Prediction markets are suddenly so popular. What can you do?

Author: Shao Jiaio

In the past year, as long as you participated in several Web3, quantitative trading or US stock-related industry events, you would almost hear one word:Prediction Markets.

One side, yesKalshiObtained a certificate from the U.S. CFTC (U.S. Commodity Futures Trading Commission)DCM (Designated Contract Market)Qualification, for the first time, the “event contract” was officially put into the federal financial regulatory system;

The other side isPolymarketFirst, it was fined US$1.4 million by the CFTC and its US users were withdrawn. Then, through a series of structural and product adjustments, it continued to grow rapidly around the world.Become synonymous with “on-chain prediction market”.

In the excitement, I have been asked the same question repeatedly recently:“The prediction market is so popular right now, is there anything I can do?”

In this article, I want to make three points clear about this matter:

• prediction marketWhat exactly is it and what is the supervision struggling with?;

• Prediction market related entrepreneurship,Which directions are relatively doable and which are high-risk areas?;

• And the most important thing:Are you making information tools, or have you quietly turned into a trading service?

What exactly are prediction markets?Don’t just think of it as a “bet”

From a technical and economic perspective, prediction markets are not mysterious.Its basic logic is very simple:

• Users use real money to invest in a certainwhether future events will occurconduct transactions;

• The contract price is essentially the market’s collective judgment on the probability of the event occurring;

• After the outcome of the event is determined, the contract will be settled based on the outcome.

What is different from traditional derivatives such as futures and options is that:Predicting market trading is not about “price movement” but “if it happens”.

For example:

• Will the Fed raise interest rates at its next meeting?

• Can a certain candidate win the election?

• Will a policy be passed by a specified time?

These areevent-based contracts.

The real question is: Who should regulate it?

The core of the regulatory differences in prediction markets around the world is not “whether it is like gambling”, but rather:Should it be included in the financial regulatory system?

In reality, there are three common processing paths:

1. Financial derivatives path

• Typical representative: United States

• Regulated by the CFTC, event contracts are considered a special type of derivatives

• Kalshi is a sample of this route

2. Betting/Gambling Path

• In many European countries and some Asian jurisdictions

• Prediction markets are directly regarded as “remote gambling” or “disguised gambling”

• The result is often: ban without a license

3. Gray area + bullish regulatory game

• Touch financial supervision, gambling supervision, and consumer protection at the same time

• It is this complex range that Polymarket is in

The same product may be “financial innovation” or “illegal gambling” in different countries.

Why did the prediction market suddenly go out of business in the past year?

This is no accident. There are at least four real forces superimposed behind it.

1. Supervision begins to confront each other head-on, rather than avoiding them

Kalshi’s DCM qualification means more than just “having a compliance platform.”

What’s really important is:

The CFTC began to respond head-on to a question: Can event contracts become part of a serious financial market?

At the same time, the game between the CFTC and the market has also significantly escalated around the contractual boundaries of political events, sports events, and public interest issues.

This means:The prediction market has entered a stage where legal boundaries are being rewritten, rather than a testing period.

2. Polymarket is not “whitewashing”, but “changing the battlefield”

Many people interpret the Polymarket story as:

“Penalized → Compliance → Return to the United States”

But from a lawyer’s perspective, a more realistic description is:

Polymarket did not become Kalshi, but put itself in a more complex, but short-term feasible regulatory gap through product structure, user positioning and technical architecture.

Its path is essentially:

• De-Americanize users;

• No legal currency or direct access to the banking system;

• Use on-chain settlement and stablecoins to weaken traditional financial regulatory connection points.

This isEngineering Capabilities for Regulatory Boundary Management, rather than “compliance success examples”.

3. AI and prediction markets naturally fit together in terms of product logic.

What do large models do best?

• information integration

• probability judgment

• Scenario analysis

The essence of the prediction market is precisely:Compress scattered information into a “probability price”.

This is why in the past year, we have seen more and more products trying:

• AI-assisted information screening;

• AI-generated event analysis summaries;

• Cross-validation of AI + market prices.

But please note:

AI participates in the “information and decision-making layer” and is not naturally suitable for “automatic order placement”.

4. Macro reality: Predicting “events” tells a better story than predicting “prices”

In a macro environment of high volatility and low certainty:

• Interest rates, regulation, elections, geopolitics, are increasingly sources of alpha;

• Compared to predicting asset prices,Predict “whether something will happen”, more intuitive for ordinary users;

• For supervision, it is also easier to be structured and discussed.

This is also a realistic reason why the prediction market will heat up significantly in 2024–2025.

If you are an entrepreneur: In the prediction market, which directions are relatively “doable”?

To see this clearly,Let me give you a more essential dismantling method first.

