Original title: Prediction Markets: They Grow Up So Fast
Original author: Alex Immerman, a16z
Original translation by Peggy, BlockBeats
Editor's Note: For a long time, prediction markets have been regarded as a "fringe product": first an academic experiment, then a tool for public opinion during election season, and later seen as an extension of sports betting. It always seems to depend on a high-profile scenario, but is rarely understood as a financial infrastructure.
However, in the author's view, prediction markets are gradually evolving from a marginal "event trading tool" focused on elections and sports into a financial infrastructure that can price uncertainty.
The author points out that the key changes in the industry are reflected in three aspects: First, the application scenarios are expanding. Although sports are still the traffic entry point, long-tail markets such as entertainment, macroeconomics, and CPI are growing faster and beginning to meet institutional demand. Second, the prediction market has for the first time provided a tradable price benchmark for the "event itself," enabling institutions to directly hedge political or macroeconomic risks instead of "secondary betting" through related assets. Third, the institutional adoption path is progressing, from data reference (looking at odds) to system integration, and then to actual participation in trading, which is still in its early stages.
The prediction market is undergoing a process similar to the early stages of the options market, namely "professionalization-institutionalization-infrastructureization". In the future, once liquidity, leverage and regulation are gradually improved, it may become a core market tool connecting retail investors and institutions, used to hedge and price real-world uncertainties.
Finance is a highly vertically segmented world, with almost every sub-sector having its own universally recognized "mecca of the year." Leaders of healthcare providers, payers, and biotech companies gather annually in San Francisco for the JP Morgan Healthcare Conference. Global macroeconomic heavyweights and political leaders travel to the Swiss Alps for the World Economic Forum Annual Meeting (Davos). TMT, real estate, industry, financial services, and virtually every other sector you can think of also have their own flagship summits.
In late March of this year, Kalshi Research, Kalshi's academic and institutional research arm, held its first research conference in New York, bringing together academics, Wall Street executives, former politicians, and traders who are the driving force behind the market. The composition of the attendees clearly reveals a trend: the industry is "mature."
The conference opened with a dialogue between Kalshi co-founder Tarek Mansour and Luana Lopes Lara, and Katherine Doherty. Below are some industry observations gleaned from this dialogue and subsequent roundtable discussions:
Markets and life are more than just elections and sports.
During major news cycles, a recurring pattern emerges: a single major event (such as the 2024 election, the Super Bowl, or more recently, the March Madness college basketball tournament) dominates the majority of media headlines and consequently drives trading volume in prediction markets. This easily creates the impression that "the value of prediction markets lies solely in these events."
However, despite early narratives often portraying prediction markets as tools that "only make sense during election cycles," Kalshi's growth in other areas has been equally remarkable.
At the time of the research meeting, weekly trading volume in sports-related transactions was just approaching $3 billion, accounting for about 80% of Kalshi's total trading volume, primarily driven by "March Madness." Tarek and Luana viewed this high concentration as a phase-specific phenomenon.
A more telling statistic is that while the absolute size of sports-related transactions hit a record high, its share of total transaction volume remained at a record low. This means that all other categories grew at a faster rate.
The two founders pointed out that categories such as entertainment, crypto, politics, and culture are showing stronger user growth and a better transaction retention structure than sports. Sports is more like a "trigger" for the mass market—it has the characteristics of high familiarity, clear time rhythm, and strong emotional participation, making it a typical entry-level product.
At the same time, the company has also observed significant growth in the longer tail markets. These markets currently account for more than 20% of Kalshi's trading volume and will play an even more crucial role in the future institutional hedging and information markets.
A subsequent roundtable discussion among institutions confirmed this assessment from the demand side.
Cyril Goddeeris, co-head of global equities at Goldman Sachs, stated that forecasts related to macroeconomic events and CPI data are currently the most closely watched category on Wall Street. Sally Shin, executive vice president of growth at CNBC, mentioned that she already uses forecast markets such as those related to "the Fed Chair's future" and "non-farm payroll data" as content storytelling tools. Troy Dixon, co-head of global markets at Tradeweb, went further, painting a future picture: large investment banks will establish dedicated forecast market trading departments, with financial contracts as their core product.
Why did Kalshi attract the attention of Wall Street?
One important reason why traditional financial markets can function is that each type of core asset has a recognized benchmark: the S&P 500 index represents the overall performance of 500 stocks, while crude oil has benchmark pricing systems such as ICE.
However, for political and macroeconomic events (such as who wins an election, whether tariffs are passed, and the outcome of Supreme Court rulings), there has long been a lack of a widely accepted and dynamically updated "pricing benchmark." Prediction markets have changed this—now, the future of almost any event can have a real-time, fluid "price anchor."
