and regulate, but in the end, they will yield because they will realize they are not competing with better products, but with superior principles.

What we are witnessing today is a complete reorganization of how the market operates. You are no longer betting on prices, interest rates, or volatility, but rather on events, the relationships between events, and most importantly, on the probabilities that can change those events.

Astute investors should be prepared for this now. It's not just about buying tokens or trading on the market, but considering what happens when there is a liquidity market for any observable event. What happens when every company's decision has a prediction market? What happens when every piece of legislation has odds? What happens when every cultural trend has a price?

Prediction markets are not just about building better markets, but about creating a machine that can incentivize people to create the future.