Have you ever watched a market print the same size over and over and wondered whether you were looking at genuine demand or a machine doing laps? That question is hanging over prediction markets this week, and it is not a small one. A regulated exchange just told the futures regulator it wants to shut a long-running volume reward scheme, while more than five billion dollars in nearly identical Ether perpetual trades sit under official review. September volume is already a record. Fundraising talk is circling a valuation that would have sounded cartoonish two years ago. Those facts can live in the same sentence, and that is what makes this moment awkward.
Why Kalshi Is Closing Its Volume Incentive Program Now
On September 28 the exchange filed to terminate its Volume Incentive Program, with the earliest possible end date of October 13. The filing does not spell out a motive. It does not mention Ether. It does not mention wash trading. It simply says the program is going away. In my experience, silence in a regulator packet is rarely accidental. Firms choose that tone when they want the paper trail to look tidy.
The program itself is older than the current controversy. It was first filed in February 2023 and was meant to lift activity on the central limit order book. Eligible traders shared a fixed reward pool based on their slice of qualifying completed volume. Most event contracts had to trade between three cents and ninety-seven cents to count. That price band did not apply to perpetual futures. Reward periods could last no more than 31 days. Event-contract payouts were capped at half a cent per contract per participant. The chief regulatory officer could kick someone off if the activity looked abusive or off-purpose.
That last clause matters. Incentive programs always create a feedback loop. Pay people for volume and you will get more volume. Some of it will be useful. Some of it will be theater. Distinguishing the two is the whole job of market surveillance, and it is never as clean as a slide deck suggests.
What The Filing Actually Changes
Ending one named program is not the same as ending every incentive on the platform. A separate Deposit and Trading Reward Incentive Program was filed on September 25 and placed under a ten-day review. So the headline is narrower than social posts made it sound. The exchange is not walking away from every carrot it ever designed. It is retiring a specific volume-share scheme that dates back to the early event-contract years.
A quiet filing can look like housekeeping until you line it up with the week’s headlines.
I’ve found that traders rarely read filings the way lawyers do. Traders read timing. The motion arrived after weeks of public argument about repeated Ether prints, after a trader highlighted a huge gap between 24-hour perpetual volume and open interest, and after a national business desk described a regulatory look at nearly identical tickets clustered around a five-thousand-five-hundred-dollar size. You can believe those facts are unrelated. You can also believe coincidences this neat are rare.
September Volume Did Not Slow Down
Through September 29 the venue had already recorded about 52.98 billion dollars in trading. August finished at 38.67 billion. July contributed roughly 37.7 billion of a 50.6 billion combined figure across the main prediction-market platforms. That is not a sleepy tape. Whatever you think of the quality of those prints, the raw number kept climbing while the argument about quality got louder.
| Period | Reported Activity | Context |
| July | About 37.7 billion on this venue | Large share of a 50.6 billion three-platform total |
| August | 38.67 billion | New monthly high at the time |
| September through the 29th | 52.98 billion | Already above the full August total |
Record months make incentive design more sensitive, not less. When a book is thin, extra volume can look like a public service. When a book is already roaring, extra volume invites a different question: who is paying whom to keep the scoreboard moving?
The Ether Prints That Sparked The Fight
The first public flare came from a trader who pointed to roughly 538.6 million dollars in 24-hour Ether perpetual volume against about 3.1 million dollars of open interest. That ratio is the kind of stat that makes risk desks sit up. Open interest is the stock of risk still sitting on the book. Volume is the flow. When flow dwarfs stock by that much, either the market is turning over at a ferocious clip or a lot of tickets are crossing without leaving much net exposure behind.
Later reporting described a review of trading patterns built from nearly identical Ether perpetual transactions, many close to 5,500 dollars each. Over about a month those repeated sizes added up to more than five billion dollars of notional. The review was described as a look, not a formal enforcement case. That distinction is real. It is also not the same as a clean bill of health.
The exchange pushed back hard. In a September 22 statement it said the identified tickets involved traders who genuinely wanted opposite sides, with hundreds of takers hitting a market maker that posted resting quotes. Self-matching, it said, is mechanically blocked. Coordinated wash trading is prohibited and monitored. The company wrote, in plain language, that wash trading does not occur on its platform. Repeated sizes, it argued, came from the way market-maker arrangements are structured and from traders who keep lifting the same quoted size.
That explanation is plausible. It is also the explanation you would expect if the prints were real. Both things can be true at once, which is why this story refuses to sit still.
Volume Rewards Are Not The Same As Liquidity Pay
This is the part a lot of commentary flattened. Perpetual market-maker programs generally pay for resting quotes of a stated size inside a stated spread for a required stretch of time. That is a payment for presence, not a payment for turnover. If a quote sits there and nobody hits it, the maker can still earn. If a quote gets hit a thousand times, the maker may earn the same presence fee plus ordinary trading economics.
The Volume Incentive Program paid for completed qualifying activity. Those are different machines. Shutting the volume machine does not, by itself, prove the liquidity machine is gone. The September 28 filing names one program. It does not say the perpetual quote arrangements cited in the Ether response have been cancelled. Readers who treat the filing as a confession are jumping a fence the document never built.