Once an event (such as "whether the 30% tariff will pass") has a credible price, institutions can trade directly around that price. This can be used to trade the event itself or to hedge the risk of other assets in a portfolio. As Troy Dixon of Tradeweb put it, "Back when Trump was first elected, there was a lot of hedging in the stock market. The logic was to short the S&P 500 because if Trump was elected, the market would definitely fall. But that trade failed. The question is: how do you price these events? What's the benchmark?"
Tarek also mentioned that this was one of the motivations behind his founding of Kalshi. During his time at Goldman Sachs, his trading desk recommended trades based on the 2024 general election and Brexit. Without market prediction capabilities, when institutions hedge political or macroeconomic events with related assets, they are essentially betting on two things simultaneously: whether the event will occur, and the correlation between the event and the traded asset. The second judgment could very well be wrong on its own.
When the event itself has an immediate price benchmark, these two layers of risk are compressed into one. As Tarek put it, "Now, the market starts pricing everything."
The three stages in which institutions truly adopt prediction markets
It's clearly premature to say that major Wall Street institutions are already trading on Kalshi on a large scale. Currently, most institutions are still using it as a "data source" rather than a "trading platform."
However, Luana points out that the path for institutional adoption of this market is clear and can be divided into three stages:
The first stage is data integration: incorporating price predictions into the daily workflows of institutions. For example, Goldman Sachs portfolio managers are encouraged to routinely check Kalshi's odds data, just like they check the VIX index. This stage has already occurred to some extent. Johns Hopkins University professor and former Federal Reserve official Jonathan Wright stated, "In areas such as Fed decisions, unemployment rates, and GDP, Kalshi is almost the sole source of information."
The second phase is system integration: including compliance and legal approval, technology integration, and internal training—essentially a process of introducing new financial instruments.
The third stage is actual trading: institutions begin to hedge risks directly on the platform, and trading volume and market depth gradually accumulate. At this point, more hedging demand attracts speculators, and tighter spreads attract even more hedgers, creating a self-reinforcing positive feedback loop in the benchmark price.
Currently, most institutions are still in the first stage, some have entered the second stage, and very few have truly entered the third stage. A significant obstacle is that market prediction trading currently requires full margin. For example, a $100 position requires a $100 margin deposit. This is acceptable for individual investors, but for hedge funds or banks that rely on leverage and capital efficiency, this mechanism is too costly.
As Tarek stated, "If you want to hedge $100, you have to put $100 in the clearinghouse. That's too expensive for institutions. Institutions like Citadel or Millennium wouldn't do that." Kalshi has already obtained a license from the National Futures Association (NFA) and is working with the Commodity Futures Trading Commission (CFTC) to introduce a margin trading mechanism.
What will happen next?
Michael McDonough, head of market innovation at Bloomberg, summed it up most directly: "The sign of success is that these things become boring." He compared the forecasting market to the options market in the 1970s, which was also full of controversy over manipulation and regulatory uncertainty, but eventually evolved into an infrastructure that is now almost forgotten.
AQR partner Toby Moskowitz said he was "willing to bet real money" and predicted the market would become a viable institutional tool within five years, or even sooner.
Garrett Herren of Vote Hub described the final state as follows: "The question is no longer whether to use prediction markets, but how to use them. Once the question becomes like this, it means that they have become indispensable."
In fact, although the current size of the forecasting market is still limited, the hedging market itself is a huge sector.
In fact, the "normalization" of prediction markets is already underway.
During a political roundtable discussion, former Congressman Mondaire Jones noted that top figures from both parties—including President Trump, House Minority Leader Jeffries, and Senate Minority Leader Schumer—have begun publicly citing Kalshi's odds data. DDHQ's Scott Tranter also confirmed that prediction market data has now become a standard input within party committees. Meanwhile, Vote Hub announced that it has directly integrated Kalshi data into its midterm election prediction model.
All of this was completely foreign two years ago. Back then, the most successful traders on Kalshi were mostly "amateurs." Now, that term is no longer even accurate.
In Kalshi's "The People Behind the Markets" roundtable, four traders shared their career paths—paths that sound remarkably similar to those of traditional professional traders: one spent 11 years studying the Billboard music charts, while another honed their skills in prediction markets since 2006, when it was merely a "somewhat geeky hobby that barely made any money." Notably, none of these four guests came from traditional finance; instead, they hailed from the music, politics, and poker industries, respectively. However, they unanimously agreed that what this platform truly rewards is deep domain knowledge, not an impressive resume.
Prediction markets have come a long way. From being initially seen as an academic experiment, to becoming a "novel tool" during elections, and then being categorized as a "sports betting product," their positioning has constantly evolved. The clear signal from this conference is that prediction markets are evolving into an infrastructure—used to price uncertainty, serving a wide range of participants and diverse applications, from retail traders to large institutions.
