[Editorial] Everyone Predicts, but 0.1% Take All the Money

People bet on who will become president, which team will win a soccer match, or what word a celebrity will say on a broadcast. Instead of showing odds, they display probabilities and attach buy and sell order windows. Those who place bets are referred to as 'investors' or 'traders.' This is the recent appearance of the rapidly growing prediction market. The screen resembles a stock trading app, but the structure of winning or losing money based on a single outcome remains the same.

Statistics have emerged showing who is making money in this market. According to an analysis reported by the Wall Street Journal in May of this year, among approximately 1.6 million Polymarket accounts that have traded since November 2022, 0.1% accounted for 67% of the analyzed profits. Fewer than 2,000 accounts took home about $500 million in profits. The number of accounts does not equate to the actual number of people, nor does it mean that all others incurred losses. However, it is clear that profits are concentrated among a very small minority.

The business is growing rapidly. The number of event contracts traded on Robinhood reached 4.7 billion in August of this year. Coinbase reported that its trading volume and revenue from prediction market contracts in the second quarter of this year increased by 106% compared to the previous quarter. This is a promising new business for the platform. However, this does not necessarily conclude that it is a promising financial investment for users.

The reason prediction markets attract people is not difficult to understand. While financial statements of stocks may be unfamiliar, soccer and politics are familiar topics. For those who watch the news daily and follow games, there may be a sense of "I know this much." However, knowing an event well and betting money on that event to make a profit are two different matters.

For instance, suppose you bought a contract that pays $1 if a certain team wins for 80 cents. If the team wins, you earn 20 cents; if it loses, you lose 80 cents. Even if you correctly predict four out of five times under the same conditions, one mistake can wipe out your earnings. After accounting for fees, you could end up losing money. It is possible to analyze diligently and achieve a high accuracy rate yet still lose money.

What is needed in this market is not merely the ability to predict outcomes correctly. It is the ability to judge more accurately than the expectations reflected in the current prices. This is why simply betting on a team that everyone thinks will win does not easily yield profits.

You must also consider your trading counterpart. Calshi has recruited its first dedicated institutional market maker from the Susquehanna group in 2024. Market makers are specialized firms that provide buy and sell prices and facilitate trades. This means that professional financial companies are entering a market where individuals place orders via their smartphones.

There is no issue with the participation of specialized firms themselves. They also play a role in facilitating trades and reducing price discrepancies. However, individuals need to be aware of what preparations their counterparts are making. Specialized firms analyze data, calculate prices, and spread risks across multiple trades. While individuals focus on the outcome of a single game, specialized firms can employ strategies that accumulate small profits across numerous trades.

The participation button is open to everyone. However, not everyone has the same level of information, analytical ability, order speed, or financial resources. Just because anyone can enter does not mean that everyone has similar odds of winning.

Prediction markets are not without value. They can aggregate the judgments of many people to provide information about future events. They may also serve as a means to reduce risks for companies that may incur losses due to specific events. There are also movements to utilize contracts related to weather, commodities, and corporate performance aimed at actual institutional clients.

However, it is not possible to equate trading aimed at reducing business risks with betting on whether a celebrity will say a specific word on a broadcast. Just because it is labeled as a prediction does not mean that all betting constitutes productive financial activity. We must be cautious of using the justification of risk management to turn every moment of daily life into a betting opportunity.

The differences from stock investments are also clear. Shareholders can share in the results as companies produce and increase profits. In simple event contracts, the contracts themselves do not generate new profits or dividends. Money is divided among participants based on the outcome, and transaction costs are deducted. While speculation exists in stocks, it is not appropriate to present the two products as the same means of asset formation.

The interests of the platform and the users may not align. Businesses that earn commissions have more opportunities to profit as trading increases. For users, stopping trading may be the way to reduce losses. We should not interpret the trading volume and revenue growth announced by the platform as a sign of user success.

Now, prediction market companies must disclose customers' profits and losses alongside trading volumes. They need to reveal how many accounts made money and how many lost over a certain period, whether profits remain after deducting fees, and where losses are concentrated. To showcase the few cases of significant profits, they must also present the results achieved by the majority.

Criteria for determining outcomes and managing conflicts of interest are also important. It must be clearly communicated before contracts what is recognized as a hit and what data will be used as a basis for judgment. They should also explain how they will prevent trades from individuals who may know the outcome in advance or influence the results. It is difficult to gain trust by merely adjusting rules every time an incident occurs after the market has grown.

If prediction markets are to be considered a new growth industry domestically, user protection should not be treated as a mere accessory to business expansion. Limits on losses, mechanisms for suspending use, and standards against excessive trading promotions should be designed from the outset. The media should also refrain from presenting probabilities displayed in prediction markets as if they were future answers.

The technology for predicting the future can advance. However, if that technology is used to stimulate people's confidence and impatience to make bets, the evaluation should change. Whether it is innovative is not determined by flashy order windows or trading volumes.

It is easy to allow everyone to place bets. It is the platform's responsibility to ensure that everyone understands the risks properly.