After Analyzing 203 Million Transactions, We Uncovered the Truth Behind Kalshi's Massive Profits

Author: Sam Schneider

Original Title: What’s Kalshi’s revenue?

Translation and compilation: BitpushNews

Imagine it’s 2005. You start a company called “Meth Labs, Inc.” You begin attracting customers and securing venture capital. Before you know it, your company is listed on the NYSE under the ticker $METH. People can buy and sell your stock, and even construct iron condor options strategies. The NYSE provides a centralized marketplace where buyers and sellers trade, with prices adjusting in real-time as information is disclosed.

I just mentioned the NYSE, but there are many other stock exchanges around the world—like NASDAQ, the London Stock Exchange, Shanghai Stock Exchange—that facilitate securities trading. In fact, markets are so vital to society that even if you’re not an active day trader, you’re constantly interacting with them… Uber connects drunk drivers and riders, Facebook Marketplace links people with secondhand furniture, and your dad is trying to leverage connections to get you a job…

Suppose you want to retire, sell your $METH stock holdings, and invest in other charitable causes. Who would buy these stocks? At what price?

• Markets define the prices people are willing to buy and sell at (price discovery).
• Markets provide a platform for trading assets because, in your neighborhood, maybe no one wants to buy your $METH (liquidity).
• But what if these markets didn’t show per-share prices, but rather the probabilities of certain discrete events? These are called prediction markets.

Whether through handwritten notes or through centralized, scaled, well-funded companies like Kalshi and Polymarket, prediction markets differ from stock trading:

• They are binary. Either the event happens or it doesn’t.
• They settle once a specific event occurs, a result is reached, or time expires.

You can’t buy $METH stock on a prediction market, but you can bet that the trading price of $METH on January 6 will be between $122 and $124.

In this article, we will examine the billions of dollars flowing through these markets, see how FTX’s “legacy” persists, understand how much sports betting accounts for, and find out how much Kalshi is actually making.

History / Introduction

Kalshi and Polymarket launched in 2018 and 2020, respectively. While these two form the current duopoly, the origins of prediction markets go back much further. One of the pioneers is the Iowa Electronic Market (IEM), which has been hosting prediction markets since 1988.

Betting on an individual’s “beliefs” can get complicated, but collective wisdom remains a valuable forecasting tool. Referencing a 2004 paper by Wolfer and Zitzewitz:

“These markets predict the vote shares of Democratic and Republican candidates with an average absolute error of about 1.5 percentage points… compared to Gallup polls, which have an error of 2.1 percentage points.”

However… over the years, something has been missing… something that hindered us from collecting predictions, building markets, scaling to millions, and leveraging them for personal gain. That missing piece is well-funded crypto web applications, which provide free resources enabling you to bet on whether the next pope is transgender. Just as Hayek envisioned in “The Use of Knowledge in Society.”

We will explore Kalshi in detail, but there are other projects and protocols in this space as well.

How does all this work?

Prediction markets expand the surface area of human gambling. Personally, I might bet $10 that I can drink 10 bottles of beer before midnight, and my wife might not believe me. On one side, I say “Yes, I can finish ten bottles,” and on the other, she says “No, you can’t.” Replace “I” with LeBron James, “beer” with points scored, “before midnight” with the end of the game, and you have a real market tradable on Kalshi.

On Kalshi, a market (Market) refers to a single binary event that settles as “Yes” or “No.” An event (Event) is a collection of such markets, and a series (Series) groups similar but independent events. For example:

• The high temperature in Miami is a series.
• Each day within that series is an event.
• Each event contains markets about temperature ranges: [68° to 69°] or [69° to 70°].
• Each market has a “Yes” and a “No,” along with relevant order books.

What is an order book? Let’s clarify some terminology first…

Every trade on Kalshi is a match between a maker (the order placer) and a taker (the order taker). Like in our hypothetical scenario that never happened, makers and takers are trading with each other, not with the platform itself. You’re not buying stock but a contract: if you’re right, it settles at $1; if wrong, it settles at $0.

As rational, price-sensitive actors, I am willing to risk $10 to win $20, and so is my wife. Since these bets are in place, the implied probability of the event is 50% (10/20). In “event contract” terminology:

• We split it into 20 contracts, each worth $1.
• Each contract is an agreement where both parties lock in $0.50.
• At settlement, the winner can keep the other’s $0.50.

As trading heats up, tracking activity via the order book becomes very helpful.

There are many ways to display an order book. They typically show resting orders, whether they are buy or sell, the quantity, and the time of order placement.

Bid prices (Bids) include all those seeking to buy; ask prices (Asks) include all seeking to sell. In the example above, bids and asks are sorted from best to worst price. The difference between the highest bid (Best Bid) and the lowest ask (Best Ask) is called the spread.

