Behind the Bet: How Prediction Market Liquidity Really Works

Prediction markets are often described as peer to peer betting platforms. But when a retail customer places a sports parlay or “combo” the other side of that wager is rarely another ordinary retail bettor or trader. Behind the scenes is a sophisticated and competitive ecosystem of market makers made up of quantitative firms, sports syndicates, specialist betting operations, and major financial institutions. Firms like White Swan Data, one of the more public liquidity providers pricing sports RFQ flow on exchanges such as Kalshi, illustrate the kind of specialist operation this ecosystem is built on, alongside prop-trading shops, financial institutions and other syndicates moving into the same space.

Prediction markets have experienced rapid growth by offering consumers a product that is similar in appearance and feel to traditional sports betting. But instead of placing bets against the bookmakers, users trade contracts. Instead of traditional odds, they see probabilities. Instead of the platform taking the other side of every bet, the exchange connects customers with liquidity providers or market makers. The distinction is important but it also creates a common misconception by the retail market about what “peer to peer” actually means. 

How Prediction Market Liquidity Actually Works

The difference between a prediction market and a traditional sportsbook is who takes responsibility for the customer's position. If a customer places a $100 bet on a conventional sportsbook, the sportsbook is the counterparty. The book sets the price, accepts the bet and manages the resulting liability of the wager. A prediction market can operate differently. Rather than the platform taking the other side of every transaction, it can connect customer demand with external liquidity.

That liquidity has to come from somewhere. Market makers exist to provide it. They look at the available opportunities, work out their own prices and decide whether they are willing to take positions from customers.  Depending on the structure of the market this can happen through an order book or through a request-for-quote (RFQ) system, where a customer requests a price for a combination of outcomes and liquidity providers compete to provide it.

From the users perspective very little of this is visible. They can choose various outcomes, see the price and click 'trade'. However behind the trade there is an automated pricing system linked to market data, risk evaluation systems, and a balance sheet able to cover large liabilities. That is the reason why the term "peer-to-peer" can be misleading; although the trade might occur on a peer-to-peer platform the other party is not usually another recreational bettor, but rather a professional market maker, a quant firm, an institutional firm, etc. 

Pricing Complexity Is the Edge

A market maker doesn't have to predict each individual result of a trade in order to earn a profit. Instead what matters is whether the price offered in the market adequately takes into account the probability, the uncertainty, and the risk associated with the position. The more complex the product is, the harder this becomes. 

A single sports contract can be quite simple, the market maker can compare the probabilities that are available with those from its own models and decide whether they are good enough to offer liquidity. But when there's a five leg combo it is a different problem entirely as the market maker has to assess the probability of each individual outcome but it also needs to understand how the outcomes interact. Two selections may be positively correlated and others may move in opposite directions. The timing of the underlying events can change the exposure while multiple positions can create overlapping liabilities. It is not sufficient to multiply together the probability of each selection in order to arrive at an effective price. This is one of the fields in which market makers can gain an edge. The more difficult it is to price a product accurately, the more valuable proprietary models, data, technology, and experience becomes. 

That is relevant to the growth of sports combos on prediction markets. Traditional sportsbooks have spent years developing correlation models and infrastructure for multi-leg products. Prediction markets are now creating an environment where external liquidity providers can compete to price those same forms of risk.

RFQs Turn Liquidity into Competition

Rather than simply accepting the prices that are shown on the order book, the user can request a quote for the specific combinations of outcomes. That request is then presented to the market makers which each determine their own price and decide whether it wants to take the risk. 

The user receives the best available quote. This creates a different competitive environment from a traditional sportsbook. The books control the price it offers to its users. In an RFQ environment liquidity providers are competing against one another for the same bet or order. If one market maker prices too conservatively another can offer a more attractive price. If a firm believes it has a better understanding of a particular market or risk it can price more aggressively to compete for the order.

Capital Is the Moat

Once a market maker has decided whether or not it has found an edge, capital determines how far that edge can be scaled. To be competitive market making requires a combination for capital, technology, and fundamental pricing ability. None of the three is sufficient on its own.

Capital enables a company to hold larger positions and take on more liabilities. Technology allows it to process event information, automate pricing and manage exposure. Fundamental pricing determines whether the risk being offered is actually attractive. 

A company that has a large balance sheet but poor sports pricing  may struggle to generate sustainable returns. On the other side a specialist sports operation with excellent models may be able to identify attractive opportunities with a comparatively small bankroll. The problem here is scale.

A market maker can have an edge without having enough capital to exploit it at meaningful volume. As trading activity grows, the balance sheet becomes increasingly important because positions need to be supported by collateral. This creates an interesting distinction between actual risk and capital requirements.

It is possible for some of the positions to have outcomes that do not overlap one another, so that the firm's theoretical maximum loss could be much smaller than the total of the liabilities linked to those positions. The exchange might still demand a large amount of collateral in order to back the activity. This implies that a firm can have a risk profile which is fairly manageable even if it needs a considerable amount of capital to operate on a large scale.

It is a natural obstacle to new entrants. It is possible to develop technology and to buy data and pricing models can also be set up. But it is much more difficult to obtain the kind of capital needed to run a large market-making operation. This is one of the reasons why the final competitive landscape might not be decided simply by who has the best model. It might instead be determined by which company has the combination of pricing, infrastructure, and balance sheet necessary to implement that model on a large scale.

Technology Is Part of the Risk Management

Capital alone does not solve the problem. Prediction-market liquidity providers are operating within infrastructure that is still developing, meaning they have to build sophisticated internal systems while adapting to exchanges that are simultaneously changing their own products.

Market makers need to connect to different APIs, map markets correctly, ingest feeds, price contracts and monitor exposure across multiple platforms. The technical challenge increases as prediction markets launch new products and introduce increasingly complex combinations. The cost of getting that infrastructure wrong can be substantial.

Market making technology is not simply a means of making decisions more quickly but it forms part of the risk-management framework. Even if the pricing engine is mathematically advanced, it becomes meaningless if the underlying market is not correctly mapped.

APIs, settlement logic, market mapping, exposure monitoring and automated execution all become potential sources of financial risk. As prediction markets expand, the firms capable of building reliable infrastructure across multiple venues could therefore have a meaningful advantage over firms that can price risk but cannot operationalize that pricing effectively.

The Liquidity Layer Is the Real Market

The fact that prediction markets are growing is not simply a matter of more people trading event contracts, it also involves the appearance of a new liquidity layer located between retail demand and the actual risk. Since sports parlays and ever more complicated products are being offered on the exchanges, the companies supplying that liquidity are becoming an increasingly important element in the way the market operates.

For the customer the experience can still be simple. They just have to choose a combination, be given a price and then decide whether or not to go ahead. Behind this interaction market makers are dealing with a considerably more complicated problem since they are constantly assessing probabilities, correlations, market conditions, capital requirements and exposure while competing with other firms for the same volume.

That trend is likely to become even more significant as the market matures. With more participants involved there will be greater competition and more rigorous pricing, but it will also raise the standard needed in order to compete. Simply being connected to an exchange won't suffice. Companies will have to have the technology to operate reliably, the models necessary for pricing ever more complex products and a balance sheet to back the positions involved when operating on this scale.

Although prediction markets are marketed to consumers as a simpler alternative to conventional sportsbooks, the financial market operating behind the user interface is becoming ever more sophisticated. The activity of retail traders is merely the obvious part of the transaction. The actual competition takes place beneath the surface, among the companies that are able to price, fund and manage the risk which makes the whole market possible.