The fight for fair markets
Prediction markets are at a crucial moment. Volume is real, capital is flowing in, regulators have opened the door, and some of the biggest names in finance are paying attention.
The central question is not simply whether one exchange will win. It is whether traders will retain any meaningful power as the market grows.
Even a market led by two major venues can become deeply concentrated. If each exchange owns its users, interface, liquidity, data, and fee structure, traders remain trapped inside separate ecosystems. The exchanges may technically compete, but the trader has little leverage unless moving between them is genuinely easy.
That is the future worth scrutinizing: not just one exchange becoming the market, but a handful of venues dividing up the users and controlling the terms on which they trade. History suggests that this rarely ends well for the people providing the order flow.
For nearly two centuries, there was one club
If you wanted to buy a share of an American company, you effectively went through one club. The New York Stock Exchange operated under fixed minimum commissions dating back to the Buttonwood Agreement of 1792. Every broker charged the same rate because the exchange required it. There was no competition on price, because there was nowhere else meaningful to go. Commissions stayed high for generations, not because the service was expensive to provide, but because the structure made it impossible for anyone to undercut anyone else.
Then came May 1, 1975, the day Wall Street called May Day, when the SEC abolished fixed commissions and forced brokers to compete. Commissions collapsed. Charles Schwab was founded that same year and built an empire on the simple premise that trading should cost less. Within a generation, the cost of buying a stock fell from a meaningful percentage of the trade to effectively zero. Nothing about the technology changed overnight. What changed was that the market for trading itself became competitive.
Something similar happened on Nasdaq in the 1990s. Dealers had settled into a convention of quoting stocks only in even eighths, which kept spreads artificially wide and quietly transferred billions from investors to market makers year after year. Two academics, William Christie and Paul Schultz, noticed the pattern in 1994 and asked a simple question: why would competing dealers all avoid odd eighth quotes unless they had stopped competing? The scandal that followed produced new order handling rules and opened the door to electronic communication networks. Spreads compressed almost immediately. Investors had been overpaying for years, and it took structural competition to fix it.
Look at where American equities ended up. Today a share of Apple trades across more than a dozen exchanges and dozens of alternative venues, all connected, all forced by regulation and by each other to honor the best available price. Spreads are the tightest in history, and retail investors get executions that institutional desks could not have matched thirty years ago. Fragmentation gets criticized for its complexity, some of it fairly, but the benefits to end users have been substantial by almost any measure. Competition among venues made trading cheaper and fairer.
That is the great equalizer. Not any single exchange, however well built. The existence of alternatives.
The same dynamic appeared in the great technology companies
Apple built a great product, then built a store on top of it, then made that store the only way to reach a billion pockets. The commission was set at thirty percent and stayed there for over a decade, not because thirty percent reflected the cost of running a store, but because there was no second store. It took lawsuits, regulators on three continents, and years of pressure from developers to move that number even slightly. Microsoft did a version of this with the browser in the 1990s. Google did a version of it with search distribution and app store billing. Ticketmaster does it with live events, where fees routinely add a third to the price of a ticket.
None of these companies are villains in some cartoonish sense. They are rational actors responding to rational incentives, which is precisely the point. When a platform becomes the only venue, it does not need to be evil to start extracting. It simply needs to answer to its own shareholders. Fees drift up, terms tighten, and the platform starts competing with the businesses built on top of it. The users have nowhere to go, so their interests slowly fall down the priority list.
These economics have a name in every industry where they appear. Switching costs, vendor lock in, monopoly rents. The vocabulary changes, but the underlying mechanism is usually the same: pricing power flows to whoever controls distribution, and sooner or later it gets used. Not immediately, and not while the platform is still fighting for adoption and needs to be generous. Later, once dependence is established. The generous early years are not a counterargument to the pattern. They are usually the first stage of it.
We already know this. It is why crypto exists.
We should be especially wary of this structure, because the entire industry is a fifteen year argument against single points of control. Bitcoin exists because trusting one institution with the ledger seemed like a bad idea. Decentralized exchanges exist because trusting one company with custody and matching seemed like a bad idea. The most expensive lesson the industry ever learned, the collapse of FTX, was a lesson about what happens when too much activity, trust, and liquidity concentrate in a single venue run by a single set of insiders.
