Original author: Oliver B

Original translation: TechFlow

The prediction market gold rush has begun. Every crypto founder, fintech entrepreneur, and contrarian is convinced they've found the secret to success. They're convinced they have the prediction market platform that can beat Polymarket and Kalshi. They've raised funding, built teams, and launched flashy interfaces, promising a better user experience, faster settlement times, or a niche market that the incumbent giants have overlooked.

However, most people are destined to fail.

This isn't pessimism; it's mathematics. In prediction markets, network effects are incredibly powerful. You need liquidity to attract traders, but you also need traders to build liquidity. In crypto-native markets, Polymarket already has an advantage in scale. In the US IPO market, Kalshi enjoys the regulatory high ground. Unseating these two players is prohibitively expensive. From marketing to regulatory response to user acquisition, these costs accumulate quickly. Even if new entrants achieve some success, they only further fragment an already thin market. This is a death sentence for platforms that rely on order book depth.

The graveyard of failed prediction market platforms speaks volumes. Remember those half-dozen markets that launched after the 2024 election cycle? Yeah, almost no one remembers them.

However, the real focus for venture capitalists is that the real profits in prediction markets lie not in operating these markets themselves, but in the infrastructure that supports them.

Why infrastructure is a better investment option

Looking back at the history of financial markets, not all wealth in the stock market is generated by stock exchanges, although some is certainly true. The real wealth comes from data providers, clearing houses, trading infrastructure providers, market surveillance systems, and deeper analytical platforms. For example, Bloomberg didn't earn billions by competing with the New York Stock Exchange (NYSE), but rather by becoming an indispensable piece of infrastructure.

Prediction markets are following the same trajectory, albeit a few decades behind. Currently, the infrastructure layer is nascent, fragmented, and inefficient, and this is where the real opportunity lies.

Here are a few specific areas where venture capitalists should focus:

Data and Oracle Infrastructure

At the heart of prediction markets lies "real data." They require authoritative data sources to provide key information, such as which candidate won, what the actual GDP figures are, or whether a company achieved its targets. This may seem simple, but it's actually complex. Different markets require different data sources, as well as diverse verification and settlement mechanisms to prevent data manipulation.

Oracle networks designed specifically for prediction markets are crucial. These companies are responsible for aggregating data, providing cryptographic proofs, and resolving disputes. As the market expands, a fragmented oracle ecosystem will become unsustainable. The ultimate winner will be the infrastructure provider that all platforms—even competitors—will have to rely on.

Cross-market infrastructure and aggregation

Currently, liquidity is fragmented across different platforms. A savvy trader might want to arbitrage between Polymarket, Kalshi, and three other platforms, but there's currently no seamless way to do this. Building an infrastructure that allows traders to view order books across all markets would be extremely valuable. This system would allow traders to hedge simultaneously and manage risk across multiple venues, unlocking enormous potential value. This is the "Bloomberg Terminal" opportunity in prediction markets: everyone benefits, and more efficient cross-market operations mean tighter spreads and deeper liquidity.

Analytics and Historical Data

As prediction markets mature, researchers, quantitative analysts, and institutions will want to delve deeper into historical forecast data. They will look for patterns and understand how the market prices events over time. Someone will establish an authoritative repository of prediction market data that is cleansed, standardized, and searchable. This will become a reference dataset for academic research, institutional analysis, and model building, creating a highly profitable and defensible business.

Processing and Settlement

As prediction markets expand and become more complex, their backend systems also need to be upgraded. More efficient settlement mechanisms, faster data processing capabilities, and improved market infrastructure are all crucial. Companies specializing in building middleware will be of immense value. They connect markets to clearing systems, automate settlement processes, and reduce operational risk. Think of it as the plumbing that makes modern markets function.

Compliance and risk management infrastructure

As prediction markets move toward mainstream adoption and gain greater regulatory clarity, complexity will also arise. Infrastructure for managing regulatory reporting will become critical. Large-scale KYC/AML (know your customer/anti-money laundering) capabilities will also become essential. Detecting market manipulation and ensuring compliance across jurisdictions will also be crucial. This type of infrastructure may seem "boring," but it's a highly defensible and sticky area. Once embedded in the market system, it's nearly impossible to easily replace.

Infrastructure layout for traders

Another key aspect of prediction markets is the infrastructure support provided to professional traders.

Currently, prediction market users are primarily retail investors and enthusiasts. However, as the market matures and attracts institutional capital, quantitative traders, and algorithmic traders, demand will shift dramatically. These professional traders will not only need access to the market but also a comprehensive suite of tools that institutional finance takes for granted.

Algorithmic Trading and Trading Robot Infrastructure

Professional traders will want to automate their strategies across multiple markets. This will require APIs, execution infrastructure, and trading bot frameworks specifically designed for prediction markets. In the future, someone might create a "Zapier" or "Make.com" for prediction markets, allowing professional users to easily create complex trading strategies. Such tools would allow them to hedge and manage risk without writing code. Furthermore, companies might even develop infrastructure specifically for professional quantitative traders, enabling them to efficiently implement these functions.

Portfolio and risk management tools

As traders accumulate positions across multiple prediction markets and platforms, they will require more advanced tools to support their operations. They will need to track, manage, and understand their risk exposure. For example, what is the net exposure to political events? How are these positions correlated? What is the optimal hedging strategy? These questions may not concern retail traders, but they will become core requirements for institutions managing millions of dollars in prediction market capital. The first platform to offer institutional-grade portfolio analysis tools will have the opportunity to capture market share from significant amounts of serious capital.

