Why the prediction market is still in the exploratory stage

Original Title: Why Prediction Markets Are Still in Beta
Original Author: Nick Ruzicka
Original Translation: SpecialistXBT, BlockBeats

Prediction markets are having their moment in the spotlight. Polymarket's coverage of the presidential election has made headlines, and Kalshi's regulatory victories are opening new avenues. Suddenly, everyone wants to talk about this "world truth machine." But behind this wave of excitement lies a more interesting question: if prediction markets are so good at predicting the future, why haven't they become mainstream?

The answer is not sexy. The problem lies in the infrastructure—specifically, regulation in the U.S. (for example, Kalshi received approval from the Commodity Futures Trading Commission (CFTC), and Polymarket has established an offshore presence), but infrastructure issues remain widespread. Even in regions where prediction markets are legal, the same fundamental challenges persist.

Platforms dominating in 2024 are throwing money at these problems. According to Delphi Digital researcher Neel Daftary, Polymarket has invested about $10 million in market maker incentives, paying over $50,000 a day at one point to maintain liquidity in its order book. Today, these incentives have collapsed to just $0.025 per $100 traded. Kalshi has spent over $9 million on similar projects. None of these are sustainable solutions—they're merely band-aids on structural wounds.

Interestingly, the challenges hindering the development of prediction markets are not mysterious. They are clearly defined, interconnected, and— for the right entrepreneurs—easily solvable. After engaging with teams in the field and analyzing the current state, we found five recurring issues. We can consider them a framework, a shared vocabulary to help us understand why prediction markets, despite their theoretically bright prospects, remain in beta.

These are not just problems; they are also a roadmap.

Issue 1: The Liquidity Paradox

The fundamental issue is liquidity. Or more precisely, the chicken-and-egg dilemma that causes most prediction markets to become ghost towns.

The mechanics are simple. When new markets launch, liquidity is low. Traders face poor execution—high slippage and price impact make trading unprofitable. They all exit. Low trading volume scares off professional liquidity providers, as they need stable fees to offset risks. Without liquidity providers, liquidity remains scarce. This cycle continues.

Data confirms this. On the Polymarket and Kalshi platforms, most markets have trading volumes below $10,000. Even larger markets lack sufficient depth to attract meaningful participation from institutional investors. Any large position leads to double-digit price fluctuations.

The root cause is structural. In typical cryptocurrency liquidity pools (e.g., ETH/USDC), you deposit two assets and earn fees as traders transact—even if the price moves against you, both sides retain value. Prediction markets are different: you hold contracts that become worthless if they fail. There’s no rebalancing mechanism, no residual value—the outcome is binary: half of the assets go to zero.

Worse still, you get "harvested." As markets approach settlement and outcomes become clearer, informed traders know more than you do. They buy the winning side from you at favorable prices while you are still pricing based on outdated probabilities. This "toxic order flow" continues to bleed market makers dry.

Polymarket switched from an automated market maker (AMM) model to a central limit order book in 2024 for this reason: the order book allows market makers to cancel quotes immediately upon realizing they are about to be trapped. But this does not solve the fundamental problem—it merely provides market makers with some defensive tools to mitigate losses.

These platforms circumvent the issue by directly paying market makers. But subsidies cannot scale. For flagship markets—presidential elections, major sporting events, hot cryptocurrencies—this model works. Polymarket's election markets are liquid. Kalshi's NFL market competes with traditional sports bookmakers. The real challenge lies in all the other areas: numerous prediction markets that could work, but lack the trading volume to support millions of dollars in subsidies.

The current economic model is unsustainable. Market makers do not profit from spreads but are compensated by the platform. Even protected liquidity providers, who face limited losses (4-5% maximum per market), require ecosystem subsidies to break even. The question is: how to make providing liquidity profitable without burning cash?

Kalshi's successful model is gradually emerging. In April 2024, they brought in Wall Street's major market maker Susquehanna International Group, making it the first institutional supplier. The result: liquidity increased 30-fold, with contract depth reaching 100,000 and spreads below 3 cents. But this requires resources that retail market makers cannot provide: dedicated trading platforms, customized infrastructure, and institutional-level capital investment. The key to breakthrough is not higher rebates but getting the first institutional investor that truly values prediction markets to see them as a legitimate asset class. Once one institution participates, others will follow: lower risk, benchmark pricing, and trading volume will naturally grow.

But there’s a catch: institutional market makers need to meet specific conditions. For Kalshi, this means obtaining approval from the CFTC and clear regulatory guidelines. For crypto-native and decentralized platforms—many of which lack regulatory moats or large-scale developers—this path is not viable. These platforms face different challenges: how to initiate liquidity without providing regulatory legitimacy or trading volume guarantees? For platforms other than Kalshi and Polymarket, the infrastructure issue remains unresolved.