- Volume rewards pay for share of completed qualifying trades.
- Liquidity rewards pay for two-sided quotes that stay inside a spread.
- Event contracts often needed mid-range prices to count.
- Perpetual futures sat outside that mid-range rule.
- Surveillance can still eject a participant if activity looks abusive.
Perhaps the most interesting aspect is how easy it is to confuse those two ledgers if you only watch the volume ticker. A healthy book needs both rest and flow. Pay only for flow and you get churn. Pay only for rest and you can get pretty quotes that never trade. Good design lives in the messy middle.
How Incentive Design Can Distort A Tape
I’ve sat through enough market-structure arguments to know the pattern. An exchange wants depth. Depth looks like size on the screen. Size on the screen looks like legitimacy. Legitimacy attracts more users and more press. Somewhere in that loop a rebate appears. Rebates work. They also teach participants to optimize the metric you pay for, not the metric you wish you had paid for.
If you pay for notional, you get notional. If the contract is small and the tick is friendly, the cheapest way to farm notional is to cross the same clip again and again. That can still be two real counterparties. A maker posts. A taker lifts. Risk transfers, then transfers back. Economically it can be close to a round trip. Legally it may not be a wash if the accounts are distinct and intent is not collusive. Optically it still looks like a photocopier.
That optical problem is now a business problem. Prediction markets sell trust as much as they sell contracts. People bet on elections, weather, sports, and now a growing list of crypto perpetuals because they believe the displayed probability or the displayed price is a signal, not a prop. If the signal is padded, the product gets cheaper in the only way that matters: reputation.
Perpetuals Changed The Texture Of The Book
Event contracts settle when the world answers a question. Perpetual futures do not wait for an election night. They roll, fund, and trade like a continuous crypto product that happens to sit inside a U.S. event-contract shop. That mix is new enough that old incentive language does not always fit. The original volume program was born in a world of binary-ish event tickets. Perpetuals brought tighter quotes, faster takers, and a culture that already knew how to farm volume on offshore books.
This year the lineup expanded quickly. In September the venue added contracts on BNB, Cardano, Worldcoin, Aave, and Venice Token, taking the crypto perpetual set to Bitcoin plus 17 altcoins at the time. Earlier, perpetual volume had crossed 5.5 billion dollars within two weeks of launch. That is a sprint, not a stroll. Sprints produce both genuine product-market fit and sloppy edges. You do not get one without risking the other.
In my view, the expansion was inevitable. Crypto traders already live on perpetual rails. If a regulated U.S. venue wants that flow, it has to list the instruments those traders actually use. The cost of that choice is surveillance complexity. Repeated 5,500-dollar clips are easier to defend as market-maker structure when the product is new. They get harder to defend after the fifth billion.
What Wash Trading Means And What It Does Not
Wash trading, in the ordinary sense, is trading with yourself or a coordinated partner to create a false appearance of activity. Intent matters. Account control matters. The economic residue matters. A market maker posting a quote that a hundred unrelated takers lift is not, on its face, a wash. A pair of accounts that pre-arrange to hit each other at the same size all afternoon is a different animal.
The exchange says the first story is the true one. Critics say the tape looks like the second. I do not have the order-level data sitting on my desk, and neither do most readers. What we have is a pattern, a denial, a regulator looking, and a program being retired. That is enough to write about. It is not enough to convict in a comment thread.
Identical ticket sizes can be a market-maker footprint or a warning light. The difference is in the accounts and the intent, not in the screenshot.
There is a practical test regular users can apply without becoming amateur prosecutors. Look at open interest next to volume. Look at whether prices move when the big clips print. Look at whether spreads stay honest when the rebate calendar changes. If volume collapses the day an incentive dies and prices barely twitch, you learned something. If volume stays bid and spreads stay tight, you learned something else.
Regulation Sits In The Middle Of The Room
The Commodity Futures Trading Commission is the venue’s primary federal overseer for these products. A review is not a charge sheet. Firms often hear nothing formal while staff request data, replay tapes, and ask quiet questions. The exchange has said it was not contacted about the reported examination. That can be true and still leave staff reading public research on their own. Agencies do that.
Prediction markets already live in a legal weather system that changes by state and by product type. Event contracts on elections have spent years in courtrooms. Crypto perpetuals add another layer because they look and feel like the offshore products that U.S. agencies spent a decade criticizing. If the volume story gets sloppy, opponents of the category will use it. That is not fair in every case. It is predictable.
I’ve found that the smartest compliance teams treat optics as a first-class risk. They do not wait for a subpoena to retire a program that has become a magnet for screenshots. Ending a rebate before a hearing is cheaper than defending the rebate inside one.
Valuation Talk Does Not Pause For Awkward Weeks
While the tape argument ran, people familiar with financing talks described advanced negotiations on a raise of about one billion dollars at a valuation near 40 billion. Sequoia and Wellington were discussed as possible leads, with Tiger Global and Dragoneer among names that could join. A completed deal at that level would follow a May 2026 round that valued the company around 22 billion. Talks can still change. They often do.
