In traditional finance, “price” typically belongs only to assets. Stocks, interest rates, commodities—these are tradable because they have standardized measurement methods and consensus-based pricing mechanisms. In contrast, the variables that truly drive market volatility—policy direction, macroeconomic data, political events—have long existed in a more primitive state: they are discussed and forecasted, but rarely priced directly. These variables have always existed, yet lacked standardized expression. Kalshi’s emergence fundamentally changes this. It does not create new information; rather, it provides a tradable pricing system for “events themselves.”

At a recent research conference, a notable data point emerged: weekly trading volume in sports-related markets has approached $3 billion—but its share of total trading volume is declining. In other words, the most visible segment is growing, while deeper structural shifts are underway. Meanwhile, institutions—including a16z—are beginning to consistently monitor this space. This isn’t because prediction markets are simply “getting hotter,” but because they are increasingly exhibiting characteristics of infrastructure. Prediction markets are evolving from a niche product into an infrastructure for “pricing uncertainty.”

01 Wall Street’s Attention: From “Discussable” to “Pricable”

Financial markets operate on one foundational premise: there must exist a benchmark price that can be traded. The S&P 500 anchors equity markets; the yield curve defines funding costs; commodity futures provide forward expectations for supply and demand. Yet in many critical decisions, the variables that truly determine outcomes lie outside these conventional assets—especially “event-type variables,” which have long lacked standardized pricing mechanisms. Examples include: whether a given policy will be implemented, whether inflation data will exceed expectations, or whether regulatory changes will occur. These factors influence markets—but cannot be traded directly.

The traditional solution has been indirect expression via “correlated assets” (e.g., hedging election risk using stock indices). The problem lies in two embedded layers of risk assumptions: first, the event itself is uncertain; second, the relationship between the event and the asset may shift. The latter layer is often far less controllable. The core significance of prediction markets lies in eliminating this structural bias—by making the “event itself” a tradable object. When the market prices “the probability that a certain policy passes” at 40%, that number ceases to be merely an opinion—it becomes a variable that can be actively traded, hedged, and modeled.

02 A Misunderstood Starting Point: Why “Sports” Isn’t the Focus—It’s Just the Entry Point

Prediction markets first scaled significantly through sports and elections—a natural outcome: clear event boundaries, discrete outcomes, and low user entry barriers. Such scenarios are inherently suited to early-market launch, yet they also created a misconception: people mistook the “most visible demand” for the “entire demand.” But Kalshi’s disclosed data reveals a structural reversal: weekly sports trading volume nears $3 billion yet its share is shrinking; macro/policy categories are accelerating in growth and attracting institutional attention; entertainment/crypto/culture segments show faster user growth and higher retention.

This highlights a crucial insight: high-traffic scenarios do not equate to high-value scenarios. Sports function more like a “cold-start mechanism,” providing users and liquidity—but the variables with genuine financial utility are those institutions can use for hedging and pricing. At the conference, participants from Goldman Sachs and Tradeweb both noted that macro events (e.g., CPI releases, interest rate paths) are becoming the most compelling category in prediction markets. These variables share a defining trait: they are not assets themselves—but they determine asset prices.

03 The Real Path to Institutional Adoption: From “Reference Indicator” to “Trading Tool”

Despite rising discussion热度, prediction markets remain in the early stages of institutionalization. According to Kalshi’s framework, institutional adoption unfolds across three phases: the Data Phase (using prediction prices as reference signals), the Integration Phase (embedding them into models, risk management, and research systems), and the Trading Phase (direct risk hedging and position allocation).

Currently, most institutions remain in the first two phases. A key constraint stems from the trading structure itself: today’s prediction markets require 100% margin to establish a position. For institutions reliant on leverage and capital efficiency, this entails significant opportunity cost. That’s why Kalshi is working with the CFTC to introduce margin mechanisms. Once this constraint is lifted, growth at the trading layer could undergo structural change.

04 From Asset Pricing to “Probability Pricing”: An Extension of the Financial System

Viewed across the broader arc of financial history, prediction markets are not an isolated innovation—they represent an expansion of the pricing architecture. Traditional markets price: assets, cash flows, and risk premia; prediction markets price: events, probabilities, and expected pathways. The key distinction lies in orientation: the former is outcome-oriented; the latter is process-oriented.

This shift brings an important transformation: information begins expressing itself as “price,” rather than remaining confined to analysis and narrative. For example, when the market assigns a “60% probability” to a given policy passing, that figure can be embedded directly into quantitative models, used for risk hedging, or serve as a decision input—bringing it far closer to how financial systems actually operate than traditional expert judgment or polling data.

05 Intersection with Agents / AI: From “Prediction Tool” to “Decision Input Layer”

Another dimension of significance lies in prediction markets’ potential integration with AI systems. Most current Agents face a shared challenge: they can generate conclusions but struggle to quantify uncertainty. Prediction markets offer an alternative path: using real capital to constrain predictions, leveraging market mechanisms to aggregate information, and expressing probability as price. As Agents begin participating in financial decision-making, risk management, or strategy generation, such “probability prices” will become critical inputs.

06 The End State Is Not Complicated: Becoming “Default Infrastructure”

At the conference, one observation was repeated: “It’s only truly successful when it becomes boring.” This isn’t dismissive—it reflects the typical trajectory of financial infrastructure. Options markets were highly controversial in the 1970s; ETFs were initially viewed as fringe instruments. But once they became standard components, they ceased being debated. Prediction markets may now be entering a similar phase: progressing from academic experiments to election/sports tools, then to macro/ institutional applications—and ultimately becoming a “default-existing” pricing layer. At that point, they won’t be called “prediction markets” anymore—they’ll simply be part of the financial system.

07 When “Uncertainty” Enters the Price System

Returning to the original question, the core of this shift lies not in trading volume or user scale—but in a more fundamental transformation: uncertainty is now being standardized and expressed. When events can be priced and probabilities traded, the future ceases to be merely an object of discussion—it becomes a variable that can be computed and allocated. In this process, prediction markets are not just a new product; they constitute a new financial language. Once widely adopted, this language reshapes not only how we trade—but the very architecture of our decision-making systems.