What Entrepreneurs Are Trying

Quality-weighted order rebate incentives improve trading—such as shortening trade times, increasing quote sizes, and narrowing spreads. While this approach is pragmatic, it does not address the fundamental issue: these rebates still require funding. Protocol tokens offer an alternative—subsidizing liquidity providers (LPs) by issuing tokens instead of tapping into venture capital, similar to the launch models of Uniswap and Compound. Whether prediction market tokens can accumulate enough value to sustain issuance in the long term remains unclear.

Tiered cross-market incentives provide diversified liquidity across multiple markets, spreading risk and making participation more enduring.

Just-in-time (JIT) liquidity provides funding only when users need it. Bots monitor large trades in liquidity pools, injecting concentrated liquidity, collecting fees, and withdrawing immediately. This method is capital efficient but requires complex infrastructure and does not solve the fundamental problem: the risk is still borne by others. JIT strategies have generated over $750 billion in trading volume on Uniswap V3, but trading activity is primarily dominated by well-capitalized participants, yielding minimal returns.

Continuous composite markets challenge the binary structure itself. Traders are no longer limited to discrete "yes/no" options but can express views across a continuous range. This aggregates liquidity that would otherwise be scattered across related markets (Will Bitcoin rise to $60,000? $65,000? $70,000?). Projects like functionSPACE are building this infrastructure, although it has yet to undergo large-scale testing.

The most radical experiments completely abandon order books. Melee Markets applies bonding curves to prediction markets—each outcome has its own curve, early participants receive better prices, and believers are rewarded. No professional market makers are needed. XO Market requires creators to use LS-LMSR AMM to inject liquidity, with market depth increasing as funds flow in. Creators earn fees, linking the incentive mechanism to market quality.

Both solve the cold start problem without requiring professional market makers. Melee's downside is a lack of flexibility (positions are locked until settlement). XO Market allows continuous trading but requires creators to pre-fund.

Issue 2: Market Discovery and User Experience

Even if liquidity issues are resolved, there’s a more practical problem: most people cannot find the markets they care about, and even if they do, the experience is clunky.

This is not just a "user experience problem," but a structural issue. The market discovery problem exacerbates the liquidity issue. Polymarket has thousands of markets online at any given time, but trading volume is concentrated in a few areas: election markets, major sporting events, and popular cryptocurrency questions. Other markets go unnoticed. Even if a niche market has some depth, if users cannot naturally find it, trading volume will remain low, ultimately leading to market makers exiting. A vicious cycle: without market discovery, there’s no trading volume, and thus no sustainable liquidity.

The concentration of market liquidity is extremely severe. In the 2024 election cycle, Polymarket's top markets accounted for the vast majority of trading activity. After the elections, the platform still sees $650-800 million in trading volume monthly, but it is distributed across sports, cryptocurrencies, and viral markets. Thousands of other markets—such as local issues, niche communities, and oddities—are virtually ignored.

User experience barriers exacerbate this situation. The interfaces of Polymarket and Kalshi are designed for those already familiar with prediction markets. Average users face a steep learning curve: unfamiliar terminology, converting odds to probabilities, what it means to "buy a YES," etc. This is acceptable for crypto-native users. But for others, these frictions can kill conversion rates.

Better algorithms help, but the core issue is distribution: matching thousands of markets to the right users at the right time without causing choice paralysis.

What Entrepreneurs Are Trying

The most promising approach is to provide services directly on platforms users already have, rather than forcing them to learn new platforms. Flipr allows users to trade markets like Polymarket or Kalshi by directly tagging a bot in their Twitter feed. For example, when users see a market mentioned in a tweet, they simply tag @Flipr to trade without leaving the app. It embeds prediction markets into the conversational layer of the internet, transforming social information flows into trading interfaces. Flipr also offers up to 10x leverage and is developing features like copy trading and AI analysis—essentially striving to become a fully functional trading terminal that happens to exist on Twitter.

A deeper insight is that for startups, distribution is more important than infrastructure. Rather than spending millions like Polymarket to kickstart liquidity, it’s better to integrate existing liquidity and compete on distribution. Platforms like TradeFox, Stand, and Verso Trading are building unified interfaces that can aggregate odds from multiple platforms, route orders to the best trading venues, and integrate real-time news feeds. If you are a serious trader, why struggle to switch between multiple platforms when you can use a single interface with higher execution efficiency?

The most experimental approach is to view market discovery as a social problem rather than an algorithmic one. Fireplace, affiliated with Polymarket, emphasizes investing together with friends—recreating the vibrancy of collective betting rather than going it alone. AllianceDAO's Poll.fun takes it further: it builds P2P markets among small friend groups, allowing users to create markets on any topic, bet directly with peers, and have results decided by creators or group votes. This model is highly localized, highly social, and completely avoids the long-tail problem by focusing on community rather than scale.

