Kalshi’s $40 billion question

A $4 billion speedometer reading

Sacra puts Kalshi’s July 2026 revenue pace at $4 billion a year. Treat that as a speedometer reading, not an odometer. And don’t take the volume at face value: about $7 billion of June’s activity came from perpetual futures that launched with no fees. 3

The $40 billion price is ten times that annualized revenue pace. Seasonal trading, rebates, and partner payments make the comparison with a mature exchange’s net revenue uncertain. The missing monthly bridge is what customers paid, what funded liquidity and distribution, and what remained for Kalshi.

I count event turnover at $1 maximum payout per matched contract, once per trade. Perpetuals use dollars of exposure traded. Economic net revenue is the fee left after transaction-linked concessions and distribution payments, before Kalshi’s operating costs. I use those denominators to keep promotional volume from masquerading as customer spending.

One round trip costs $3.50

Buy 100 contracts at 50 cents and sell them at the same price, taking liquidity twice at the default multiplier. The price gain is zero. The fee bill is $3.50. I count $200 of turnover across the two executions, funded by recycled collateral instead of $200 of new deposits.

Kalshi’s July default taker fee is 0.07 × contracts × price × (1 − price), subject to market multipliers, maker exceptions, and rounding in its fee schedule. The price mix changes the fee even at the same contract count.

Spread, rebates, and special schedules can change the trader’s total cost. The bid–ask spread pays the liquidity supplier, which may be an outside market maker. I value the fees Kalshi keeps. Repeat customers must be willing to replenish their accounts after fees, trading losses, and withdrawals if that fee stream is to last.

Scroll horizontally to inspect the figure →

Liquidity has to be assembled for each event

Kalshi’s product is an executable price, enough size to trade, a clear settlement rule, and confidence that the contract will pay. A good experience can bring more orders and improve the next customer’s execution.

But depth in an election contract doesn’t fill tomorrow’s basketball book. I value repeated success at assembling markets. Messari’s April study found in-game order-book depth falling 76% in its historical sample, a warning that busy trading can coexist with thin execution. 5

Kalshi’s filed liquidity-incentive terms pay for qualifying resting orders. I would pay for the advantage when the depth and customer demand produced by those payments justify their cost. I count the expense once, either as a fee concession or an operating cost.

CME’s second-quarter release reported $95 billion of average daily margin efficiencies. Portfolio offsets give its customers a reason to stay that a competing screen can’t easily copy. Kalshi has to build comparable attachment through its own products and capital relationships. 14

The broker can charge for access and choose the venue

Robinhood’s June announcement sends selected World Cup and baseball contracts to Rothera, while its disclosures also name KalshiEX and ForecastEx. Robinhood describes its own commission plus an exchange fee. Both broker and venue can get paid. 6

A lower-yield broker trade can contribute more than a direct trade when the broker’s funded customer arrives with enough acquisition and service costs already paid to outweigh the venue’s concessions, provided the order adds business instead of simply moving an existing customer to a cheaper route. I’d follow contribution per route. The planned Alpaca integration points toward that opportunity, with realized volume still unreported. 7

Rothera also gives the broker an alternative venue. The ability to send the next order elsewhere can affect the price of the orders Kalshi keeps. I would compare contribution after the full bill for each route.

May’s perpetual-futures launch creates another growth route. Exposure traded in those markets earns a few modeled basis points, far below binary-contract yields. 8 Data may be more recurring, but ICE’s February Polymarket signals launch shows that institutional distribution is already contested. A paying buyer needs coverage, history, and integration worth renewing. 9

AI can improve the market and compress the fee

Automated participants can improve prices and route cheaply between venues. That may encourage trading while making the customer more sensitive to fees. Kalshi keeps bargaining power when reliable settlement and executable depth outweigh a cheaper alternative.

Within a closed trading group, gains and losses cancel before costs; fees leave the group. Fresh funding, a hedge, or entertainment value has to sustain the willingness to trade. Professional participants can improve prices, but more turnover doesn’t create the customer budget that pays the venue.

I would judge automation by repeat funded demand and contribution after liquidity incentives. Fills and available size around an important event reveal more than a growing order counter.

Sports access belongs in the cash flows

On October 9, five days before the valuation cutoff, the CFTC proposed expressly including event contracts based on sports, politics, culture, and weather in the definition of a swap. The proposal leaves state challenges and final rulemaking unresolved. 10

The agency describes favorable Third Circuit reasoning in April, opposing Ninth Circuit reasoning in August, and an adverse Sixth Circuit ruling in September. It is defending its jurisdiction in that text. 11 Broader federal access would help Kalshi, while surveillance, disputes, and competition would still cost money. The announced Nasdaq surveillance integration is one such investment. 12

A loss of sports access would cut customers and turnover before the cost base could shrink. My restricted case therefore contracts revenue and runs early operating losses. The base case allows a substantial sports business with compliance costs, and the breakout case combines broader access with strong execution. Legal outcomes enter those cash flows once.

What I am underwriting

I project $8.85 billion of economic net revenue in 2031, up from $4.83 billion in 2027. The $4 billion starting normalization uses the reported pace as a scale anchor. The table contains the author’s operating assumptions.

I put most of the growth in sports. Annual turnover doubles from $300 billion to $600 billion while the fee Kalshi keeps falls from 1.35% to 1.10%. Other events contribute at lower yields. Even $2 trillion of perpetuals exposure produces only $500 million at 2.5 basis points.