In a highly liquid market (like the Super Bowl), the spread can be tiny—perhaps just a penny. In less active markets, the spread can be large because no one cares enough to be a counterparty. That’s why Kalshi offers incentives for liquidity and market making.

Reviewing our order book, when a taker decides that $0.52 is a good deal and buys, they fill 50 contracts with my mom (you can see her on the buy side). The asset’s price just “moved” to $0.52, and that order disappears from the book. This is price discovery, or more precisely, the market allocating probabilities. Traders update the likelihood of me being liver failure-prone to 52%.

Looking at Kalshi’s real order book, things look different:

You will notice the “Yes” and “No” sides. On Kalshi, contracts can settle as “Yes” or “No,” and traders have demand on both sides.

• On the “Yes” side, there are 13 contracts priced as low as $0.44, and 58 available at $0.42, with a $0.02 spread.
• On the “No” side, 58 contracts are priced as low as $0.58, and 13 can be bought at $0.56, also with a $0.02 spread.

Wait… why are these complementary? Why do the “Yes” bid prices and “No” ask prices have the same number of contracts, and why does their sum equal exactly $1? Because buying “Yes” is equivalent to selling “No.”

Now that we understand order books, how does matching work?

Kalshi uses a price-time priority algorithm on its central limit order book to match orders. On the surface, it sounds simple: orders are sorted by price, then by submission time. But building an exchange that handles these orders at scale is no small feat. On Kalshi, orders must be fully collateralized, so margin trading isn’t currently available.

For a time, MIAXdx (via Miax) was the clearinghouse for Kalshi’s traded contracts (the central venue for trading). MIAXdx was formerly called LedgerX, but Miax, through… (here should be a drumroll)… was acquired and renamed during FTX’s bankruptcy proceedings!

After some time, Kalshi decided “We want our own clearinghouse,” and in August 2024, registered with the CFTC and established Kalshi Klear. To complete the loop, Robinhood recently bought back… yes, LedgerX from Miax!

Data

I accessed Kalshi’s market and trading APIs, obtaining historical data on approximately 30 million markets, 203 million trades, and over $41.7 billion in total trading volume.

Where does this volume come from? Kalshi’s website and app generate massive traffic. It also partners with some futures commission merchants (FCMs) that facilitate futures trading by accepting contracts on behalf of clients. You might recognize them:

For example, where you transfer your IRA… Robinhood;

Where teens trade… Webull;

And where you store your junk coins… Coinbase.

The top two event categories by volume are the 2024 presidential election (over $535 million) and the 2026 Super Bowl champion (about $244 million).

Wait… the Super Bowl… isn’t that… sports betting?

Isn’t this just sports betting?

Kalshi is regulated by the CFTC, which oversees the US derivatives markets. When I say “regulated,” I really mean “not regulated.” The Commodity Exchange Act (CEA) establishes the legal framework for the CFTC. It grants the CFTC the power to ban onion futures trading but also allows 18-year-olds to trade contracts on Kalshi. Kalshi even boasts on its FAQ page that the “minimum age to register and participate” is 18+, directly comparing itself to… sportsbooks.

Bookmakers are subject to state regulation; some states ban sports betting outright, others require bettors to be at least 18 or 21 (usually 21+).

Cool, but… Kalshi is different. You’re trading contracts with peers, and these contracts can be about anything. Not just sports, right? You’re not betting against a bookmaker, right?

Over 82% of contracts are about… sports. Kalshi is a volume-driven business; the more contracts traded, the more fees collected. As the first platform targeting 18+ bettors, this is ideal. By the way, they have also offered parlays, accounting for over 5% of total volume!

Another point about not betting against a bookmaker… from Kalshi’s article “Who am I trading with?”:

“Another important participant in the exchange is Kalshi Trading. Kalshi Trading is an entity independent of Kalshi Exchange… They, like everyone else, are participants in the exchange.”

If it looks like a bookmaker and trades like a bookmaker, then it might just be…

Back to the data!

The trading volume of these markets follows a power-law distribution. By bucketting the total volume (in USD) in orders of magnitude, this becomes clear.

80% of the volume is in markets with zero trading activity (i.e., multi-variable events, also known as parlays). Each is a unique market generated via Kalshi’s RFQ (Request for Quote) system, many of which have no counterparties.

Using the same volume buckets and splitting by settlement outcome, we see that as volume increases, the proportion of markets settling as “Yes” also increases.

This suggests that your baseline expectation for any particular market should lean toward “No.” This makes sense given the volume relationship: events with many markets dilute volume, and ultimately only one or two markets settle as “Yes.”

In these markets, the average number of contracts per trade hovers between 150-250, except for 2024, where some large orders of 1 million contracts due to the US election create a huge spike. The median is much lower, with most months below 50 contracts.

Fees: Bookmakers vs Kalshi

If sports bookmakers are like roulette, making money through odds, then Kalshi is more like poker, earning through rake, regardless of who wins or loses.

