Prediction markets transform what were once private or illegal games into publicly available, serialized data of on-chain data and collective expectations. This creates real-time probability signals and alternative data, which financial institutions, data providers, and AI models can use as an economic mechanism for information aggregation and probability estimation. Regulation has created a fragmented system: prediction markets tend to be institutionalized in the West, while they are suppressed in Asia. This constitutes a major short-term constraint, but it has also paved the way for prediction markets to evolve into "infrastructure that transforms collective beliefs into information and markets." 1. How HIP-3 is enabling a new PvE-style growth pattern. The business models of trading platforms are changing. Centralized exchanges (CEXs) maintain their position due to their structural advantages based on institutional trust (fiat currency entry and exit, custody, and regulatory access). This makes them a natural gateway for institutional capital and provides stability in terms of liquidity and operational reliability. However, the same regulatory obligations, internal controls, and custody infrastructure also generate high fixed costs. As a result, CEXs are slower to experiment and make decisions, limiting their pace of innovation. In contrast, decentralized exchanges (DEXs) grow through incentive structures. They natively coordinate rewards between LPs, traders, and builders on-chain. Previously, launching a new exchange or market required teams to build a matching engine, margin and clearing systems, and oracles from scratch. This created an extremely high technical barrier to entry. HIP-3 removes this barrier. Hyperliquid now allows anyone who stakes 500,000 HYPE to deploy their own perpetual contract market using the same CLOB engine, margin logic, and liquidation system as the main site. The technical burden of building a trading platform has disappeared. Market creation has become a standardized on-chain deployment process that requires capital and reliable oracles, not an entire engineering team. The barrier has shifted from technical capabilities to capital and oracle design. This change is not just about increased efficiency; it has transformed where innovation occurs. Builders can now experiment with different liquidity structures, fee designs, oracle definitions, and leverage limits without rebuilding the backend. The challenge has become identifying the “demand surface” (i.e., how many people want to speculate on something) and anchoring it to reliable oracles. In reality, the market can now consist of three components: **market + oracle + demand**. This expands the range of tradable assets. As Ventuals founder Alvin Hsia describes it, the "Fat Head" comprises asset classes already covered by traditional finance (index products, forex, commodities); the "Chunky Middle" includes equity financing, real-world datasets, and commodity indices; and the "Long Tail" extends to niche signals such as localized real estate prices, product premiums, or cultural trend indices. Traditional finance cannot easily commoditize these data points, but on-chain settlement systems can. HIP-3 effectively opens up a demand-driven market creation model.

Source: X (@alvinhsia)
This transforms DEXs from competitors to CEXs into a structurally different entity.
HIP-3 is no longer about competing for fixed crypto-native liquidity (PvP dynamics), but allows DEXs to expand into non-crypto assets and real-world data. This brings new traffic, new users, and new forms of demand—a PvE dynamic of **constantly increasing market size rather than redistribution**. It also deepened revenue at the protocol layer. A prime example is Hyperliquid's XYZ100 market, which surpassed $1.3 billion in cumulative trading volume within three weeks of its launch, demonstrating the rapid scaling of new asset classes once infrastructure is standardized. In short, CEXs continue to offer stability and regulatory access, but HIP-3-based perpetual DEXs have gained an advantage in speed, experimentation, and asset expansion. They are not substitutes, but rather distinct growth paths. The competitive advantage of trading platforms will shift from back-end engineering to market design and user experience, and leadership will depend on which protocol can translate that into sustainable value. 2. From Narrative-Driven Valuation to Cash Flow-Driven Valuation The market in 2025 is fundamentally different from previous cycles. The era of abundant liquidity that once fueled all assets is over. Capital now flows selectively. Prices reflect actual performance more than narrative, and projects that fail to generate revenue are being naturally squeezed out. Most altcoins have yet to recover to their 2021 highs, while protocols with clear revenue streams have shown relative strength even during market corrections. The arrival of institutional capital has solidified this shift. Traditional finance (TradFi) frameworks are being applied directly to the crypto space. Revenue, net profit, fee generation, user activity, and profit distribution are becoming the primary metrics for project evaluation. The market is moving away from valuations based on "storytelling" or anticipated