These are not just improvements in user experience; they are distribution strategies. The platform that ultimately wins may not have the best liquidity or the most markets, but rather the one that can best answer the question: "How can we push prediction markets to the right users at the right time?"

Issue 3: The Problem of User Expression of Views

The following data should concern anyone optimistic about prediction markets: 85% of Polymarket traders have negative account balances.

To some extent, this is inevitable—predicting is inherently difficult. But part of the reason lies in the platform's hard flaws. Because traders cannot effectively express their views, the platform forces them to establish suboptimal positions. Do you have a nuanced theory? Too bad. You can only make binary bets: buy, don't buy, or choose position size. There’s no leverage to amplify your conviction, no way to consolidate multiple views into a single position, and no conditional outcomes. When traders cannot effectively express their beliefs, they either tie up too much capital or have positions that are too small. In either case, the platform captures less traffic.

This problem can be divided into two distinctly different needs: traders who want to leverage to amplify single bets and those who want to combine multiple views into a single bet.

Leverage: Continuous Settlement Solutions

Traditional leverage strategies do not apply to binary prediction markets. Even if your directional prediction is correct, market fluctuations can wipe you out before settlement. For example, a leveraged position on "Trump wins" could be liquidated during a week of poor polling results, while Trump ultimately wins in November.

But there are better ways: continuous settlement perpetual contracts based on real-time data streams. Seda is building true perpetual contract functionality based on Polymarket and Kalshi data, allowing positions to settle continuously rather than waiting for discrete event settlements. In September 2025, Seda enabled perpetual contracts for the Canelo vs. Crawford fight's real-time odds (initially at 1x leverage), proving the model's viability in sports betting.

Short-term binary options are another increasingly popular trading method. Limitless surpassed $10 million in trading volume in September 2025, offering binary options on cryptocurrency price movements, which provide implicit leverage through their payout structure while avoiding liquidation risks for traders during the contract's duration. Unlike fixed-income options, binary options settle at a fixed time, but their immediacy (hours or days rather than weeks) provides the rapid feedback retail traders need.

Infrastructure is rapidly maturing. Polymarket partnered with Chainlink in September 2025 to launch a 15-minute cryptocurrency price market. Perp.city and Narrative are experimenting with continuous information flow trading based on polling averages and social sentiment—true perpetual contracts that never yield binary results.

Hyperliquid's HIP-4 "event perpetual contracts" is a groundbreaking technology—it trades changing probabilities rather than just final outcomes. For example, if Trump's winning probability rises from 50% to 65% after a debate, you can profit without waiting for election day. This addresses the biggest issue with leveraged trading in prediction markets: even if the final prediction is correct, you may be forcibly liquidated due to market fluctuations. Platforms like Limitless and Seda are also gaining increasing attention with similar models, indicating that the market needs continuous trading rather than binary bets.

Composite Betting: An Unresolved Issue

Composite betting is different. It expresses complex, multifaceted hypotheses, such as: "Trump wins, Bitcoin price exceeds $100,000, and the Fed cuts rates twice." Sports bookmakers can easily do this because they operate as centralized entities managing dispersed risks. Conflicting positions offset each other, so they only need to collateralize for the maximum net loss rather than for each individual payout.

Prediction markets cannot do this. They act as custodial agents—every trade must be fully collateralized once completed. This quickly drives up costs: even small composite bets require market makers to lock up several orders of magnitude more capital than what sports bookmakers would need to cover equivalent risks.

The theoretical solution is a net margin system that only collateralizes for the maximum net loss. But this requires complex risk engines, real-time correlation modeling across unrelated events, and potentially centralized trading counterparts. Researcher Neel Daftary suggests that professional market makers underwrite limited market combinations before gradually scaling up. Kalshi has adopted this approach—initially offering composite bets only for events in the same context, as the platform can more easily model correlations and manage risks within a single event's framework. This approach is insightful but acknowledges that a true composite market, the "pick-and-choose" experience, may be difficult to achieve without centralized management.

Most prediction market entrepreneurs believe these novel prediction market formats have limitations: for example, leverage restrictions on short-term markets, pre-approved event combinations, or simplified "leveraged trading" that platforms can hedge. The issue of user expression of views may be partially addressed (e.g., through continuous settlement), but other aspects (e.g., arbitrary composite markets) remain out of reach for decentralized platforms.

Issue 4: Permissionless Market Creation

Solving the market expression problem is one thing, but a deeper structural issue is: who has the right to create markets?

Everyone agrees that prediction markets need diversity—events of significant regional interest, niche community concerns, and quirky one-off events that traditional platforms would never touch… But permissionless market creation has always been a challenge.

