I wrote the following as a "research report". You can use it directly as a material, and it will also facilitate the connection with the IFC / ISO / ICR system later.

Prediction Market Research Report (2025 Short Version)

1. Brief description of concepts and mechanisms

1. What is a prediction market?

Prediction Markets, also called Event Contracts, are essentially:

Allow participants to "bet" on the outcomes of future events (elections, policies, economic data, sports scores, technological milestones, etc.),

The price is the collective judgment of the group on the probability of the event occurring.

Typical form:

• "A certain candidate will be elected president in 2028?" - "Yes/No" contract, price 0–1 (or 0–100);
• “Will AI pass a certain benchmark test by 2026?”
• “Does a company’s revenue exceed $X billion in 2025Q4?”

The contract price approaches the subjective probability: for example, the price is at 0.73, which means the market believes that the probability of the event happening is about 73%.

2. Basic transaction structure

• The user buys the "YES/NO" contract;
• If the event occurs at expiration, the "YES" contract will be redeemed for 1 US dollar, and the "NO" contract will be reset to zero, and vice versa;
• You can sell it at any time like a stock;
• The platform makes money through spreads, handling fees or other value-added services.

3. Comparison with traditional forecasting methods

• Compared with expert predictions/research questionnaires, the prediction market has three advantages: 1) Participants have to express their opinions with "real money or limited chips" and there is less noise; 2) Prices are dynamic and continuously updated, which can reflect new information in real time; 3) Correction through the arbitrage mechanism: when the price deviates from the true probability, it will attract more rational participants to trade in the opposite direction.

2. Academic research conclusion: Is prediction market “really useful”?

1. Overall conclusion on accuracy

• Wolfers & Zitzewitz's data analysis of multiple electoral, macroeconomic and other scenarios shows that probabilistic forecasts of market output often outperform traditional methods such as polls and simple averages.
• The systematic review and Meta-Analysis found that: On average, prediction markets are approximately 79% more accurate overall than other alternative prediction methods (experts, questionnaires, simple models, etc.).

2. Typical experiment: IARPA prediction competition

• The U.S. Intelligence Advanced Research Projects Agency (IARPA) has organized prediction competitions spanning many years and thousands of problems: multiple research teams participated, using prediction markets, superforecasters, combination methods, etc. to compete;
• The results show that a carefully designed combination of prediction markets and top forecasters can significantly outperform traditional intelligence analysis and polling in long-term predictions.

3. “Internal Prediction Market”: Enterprise Scenario

• Internal enterprise prediction markets (such as Google, HP and other companies have done) are mainly used for: product launch time
• Sales target completion
• Project risk assessment

3. Current industrial structure and mainstream platforms

It can be roughly divided into three categories: regulated legal currency prediction markets, cryptocurrency prediction markets, and virtual currency/reputation prediction games.

1. Regulated fiat currency prediction market: Kalshi system

• Kalshi: Founded in 2018, headquartered in New York, USA;
• It has been authorized by the US CFTC to become a Designated Contract Market (DCM) and is currently one of the most important "compliance event contract exchanges".
• Contract coverage: inflation rate, interest rate, unemployment rate, election results, government shutdown, sports event indicators, etc.

Features:

• It has strong compliance and is suitable as a model for "financial derivatives + information market";
• The regulatory game is fierce and has a huge impact on product design and expansion pace.

2. Crypto prediction market: Polymarket, etc.

• Polymarket: A decentralized prediction market platform denominated in stablecoins such as USDC;
• Transaction volume hit a peak during the 2024 U.S. election, and there will be a second explosion in 2025 through token incentives.

Features:

• Growing extremely fast, event coverage is almost borderless (politics, sports, entertainment, encryption, technology...);
• Subject to regulatory restrictions in the United States (such as the previous settlement with the CFTC), it is mainly targeted at non-U.S. users or in gray areas;
• Token incentives make it more like a hybrid of “DeFi + Social + Speculation”.

3. Virtual currency/reputation-based prediction market: Manifold et al.

• Manifold Markets / Manifold: Use virtual currency Mana for trading, which is essentially a “toy money + reputation points” type prediction platform;
• Users can create a market on any topic, and the platform provides liquidity through the automatic market maker algorithm (CFMM/Maniswap);
• Emphasis on “predictive games + social”, not strictly financial products;
• The measured calibration data shows that even with "virtual currency + non-monetary income", the prediction accuracy still has reference value.

Other forms:

• Metaculus: a scientific research and technology prediction community;
• Open source project PlayMoney: a community-built prediction market framework.

4. Regulatory and legal environment

1. United States

• Federal level: The CFTC is responsible for regulating whether "event contracts" are legal derivatives;
• Kalshi, as a DCM, is under strict supervision.

Core controversy:

• Is the prediction market an “information market/hedging tool” or a “disguised gambling”?
• The regulatory boundaries have not yet been fully clarified, and the game will continue in the coming years.

2. EU, UK, etc.

• The overall strategy is cautious, incorporating more prediction markets into the financial derivatives/gambling regulatory framework, and strictly limiting large-scale promotion to retail investors;
• At the same time, it is relatively relaxed in scientific research, policy consultation, and internal prediction experiments.

3. China and East Asia

• China's supervision of "online gambling/lottery/feudal superstition" is very strict, and a monetary prediction market directly facing the public is basically unfeasible;
• But as: internal enterprise tools (e.g. project risk forecasting),
• Teaching/scientific research tools (virtual currency/statistics only, no redemption),
• There is still room for gamified Q&A (no money or winning or losing involved), but it just needs to be very careful about compliance and narrative.

5. Typical application scenarios

1. Politics and Macroeconomics: Election results, policy implementation, inflation rate, interest rates, unemployment rate, etc. - are currently one of the most active trading directions.
2. Sports: Championship ownership, game scores, player data, etc.;
3. However, it highly overlaps with traditional sports betting and is one of the most controversial areas of current regulation.
4. Technology and scientific research: Predict the results of a certain scientific research project, paper publication, technological breakthrough time, etc.;
5. For example, the prediction market is used to predict the scientific research evaluation results (REF) of British universities, and the effect is better than traditional expert scoring.
6. Enterprise management and product decision-making: Use the internal prediction market to replace "head-beating": new product launch time, sales target, project delay probability, etc.
7. Community governance and public decision-making: DAO, open source communities, etc. can use prediction markets to estimate proposal approval rates, changes in community activity, token price ranges, etc.

6. Opportunity and risk assessment (an inspiration for you)

1. Opportunity points

• As a "collective intelligence aggregation layer": combined with large models/Agents to form an "AI + human predictor" collaborative system;
• Form a more sensitive "emotion & expectation radar" in terms of macro, technology trends, policy changes, asset pricing, etc.

2. Risk points

• Compliance: Any product involving the winning or losing of real money may be regarded as "gambling/gambling", with high compliance costs and great regulatory uncertainty;

7. Summary

• From the perspective of academic research: the prediction market is indeed a high-quality collective prediction mechanism in most scenarios, and is overall better than experts and polls;
• From the perspective of industry data: The transaction volume and number of users of platforms such as Kalshi and Polymarket will increase rapidly in 2025, and a multi-billion-dollar "information derivatives market" has been formed;