In sportsbooks, you’ll find “even odds,” where each side has a 50% chance, like flipping a coin at the Super Bowl. They won’t give you true even odds but will offer implied probabilities of 52.4% for each side, which is above reality.

In a fair market, if you bet $10 on a coin flip and guess correctly, you expect to win $10… but with a higher implied probability, say 52.4%, you only win $9.09 on a $10 bet.

In sports betting, if two people each bet $10 on a coin flip, one ends up with $19.09, the other loses everything, and the bookmaker pockets about $0.91, or 4.5% of the total bets. This 4.5% is the “rake”—also called “juice,” “commission,” or “house edge.”

For example, the market for Islanders vs. Devils might be set at 52.4%/52.4% by a bookmaker. On Kalshi, the contract price is $0.51 (51%). So, trading on Kalshi would seem better (51% vs. 52.4%), right?

No! Although the bookmaker’s odds are worse, Kalshi’s taker fee ($0.35) offsets this:

• Kalshi: Bet $10.55 (buy 20 contracts, total $10.20 + $0.35 fee), winning $20.
• Bookmaker: Bet $10.55, winning $20.14 (more profit on the same stake).

This isn’t the full picture, as Kalshi offers liquidity incentives and volume rewards. Plus, fee structures vary across markets, and Kalshi also pays you daily accrued returns on your positions.

Fees

How much does Kalshi earn from all this volume? First, let’s look at the taker fees. The formula is:

Fee = ceil(0.07 × C × P × (1 - P))

where C is the number of contracts, P is the price (ranging from $0.01 to $0.99).

We can plot fees as P and C vary. Fees increase linearly with contract count; controlled by P(1-P). As shown in the right chart, when implied probability P is far from 50%, fees are lower, since extremely likely or unlikely contracts cost less.

This is the Bernoulli distribution. It models the outcome of a single “Yes” or “No” question (our market). Its variance is modeled as P(1-P), which is the Y-axis in the right chart.

• Variance ranges from 0 to 0.25.
• Entropy (uncertainty or randomness in the probability distribution) peaks at P=0.5.

Why doesn’t Kalshi implement a fixed fee rate to avoid plotting? Likely for trading considerations:

If a contract is priced at 98 cents, your maximum profit is 2 cents. If Kalshi charges a flat 2-cent fee, your profit drops to zero… nobody will trade it.

The slope of the taker fee is the same, just scaled by a factor of 0.0175 (a quarter of the taker fee):

Fee = ceil(0.0175 × C × P × (1 - P))

After analyzing all 203 million historical trades on Kalshi, I know the exact trading price per contract. Plugging P and C into these formulas, I estimate Kalshi’s total revenue from all these contracts at approximately $546 million.

Below are Kalshi’s monthly trading volumes categorized by implied probability:

And here are Kalshi’s monthly fee revenues:

Clearly, these markets are exploding in volume, with DraftKings, FanDuel, and Fanatics rushing into the frenzy—a “we do basically the same business, but now in a pseudo-regulated environment where 18+ can play” party.

Settlement

An interesting settlement case is when the Dallas and Green Bay teams tie in an NFL game. The market settles at 50/50, not “Yes” or “No,” or 100 or 0… In prediction markets, there’s no concept of refunds (pushing) or voiding. When the outcome is uncertain, Kalshi intervenes. In the data, Kalshi marks these as “scalar” markets, with over 170,000 markets labeled as scalar.

Market settlement on Kalshi appears quite manual. They have a dedicated team to thoroughly review outcomes. Each market has an authoritative reference point. For example, the Super Bowl markets list multiple sources and include specific rules. Still, they can’t determine whether Cardi B performed at the halftime show and settle based on the last traded price.

Polymarket, on the other hand, says “Yes, she performed,” highlighting their use of UMA’s optimistic oracle for different settlement mechanisms… but we’ll save that for another time.

Conclusion

At this point, I have 9 bottles of beer to drink, and this article has already gone way over the word limit.

A quick legal aside: there’s also a prediction market called PredictIT, focused on political forecasts. Operated by a company called Aristotle, it provided data mining services for political campaigns. Launched in 2014 as a nonprofit educational project at Victoria University of Wellington, New Zealand, it obtained a no-action letter from the CFTC, allowing it to operate under certain protections for academic purposes. By 2022, PredictIT was heavily penalized by the CFTC for non-compliance. By 2025, they fought back in federal court, and the “Cadillac of prediction markets” is back!

In summary, all these companies involved in “event contracts” keep submitting letters to the CFTC, asking not to be penalized for not meeting typical reporting standards. So far, the CFTC seems to agree, citing that “the applicability of traditional swap reporting rules to exchange-traded event contracts is limited.”

Of course, other issues remain, such as classification questions around Kalshi and Robinhood, but future discussions will revolve around how to regulate, tax, and report on these “new” entities.