Keeping 1.10% across that sports volume requires a favorable contract mix and enough fees left after incentives and negotiated terms. A broker may charge its own commission, but its ability to switch venues still shapes the bargain.

I translate the model’s $6.6 billion of sports revenue into $220 per customer at 30 million annual fee-paying customers, or $660 at ten million. These illustrative customer counts show how much fee-paying demand the forecast needs, before traders’ losses and payments elsewhere.

Of $450 million in ancillary revenue, an illustrative $300 million comes from data and services and $150 million from eligible customer balances. The latter needs $5 billion of average balances earning a 3% net yield for Kalshi. The principal belongs to customers and never enters equity value.

Scroll horizontally to inspect the figure →

A 50% margin still leaves a reinvestment bill

I let operating margin rise from 42% to 50%, including stock compensation. That is ambitious: ICE’s second-quarter 2026 Exchanges segment earned a 74% GAAP margin, while its whole company earned 52%. ICE supplies the comparison; Kalshi’s margins are author assumptions. 13

Kalshi must earn its margin after development, marketing, surveillance, support, and liquidity costs. Labeling a fee “net” doesn’t make the incentive bill disappear.

I turn operating profit into cash after taxes and growth investment. A 50% operating margin still has to fund the capital needed for new revenue.

I use the downloadable model for the tax, capital, growth-fade, and dated-discounting conventions, including the partial 2026 period. The assumptions file states every input. Losses in the restricted case receive no immediate tax refund or speculative later tax benefit.

Scroll horizontally to inspect the figure →

Before applying a multiple, the Hyperliquid valuation model follows exchange activity through protocol receipts to the economic claim being valued.

Three businesses, three values

I value the restricted case at $2.6 billion of enterprise value. It reaches $2.04 billion of 2031 net revenue, with $150 billion of sports turnover at a 0.80% yield and margin recovering from negative 30% in 2027 to 30% in 2031. Worse legal, liquidity, or funding outcomes could destroy more value.

The base case supports $26.4 billion. Its $8.85 billion of 2031 revenue and $4.42 billion operating profit already describe a successful exchange.

The breakout supports $58.7 billion. It needs $1 trillion of sports turnover at 1.25%, $400 billion of other events, $4 trillion of perpetuals, and $900 million of ancillary revenue. Together those produce $18 billion of 2031 revenue at a 58% margin. Broad access, direct demand, distribution, and fee resilience all have to work.

My 25% restricted, 50% base, and 25% breakout weights produce $28.5 billion of enterprise value. Adding $1 billion of assumed unrestricted net cash gives $29.5 billion of aggregate equity. With no excess net cash it is $28.5 billion, and with $2 billion it is $30.5 billion.

A financing headline can’t supply the cash bridge: money may have been spent, earmarked, or paid to selling shareholders. A per-share price also needs the capitalization table, options, and each class’s rights.

Scroll horizontally to inspect the figure →

What $40 billion requires

At base margins and investment rules, about 50% more revenue in every forecast year is needed to support $40 billion of equity with $1 billion of net cash. That is about $7.22 billion in 2027 and $13.25 billion in 2031, against the base’s $4.83 billion and $8.85 billion. Reinvestment rises with the larger business.

Arriving at $13.25 billion only in the final year wouldn’t buy the same value. The years of cash flow on the way there matter, as do the margins and capital needed to earn them.

Fees can move the answer by billions. A 25-basis-point cut in sports yield during 2027–2031, carried into later years, takes base equity from $27.4 billion to $23.2 billion. This allows costs to fall proportionally with revenue. Sticky costs would make the loss larger.

At an 11% required return and 3% perpetual growth, base equity reaches $35.0 billion. At 15% it falls to $22.4 billion. Even the table’s favorable 11%/4% pair remains below $40 billion.

Holding the restricted weight at 25%, a $40 billion weighted equity value needs a 57.4% breakout weight, leaving 17.6% for base. That is a different investment conviction, far beyond a small adjustment to a multiple.

The evidence that would change my price

I would move toward $40 billion with a reconciled history supporting the reverse model’s approximately $7.2 billion of 2027 economic net revenue at base-like margins and investment. It should show what customers paid, what funded liquidity, and what remained after each distribution route.

Cheaper broker distribution would strengthen that case if it added enough profitable demand. Lower fees alongside continuing incentives and dependence on a few routing partners would weaken it. Repeat direct customers, paid data, and non-sports contribution could make the franchise less dependent on one legal and demand regime.

My price stays near $30 billion

The July fee schedule, Robinhood’s broker and exchange charges, and Kalshi’s filed liquidity-incentive terms explain how trading revenue is divided. The income statement, broker agreements, and capitalization table are undisclosed. The scenarios and legal cutoff are dated October 14, 2026.

Contrary’s regulation-first history and Messari’s execution-quality analysis provide historical context, followed by company releases and the October CFTC proposal. Contrary’s profile dates to 2023; the PitchBook reference is its public abstract, rather than the paid report or model. 17

The valuation method follows Damodaran’s distinction between activity, revenue, operating conversion, and reinvestment. 15 The model and figures use the declared authored inputs and can be reproduced without paid credentials.

I value the aggregate equity at about $30 billion under the declared cash and security assumptions. At the $40 billion price discussed in financing talks, I would pass until collected contribution and the security terms make that bridge credible.