Entrepreneurship related to prediction markets is essentially divided into four levels:

1. Information layer: data, visualization, aggregation;

2. Decision-making layer: strategy, signal, probability judgment;

3. Execution layer: placing orders, following orders, and automated trading;

4. Clearing layer: the market itself, settlement rules, and capital flows.

Which floor are you standing on?Determines what your regulatory attributes are.

1. Information layer: data aggregation/search/visualization (lowest risk)

This layer only does three things:

• Aggregate prediction market data;

• Do visualization, screening, and ranking;

• Help users “understand what the market is betting on.”

Not trusting funds or placing orders on behalf of customers is a life-or-death situation.

Typical forms include:

• Multi-platform odds and volume aggregation;

• Popular events, price curves, sentiment changes;

• Data panel used by media, institutions, and researchers.

At this level, in most legal jurisdictions,Closer to information services or alternative data providers.

2. Decision-making level: arbitrage/strategy/signaling tools (high demand, but restraint)

There are natural price differences in the prediction market:

• The same event has different prices on different platforms;

• There is room for hedging with traditional financial markets and crypto derivatives.

Therefore, doStrategy center, signal scanning, arbitrage tips, the business needs are real.

The key differences are:Do you “prompt the opportunity” or “execute it for him”?

The former is a tool, the latter can easily be identified as:

• Investment advisory services

• Brokerage services

• Even the prototype of asset management

3. Execution layer: copying/automated trading (the most profitable, but also the most dangerous)

The reason why copying is attractive is not that it allows users to make money, but that it is extremely easy to turn into a sustainable source of income for the platform.Because of this, it is often the first function targeted by regulation.

But from a regulatory perspective:

Copy trading + automatic execution are often regarded as a combined service of “investment advice + client execution”.

In an already regulatory-sensitive track like the prediction market, this level of risk will be further amplified.

There is only one key boundary here:

Does the user still retain the “right of final confirmation”?

4. Liquidation layer: Make your own prediction market (focusing on supervision route)

If what you want to do is:

• A market with settlement rules;

• A platform for hosting funds and matching transactions;

Then what you are facing is not “Web3 entrepreneurship”, but:The twin challenges of financial market infrastructure + gambling regulation.

This path is not impossible, but it must be:

• high cost

• long cycle

• Strong regulatory game

The most common and dangerous pitfalls in prediction markets

1. Escrow funds and transfer on behalf of others

As long as you touch these actions, two words will basically appear in the regulatory language:

• Custody

• Money Services

Whether this step is necessary and worthwhile is something every team must think clearly about.

2. AI directly gives “buying and selling advice” + one-click order placement

AI summarizes and organizes information, so there is no big problem; AI gives clear trading suggestions and cooperates with automatic execution.It is already very close to regulated financial services in many jurisdictions.

3. Dare to talk about any topic: elections, sports, public events

Topic pool design itself is a compliance issue.Some events are naturally off-limits in some countries; platform rules themselves are becoming “quasi-regulation.”

4. Platform currency + rebate + income narrative

Prediction markets are gray enough.If another one is superimposedToken with loose design, it is easy to push yourself to:

• Securities Regulation

• Gambling Regulation

• Prepaid card/fund pool supervision

Finally, let me be honest: this is not a track where you can “think clearly and do it”

The real complexity of the prediction market lies not in the product form nor in the technical implementation.From a regulatory perspective, the question has never been “are you predicting the market?” but –Where are you and what role do you play?

When many teams introduce their products, they will emphasize:

• I am just a tool;

• Does not provide investment advice itself;

• You are not responsible for the results.

But in the real world,Roles are not determined by self-declaration.

When supervision judges a project, it does not start from your white paper or disclaimer, but directly looks at the three simplest questions:

• Has the user changed their trading behavior because of you?

• Did the funds flow because of you?

• Are risks amplified or concentrated because of you?

As long as “yes” appears in the answer once, your project is no longer a “peripheral tool” from a regulatory perspective.

The reason why the prediction market has been repeatedly controversial is precisely because it is naturally vague:

• It’s like finance, but not quite finance;

• It’s like gambling, but it’s cloaked in the guise of information efficiency;

• It stands at the intersection of public events, price signals and user sentiment.

This means that this trackThere is no “one-time design and everything will be fine” solution.Every functional choice you make today is essentially defining a certain kind of supervision in the future.Bet in advance.

Therefore, if I must say a conclusion, it should be:

It’s not that prediction markets can’t be done, but you have to accept it: it is a track that does not allow for long-term survival by ambiguity and luck.

The real danger is never the supervision itself, but that you have pushed yourself into a situation without knowing it.in a position that needs to be supervised.