So it is strange to watch parts of the prediction market world drift toward the exact structure this industry was built to escape. A winner take all outcome, where one venue owns the liquidity, the data, the interface, and the fee schedule, would be a regression. A two horse race is not much better. Duopolies tend to settle into comfortable parallel pricing, the way the credit card networks have for decades.
To be clear, this is not about any exchange's intentions. Kalshi has done real, difficult, admirable work legitimizing event contracts with regulators, and we have written our own Kalshi review precisely because it deserves a serious look. Polymarket proved global demand at a scale nobody thought possible. These are impressive companies and the space is better because they exist. The concern is structural, not personal. Even a well run exchange will eventually face pressure to monetize its control over liquidity, distribution, and user behavior.
Exchange owned interfaces are not neutral products. They are distribution channels designed to keep order flow inside the exchange that owns them. That does not make them bad products. It makes them misaligned. Traders should understand the incentive behind the screen they are looking at.
We are already seeing how fluid the economics can be. Polymarket restructured its entire fee model this spring, moving from a fee on profits to per trade fees that vary by category, and has adjusted the rates more than once since. Every venue tunes its pricing, and that is its right. But a trader on a single venue experiences those changes as weather. Something that happens to you, that you cannot negotiate with, and that you cannot escape. A trader connected to several venues experiences them as information, and can simply send the next order somewhere else.
The case for many marketplaces
Recent market developments point in the opposite direction, which is encouraging. Rothera, the exchange backed by Robinhood and Susquehanna, went live and cleared three billion dollars in volume during the World Cup almost immediately. Novig won CFTC designation in June and will run a peer to peer model that puts direct pricing pressure on everyone. Polymarket entered the US market. Kalshi keeps expanding its product surface, including its new Pro terminal. Onchain venues like Predict.fun are growing their own liquidity. New exchanges are filing with regulators every quarter.
This is what a healthy market looks like at this stage. More venues means fee pressure in every direction. It means an exchange that degrades its product or raises its costs actually loses flow, which is one of the strongest forms of discipline a market can apply. It means better prices, because the same event trading in several places creates a real reference price instead of one company's opinion. It means resilience, because no single outage, policy change, or failure can take the whole asset class down with it. And it means innovation, because exchanges competing for order flow have to earn it.
Onchain exchanges deserve a particular word here, because they encode this argument into the system itself. On a venue like Predict.fun or Hyperliquid, the order book is public, settlement is verifiable, and the rules are written in code that anyone can read before committing a dollar. Custody stays with the trader. Fee changes happen in the open, where the market can see them and react. No support ticket decides whether you can withdraw. They are not a replacement for regulated exchanges, and the two will likely grow side by side for a long time. But as long as a credible onchain alternative exists, every centralized venue has to price and behave as if its traders can leave, because they can.
How we fix this system
While trading today, you may not notice that the same market trades at meaningfully different prices across venues, sometimes for hours. The issue is not a lack of information. The exchange that owns your screen owns your defaults, your data, and your habits. You do not notice the fee that ticks up or the term of service that shifts, because you have nothing to compare it against and no muscle memory for leaving.
So keep your freedom to move, and use it. Compare prices across venues before you trade. Route your order to whoever offers the best fill, not whoever you happened to sign up with first. Reward the exchange that treats you well and punish the one that does not. A trader who cannot leave is not a customer. They are revenue.
This is why we built Kairos. Not as another venue, but as the layer that makes that freedom practical. One terminal across Kalshi, Polymarket, Predict.fun, Hyperliquid, and every serious venue that comes next, so you can express conviction wherever the price is best without being owned by the place you express it. I will not pretend a cross venue terminal solves every problem fragmentation creates. Traders still face different collateral systems, settlement rules, market definitions, and regulatory constraints, and those are hard problems that we are here to work through. But it does solve one important problem well. It makes comparison and exit materially easier, and comparison and exit are what keep venues honest.
The venues will keep competing. They should. Our job is to make sure that no matter how the venue race unfolds, the trader stays free. Free to compare, free to route, free to leave. That freedom is the great equalizer, and it is worth protecting while this market is still young enough to choose its shape.
Kairos.