The core issue is that the lifecycle of hot topics is limited. The most explosive trading opportunities often arise from breaking news and cultural events. For example, a market like "Will the committee revoke Will Smith's Oscar for slapping Chris Rock?" can generate huge trading volumes within hours of the event. But by the time a centralized platform reviews and launches it, interest has already waned.

However, completely permissionless creation encounters three problems: semantic fragmentation (ten versions of the same question split liquidity into useless pools), liquidity cold starts (zero initial liquidity makes the chicken-and-egg problem extreme), and quality control (platforms are flooded with low-quality markets or worse—bets on assassination events that pose legal risks).

Both Polymarket and Kalshi have chosen a curated platform model. Their teams review all markets to ensure quality and clear solution standards. While this helps build trust, it sacrifices speed—the platform itself becomes a bottleneck.

What Entrepreneurs Are Trying

Melee adopts a strategy similar to pump.fun to address the cold start phase. Market creators receive 100 shares, while early buyers' shares decrease (3 shares, 2 shares, 1 share…). If the market gains traction, early participants will receive outsized returns—potential returns of up to 1,000 times or more. This is a "market of markets," where traders predict which markets themselves will grow by establishing early positions. The core idea is that only the highest-quality markets—those created by top creators or truly meeting market demand—can attract sufficient trading volume. Ultimately, quality markets will naturally rise to the top.

XO Market requires content creators to use LS-LMSR AMM to provide liquidity. Creators earn revenue by paying fees, linking the incentive mechanism to market quality. Opinion market platforms like Fact Machine and Opinions.fun allow influencers to monetize cultural capital by creating viral markets around subjective topics.

The theoretical ideal is a hybrid, community-driven model: users invest reputation and liquidity when creating markets, which are then reviewed by community administrators. This model allows for rapid permissionless creation while ensuring content quality. However, no mainstream platform has successfully implemented this model yet. The fundamental contradiction remains: permissionless creation can bring diversity, while administrators can ensure quality. Breaking this balance would unleash the localized, niche markets that the ecosystem needs.

Issue 5: Oracles and Settlement

Even if you solve liquidity, discovery, expression, and creation issues, there remains a fundamental question: who decides what happened?

Centralized platforms make decisions through their teams, which is efficient but carries the risk of a single point of failure. Decentralized platforms require oracle systems to handle arbitrary questions without ongoing human intervention. But determining the outcomes of these questions remains the hardest part.

As researcher Neel Daftary articulated for Delphi Digital, emerging solutions involve a multi-layer stack that routes questions to appropriate mechanisms:

Automated data feeds for objective outcomes. Polymarket integrated Chainlink in September 2025 for instant settlement of cryptocurrency price markets. Fast and highly deterministic.

AI agents for answering complex questions. Chainlink tested AI oracles on 1,660 Polymarket markets, achieving an accuracy rate of 89% (with sports events reaching 99.7%). Supra's Threshold AI oracle uses multi-agent committees to verify facts and detect manipulation, ultimately providing signed results.

Optimistic oracles like UMA are suitable for ambiguous questions, proposing several outcomes that disputing parties can fund to challenge. While it is based on game theory, it is effective for clear questions.

For high-risk disputes, reputation-based juries are employed, where voting power is linked to on-chain performance records, not just capital.

Infrastructure is rapidly maturing, but market settlement remains the trickiest issue. If a settlement scheme goes wrong, it undermines trust; if it is correct, it can scale to millions of markets.

Why These Issues Matter

Liquidity, market discovery, trader expression, market creation, and settlement are interconnected issues. Solving the liquidity problem can enhance market attractiveness, thereby improving market discovery mechanisms. Better market discovery can lead to more users, making permissionless market creation possible. More markets mean a greater demand for robust oracles. This is a system, and currently, the system has bottlenecks.

But opportunities arise: existing projects are trapped in established models. The successes of Polymarket and Kalshi are built on certain assumptions about how prediction markets operate. They are optimizing within established constraints. Meanwhile, the next generation of developers has the advantage of completely ignoring these constraints.

Melee can experiment with different bonding curves because their goal is not to become Polymarket. Flipr can embed leverage mechanisms into social information flows because they do not need regulatory approval in the U.S. Seda can generate perpetual contracts based on continuous data streams because they are not limited by binary parsing.

This is where prediction market entrepreneurs truly have their advantage. It’s not about replicating existing models but directly tackling fundamental issues. These five issues are the basic requirements. Platforms that can solve these problems will not only capture market share but also unleash the full potential of prediction markets as a coordinating mechanism.

2024 proved that prediction markets can be adopted on a large scale. 2026 will prove they can operate anywhere.

"Original Link" ```