Backtesting and Research Framework

Before committing capital, institutional traders want to backtest their strategies using historical prediction market data. However, this data is currently not organized in a format that facilitates backtesting, nor are there tools to support this need. Therefore, companies are needed to build robust backtesting frameworks that provide clear historical data and realistic simulations of market microstructure. Furthermore, these tools need to be easily integrated into existing research tools. This type of infrastructure will become a key pillar for the quantitative trading community to enter the prediction market.

Market Microstructure and Intelligence Tools

Professional traders know that the market is not only about correctly predicting outcomes, but also about a deep understanding of liquidity.

They need to identify market inefficiencies, monitor information flows, and precisely time entries and exits. As prediction markets mature, the demand for real-time market intelligence tools will grow rapidly. Microstructure analysis tools, such as heat maps showing where "smart money" is flowing, real-time alerts on unusual activity, and tools to detect mispricings, will become particularly important. These capabilities will be similar to the role of the Bloomberg Terminal in traditional financial markets, but tailored specifically for prediction markets.

Real-time aggregation and one-click trading: the essential cornerstones of institutional funds

For professional traders, trading simultaneously on multiple platforms is a fundamental requirement. In the future, platforms will inevitably emerge that can aggregate order books from Polymarket, Kalshi, and other prediction markets in real time. Through such platforms, traders can view liquidity across all markets in a single interface and trade across platforms with a single click. This is not only a dream for market makers, but also critical infrastructure for the efficiency of the entire prediction market ecosystem.

This trader-facing infrastructure is just as important as the market-side infrastructure. These tools aren't just nice-to-haves; they're essential prerequisites for institutional participation. As institutional capital pours into prediction markets, these tools will become essential and core. Companies building this layer of infrastructure will capture a different kind of value than market operators. This value is not only highly defensible but, to some extent, even more scalable.

The ultimate question of valuation: How much room for growth is there in the prediction market?

The recent funding rounds of two major prediction market players have garnered significant attention. Kalshi recently achieved a $5 billion valuation, while Polymarket, with investment from Intercontinental Exchange (ICE), the parent company of the New York Stock Exchange, reached a $9 billion post-money valuation.

This is no small increase. Just a few months ago, Kalshi's valuation was $2 billion, while Polymarket's valuation was only $1.2 billion at the beginning of 2025. In just a few months, these valuations have soared by 2.5 times to 7 times, respectively.

This raises an unsettling question for venture capitalists: How much further can the prediction market grow?

Currently, both companies have reached sufficiently high valuations that future exit multiples are limited. Assuming Kalshi or Polymarket ever reaches a valuation of $50 billion to $100 billion, this would undoubtedly be a decent, but not spectacular, return from the current base of $5 billion to $9 billion.

More importantly, these platforms are increasingly becoming potential acquisition targets for traditional financial giants. Exchanges, brokers, and financial institutions are showing strong interest. A sale to Intercontinental Exchange (ICE), Chicago Mercantile Exchange (CME), or another large brokerage firm at 2-4 times their current valuation is entirely possible. However, this isn't the "power law" investment with a 100x return that venture capitalists seek.

In contrast, investments in infrastructure offer a completely different return curve. Whether it’s an oracle service provider, an analytics platform, or a cross-market execution layer, once they become the core infrastructure of the prediction market ecosystem, their returns will spread across all platforms, all traders, and all institutions.

This type of infrastructure typically starts at a low valuation, but its expansion potential is virtually unlimited.

Asymmetry of risk

In the fiercely competitive platform space, venture capital firms often bet on multiple projects, hoping one will become the next Polymarket. This is a classic "power law" bet: most projects will fail, and even those that succeed may struggle to create significant value due to market fragmentation and liquidity fragmentation.

In contrast, infrastructure investments have a completely different risk profile. For example, an oracle provider doesn't care whether traders use platform A or platform B—it benefits regardless of which platform wins. The value of an analytics platform, on the other hand, increases, not decreases, with the number of markets. Infrastructure doesn't need to choose winners; it simply needs to be usable by all platforms.

Furthermore, infrastructure often forms strong defensive capabilities through data advantages, network effects, or technological barriers. This is not just a race to burn money, but also a contest of technological depth and ecological stickiness.

What does this mean for investors and entrepreneurs?

If you’re evaluating a business plan focused on building a new prediction market platform, whether it’s selling it on the promise of a better user experience (UX) or targeting an untapped market segment, you need to ask more pointed questions:

• How to solve liquidity problems?
• How to achieve profitability in the face of competitive pressure from existing giants?
• Among the many competing platforms, how many will succeed?
• More importantly, even if successful, what is the realistic likelihood of an exit multiple from a base of over $100 million in funding?

If you focus on infrastructure opportunities, you face a completely different risk-reward model. Build data layers, develop cross-market tools, design settlement mechanisms, create trader analytics, and establish intelligent intelligence platforms. These businesses grow with the expansion of the entire market, not against a single competitor. They benefit from market prosperity, not from fragmentation. They offer the kind of unconstrained growth potential that venture capitalists truly seek.

The prediction market ecosystem is still in its early stages, which means there is a huge opportunity. However, the real opportunity lies not in replicating what Polymarket has done, but in building the foundational layer that makes the entire ecosystem run more efficiently .

Platforms will fight each other, and infrastructure will continue to expand.