Author: Jiawei @0xjiawei
The prediction market is undoubtedly one of the most talked about tracks in crypto this year. The leader, Polymarket, has a cumulative volume of more than $36 billion and recently closed a strategic funding round at a valuation of $9 billion. Meanwhile, platforms such as Kalshi (valued at $11 billion) have also received large capital raises.
However, behind this strong influx of capital and impressive growth, we see that the prediction market, as a trading product, is still facing some structural problems.
In this article, I will try to put aside the mainstream optimism and provide some observations from a different perspective.
- Predictions are event-based — events are inherently discontinuous and unrepeatable.
Compared to the price of assets such as stocks and FX, which change over time, the prediction market relies on a limited and discrete number of events in the real world. It is low-frequency relative to trading.
There are only a limited number of real-world events with truly widespread interest, clear outcomes, and settlements within a reasonable period of time — presidential elections happen once every 4 years, the World Cup happens once every 4 years, the Oscars happen once a year, and so on.
Most social, political, economic, and technological events do not have ongoing trading needs. The limited number and low frequency of such events per year makes it difficult to build a stable trading ecosystem.
In other words, the low-frequency nature of the prediction market is not something that can be easily changed by product design or incentives. This underlying characteristic determines that the trading volume of the prediction market will inevitably not remain at a high level when there are no major events.
2. Prediction markets do not have fundamentals like the stock market: the source of value in the stock market is the intrinsic value of a company, including its future cash flow, profitability, assets, etc.
Prediction markets, on the other hand, ultimately point to an outcome, and rely on the user’s interest in the outcome of the event itself.
(Of course, we discuss the original purpose of the product here, excluding arbitrage, speculation, etc.; even the stock market has many speculators who don’t necessarily care about the nature of the underlying assets.)
In this context, the money people are willing to bet has a significant positive correlation with the significance of the event, the level of attention, and the time period: scarce, high-profile events, such as the finals, presidential elections, and so on, will attract a large amount of money and attention.
It stands to reason that a casual fan would be more likely to care about the outcome of the annual finals and place heavy bets on it, and less likely to do so in the regular season.
On Polymarket, the 2024 presidential election alone accounts for over 70% of the platform’s total OI. At the same time, the vast majority of events are chronically low liquidity and high bid-ask spreads. At this level, it is difficult to scale the prediction market exponentially.
3. The prediction market itself has a betting nature, but it is difficult to generate the retention and expansion that betting does.
We all know that the real addiction mechanism of gambling is instant feedback — slot machines are played once every few seconds, Holdem is played once every few minutes, and perps and memecoin trades change rapidly every second.
Prediction markets have long feedback cycles, with most events taking weeks to months to settle. If it’s a fast-feedback event, it’s not necessarily interesting enough to warrant a heavy bet.
Immediate positive feedback significantly increases the frequency of dopamine release and reinforces user habits. Delayed feedback, on the other hand, does not create stable user retention.
4. In some types of events, there is a high degree of information asymmetry between participants.
For competitive sports events, in addition to the paper strength of the teams, much depends on the athletes’ on-field performance, so there is still a large degree of uncertainty.
But for political events involving internal information, channels, connections and other black-box processes, insiders have a great information advantage, and the certainty of their bets is much higher.
For example, the vote counting process, internal polls, and the organization of key regions in an election are very difficult for outside participants to access. Currently, there is no clear definition of “insider trading” in the prediction market, and this part of the market is still a gray area.
In general, the party with the least information can easily become the exit liquidity.
5. Due to the ambiguity of language and definitions, it is also difficult to be completely objective about the events in the prediction markets themselves.
For example, “whether there will be a ceasefire between Russia and Ukraine in 2025” depends on which statistical caliber is used; “whether a cryptocurrency ETF will be approved at a certain point in time”, which may be fully approved, partially approved, conditionally approved, etc.
This brings us to the issue of “social consensus” — in a situation where two parties are 50/50, the loser will not honestly admit defeat.
Such ambiguity requires the platform to establish a dispute resolution mechanism. And once a prediction market touches on linguistic ambiguity and dispute resolution, it can’t fully rely on automation, and there is room for human manipulation and corruption.
6. The main value proposition of prediction markets in the market is “collective wisdom”, i.e., that prediction markets can bring together the best information globally to achieve a group consensus, over and above the low level of trust in the media and mainstream discourse.
However, until prediction markets reach a very large scale of adoption, this “information sampling” will necessarily be one-sided and the sample will not be sufficiently diverse. The user base of a prediction market platform can be highly homogenized.
For example, in the early stages of a prediction market, it will certainly be a platform comprised primarily of cryptocurrency users, whose views on political, social, and economic events are likely to be highly homogenized, creating a cocoon of information.
In this case, the market reflects the collective bias of a particular circle, which is still a long way from “collective wisdom”.
Conclusion
The point of this article is not to be bearish on the market, but to keep our cool in the midst of fomo sentiment. Over-reliance on special events like the election, short-term sentiment on social media and airdrop incentives tends to amplify data appearances and isn’t enough to support a judgment on long-term growth.
Nonetheless, the prediction market still has an important place in the next three to five years from a user education and user attraction perspective. Similar to on-chain income savings products, they have an intuitive product format and lower learning costs, and have a better chance of attracting web2 users to crypto than others. Based on this, there is a high probability that prediction markets will grow further and become somewhat of an entry-level product for the crypto industry.
Future prediction markets are also likely to occupy certain verticals, such as sports and politics. They will continue to exist and expand, but they do not have the potential for exponential growth in the long term. It is more important to think about investing in the prediction market with a cautiously optimistic perspective.
Pouring Cold Water on the Prediction Market was originally published in IOSG Ventures on Medium, where people are continuing the conversation by highlighting and responding to this story.
