Prediction markets have reshaped the information aggregation model by converting collective wisdom into tradable event contracts. This report takes Kalshi (regulated U.S. benchmark) and Polymarket (crypto-native platform) as research objects, analyzing their operational mechanisms, pricing logic, and liquidity supply models. The study finds that the core value of prediction markets lies in "information pricing" rather than speculation, and their legality stems from verifiable outcomes and controllable risks.
- Kalshi: Co-founded by Luana Lopes Lara (youngest self-made female billionaire at 29), holds the first federal CFTC license
- Polymarket: Peter Thiel's Founders Fund led a $200M round at $1B valuation; ICE invested up to $2B at ~$8B valuation
Kalshi vs. Polymarket Comparison
| Dimension | Kalshi (Centralized Regulated) | Polymarket (Crypto-Native) |
|---|---|---|
| Regulatory | U.S. CFTC DCM license | Seychelles FSA; complies with select U.S. states |
| Settlement | U.S. Dollar (fiat) | Stablecoins such as USDC |
| Core Technology | Centralized order book + regulatory risk control | Polygon blockchain + off-chain matching + on-chain settlement |
| Liquidity Model | User-driven maker-taker | CLOB + maker fee rebates + quadratic scoring |
| Target Users | Institutional investors, professional traders | Retail users, crypto asset holders |
Kalshi: Regulated Fiat-Based Mechanism
Kalshi strictly limits events to "verifiable non-gambling events" with three contract types:
- Binary Yes/No Contracts: Clear dual outcomes
- Numerical Range Contracts: Quantitative outcomes (e.g., CPI above 0.3%?)
- Non-Exclusive Contracts: Multiple potential results
All contracts priced between 1¢ and 99¢, with $1 payout if the event occurs.
Polymarket: Two Hard Constraints
Cornerstone 1: Mathematical Constraint P(Occur) + P(Not Occur) = 1
Cornerstone 2: Financial Constraint (Prices Sum to $1) The "One Dollar Redemption Guarantee" ensures arbitrage opportunities correct any price deviations.
Kalshi's Fee Structure
Fee = 0.07 × p × (1-p) where p = contract price in dollars
This creates a "U-shaped" fee characteristic: highest at 50¢, lowest at extremes.
Favorite-Longshot Bias
Empirical research shows longshot contracts (1-10¢) only win ~38% of the time relative to break-even probability, while favorites (90-99¢) win 5% more often than required.
Polymarket's Fees
- Trading Fees: None on main platform
- Net Winnings Fee: 2% on profits
- Gas Fees: Small MATIC costs for blockchain transactions
| Dimension | Prediction Markets | Gambling |
|---|---|---|
| Core Purpose | Information aggregation and price discovery | Entertainment and pastime |
| Outcome | Objectively verifiable by third parties | Random or intervenable |
| Risk-Return | Positive expected returns via information | Negative expected returns (house edge) |
| Regulatory | Financial regulation (KYC/AML) | Gambling regulation |
Core Challenges and Proposed Solutions
| Challenge | Proposed Solution |
|---|---|
| Ambiguous Regulatory Classification | Rely on HKMA's "Fintech Sandbox"; clarify "information-based derivatives" |
| Retail User Risk | "Institutional priority" model; asset certification thresholds |
| Insufficient Liquidity | Polymarket-style incentives; connect local brokers |
| Cross-Border Compliance | Refer to EU MiCA and U.S. GENIUS Act standards |
| Technology Adaptation | Combine Kalshi's centralized risk control + Polymarket's blockchain |
The rise of Kalshi and Polymarket marks the transformation of prediction markets from academic concepts to commercially mature financial tools. For Hong Kong, developing a prediction market ecosystem is not only a natural extension of its fintech layout, but also an important grasp to enhance its global information pricing power.
By following the path of "clarifying regulatory positioning through sandbox pilots, controlling risks with institutional priority, and ensuring liquidity with incentive mechanisms," Hong Kong is fully capable of building an "Eastern Prediction Market Hub."
Standard Kepler Research | standardkepler.com