growth. Only projects with real revenue flowing back to tokens will receive higher market valuations. In this context, Uniswap's recent proposal to activate the fee switch is symbolic. A flagship DeFi protocol explicitly choosing to link cash flow to token value signals that fundamentals (rather than narrative) are now at the heart of market pricing. A clear group of frontrunners has emerged. Hyperliquid (HYPE) and Pump.fun (PUMP) are typical examples: Hyperliquid is the largest perpetual DEX in terms of trading volume, open interest (OI), and number of traders. As of November 2025, its cumulative trading volume reached $3.1 trillion, and its open interest reached $9 billion. Notably, Hyperliquid uses 99% of its perpetual contract fees to buy back HYPE, directly linking the protocol's cash flow to the token's value. Total buybacks have reached 34.4 million HYPE (approximately $1.3 billion), representing about 10% of the circulating supply. Pump.fun, a leading meme exchange, has generated approximately $1.1 billion in cumulative fees. Its buyback program has purchased approximately 830,000 SOL (approximately $165 million), equivalent to 10.3% of its (estimated) circulating value. Other projects also demonstrated strong revenue momentum: Aave (AAVE) and Jupiter (JUP) continued to generate stable and growing cash flows. Aave's annual revenue grew from $29.75 million in 2023 to $99.39 million in 2025. Jupiter's revenue growth was even more impressive, soaring from $1.42 million in 2023 to $246 million in 2025. Coinbase (COIN), despite being a publicly traded stock, is also benefiting from the increasingly clear path to its Coinbase blockchain token. Coinbase has broadened its revenue structure: subscription and service revenue reached $746.7 million in Q3 2025 (a 13.9% increase quarter-over-quarter). This shift is spreading from individual dApps to the L1 and L2 ecosystems. Technical prowess or investor endorsement alone are no longer sufficient. Chains with real users, real transactions, and protocol-level revenue are gaining stronger market recognition. The core evaluation metric is shifting towards the sustainability of economic activity. In short, the market is undergoing a structural transformation. The market in 2026 may be restructured around these performance-driven participants. 3. Quantifying Market Expectations Through Prediction Markets Prediction markets are an experiment that transforms previously private or illegal gambling activities into public on-chain data. At their core, they quantify the probability people assign to future events by getting people to invest real money in their beliefs. This makes them not just gambling venues, but economic mechanisms for aggregating information and estimating probabilities.

The prediction market has grown rapidly since 2024: As of October 2025, weekly nominal turnover was approximately $2.5 billion, with over 8 million transactions per week. Polymarket holds a 70–75% share of activity, while Kalshi's share climbed to around 20% after receiving CFTC approval and expanding into the sports and political markets. The uniqueness of prediction market data lies in the fact that: Polls, social media sentiment, and institutional research are often slow to react and costly. Prediction markets, on the other hand, price expectations in real time. For example, Polymarket reflected an increase in Donald Trump's probability of winning the 2024 election significantly earlier than traditional polls. In fact, prediction markets create serialized data of collective expectations. These curves serve as real-time probability signals for political, economic, sporting, and technological events. Financial institutions and AI models are increasingly viewing these markets as alternative data sources (Alt-data) for quantifying expectations. From an institutional perspective, prediction markets represent not "the datafication of gambling," but rather "the financialization of uncertainty." Because prices reflect consensus probabilities, macro traders can use them to manage risk. Kalshi has already provided markets linked to inflation, employment data, and interest rate decisions, attracting significant hedging interest. As prediction markets mature, they create a new value chain: Market (generating signals) → Oracle (solving outcomes) → Data (standardized datasets) → Applications (finance, media, AI, consumer). Current obstacles are primarily regulation: In Asia, regions like South Korea, Singapore, and Thailand largely prohibit prediction markets, classifying them as illegal gambling and penalizing users. In the West, the US regulates prediction markets as "event contracts" under the CFTC. Kalshi operates legally with a DCM license, while Polymarket plans to re-enter the US market in 2025 through the acquisition of QCX. This regulatory disparity has created a divide: the West is moving towards institutionalization, while Asia is suppressing it. While this is a short-term constraint, in the long run, prediction markets will evolve into infrastructures that translate collective beliefs into information. They will shift from "markets that interpret information" to "markets that produce information," reinforcing a world where "prices become the primary expression of collective expectations."