ARK funds have also taken direct marks in the private company across several of the flagship portfolios. The same shop has publicly sketched a world where prediction markets someday turn over one to five trillion dollars a year. Those are long-horizon numbers. They assume the category keeps its license to operate and its claim to honest prices. A messy volume story is not fatal to that thesis. It is a speed bump that investors will price if it lingers.
Is a 40 billion dollar private mark compatible with a live debate about printed size? Markets answer that every cycle. Growth stories absorb controversy until they cannot. The honest read is that demand for the equity and demand for the contracts are both real, and that both can be inflated by narrative. Separate the two in your head or you will misread both.
What Traders Should Watch After October 13
If the volume program actually sunsets on or after that date, the first useful experiment begins. Watch Ether perpetual turnover versus open interest. Watch whether the famous clip size fades. Watch whether event-contract volume in the three-to-ninety-seven-cent band softens more than the perpetual book. That split will tell you which activity was subsidy-sensitive.
- Compare daily perpetual volume to open interest for two weeks after the cutoff.
- Note whether repeated mid-size prints persist when the rebate is gone.
- Check spreads and quote depth during overlapping U.S. hours.
- Separate event-contract flow from crypto perpetual flow instead of using one headline number.
- Treat any remaining incentive filings as a new rulebook, not a footnote.
Do not expect a single clean drop. Market makers can keep quoting because presence fees still pay. Fast takers can keep lifting because the trade is still a trade. The interesting change is at the margin: the last few billion that only existed because someone was farming a share of a pool.
A Fair Reading Of Both Sides
Critics are not crazy to stare at identical sizes and a volume-to-open-interest gap. Those are classic tells. Defenders are not crazy to say a quoted size that gets hit all day will print the same number all day. Those are classic market-maker footprints. The grown-up position is to hold both ideas without turning either into a personality test.
The exchange has rules against self-matching and coordinated wash activity. Rules are not the same as outcomes, but they are not decoration either. A regulated book with a named chief regulatory officer and a published eject clause is still a different animal from an anonymous offshore venue with a cartoon mascot. Collapsing those two worlds into one rant helps nobody who actually trades.
At the same time, “we have a program for that” is not an infinite shield. If a reward pool pays for completed volume, someone will complete volume. If perpetuals sit outside the mid-price filter that constrained event tickets, someone will notice. Design choices have children. Those children show up on the tape.
Why Prediction Market Integrity Is The Real Product
People do not wake up craving another derivative ticker. They wake up craving a number they can argue about with a straight face. Election odds. Rate-cut probabilities. The chance a token holds a level. That number is only useful if the path that produced it was not a carnival ride funded by a rebate calendar.
I keep coming back to a simple analogy. A restaurant can fill seats with a two-for-one coupon. The dining room looks busy. The kitchen may even be good. But if every full house depends on the coupon, you do not yet know whether people like the food. Prediction venues are in that kitchen-versus-coupon phase for parts of the crypto book. September’s 52.98 billion dollars says the room is packed. The Ether clip story asks whether some of the guests were comped.
None of that requires you to pick a villain. It requires you to keep the scoreboard honest. Volume is a vanity metric until it is paired with open interest, price impact, and post-incentive persistence. Use the full set or you are reading marketing.
The Broader Category Is Still Growing Up
Combined prediction-market activity has been setting records for months. That growth is not a mirage. Political contracts, sports questions, and macro events pulled in users who never wanted a perpetual in their life. Crypto listings then pulled in a second crowd that lives on leverage and twenty-four-hour screens. Housing both crowds on one regulated platform is ambitious. Ambition leaks. Sometimes it leaks as a filing that kills a rebate.
State-level fights over whether event contracts are gambling will continue in parallel. Those cases do not need a volume scandal to stay alive. A volume scandal would still be useful to people who want the category boxed in. That is why the October cutoff is more than housekeeping. It is reputational first aid, whether or not anyone admits that in a cover letter.
Will ending the program settle the argument? Probably not on day one. Arguments like this die when the tape changes or when a regulator speaks in complete sentences. Until then, the adult move is to read the filing as written, keep the denial on the record, and watch the data after mid-October like it actually matters. Because it does.
Practical Takeaways If You Trade These Books
If you are a casual event-contract user, this fight may never touch your ticket. A contract on a confirmation hearing or a weather threshold does not need a 5,500-dollar crypto clip to function. If you trade the new perpetual list, treat headline volume as a starting point, not a finish line. Size your risk off open interest and realized impact, not off a 24-hour banner.
If you make markets, assume screenshots of your default size will circulate. Uniform clips are efficient. They are also photogenic in the worst way. Varying displayed size is not a moral duty. It is hygiene in a week when hygiene is the product.
If you allocate to the private equity story, separate operating momentum from metric quality. A company can be valuable and still need to retire a clumsy rebate. Those two sentences can share a paragraph without canceling each other.
Read the tape in layers: Headline volume Open interest Repeat-size clustering Spread stability after incentives change
That stack is not glamorous. It is how you avoid getting fooled by a busy board. I would rather be slightly late and clear than early and impressed by a number that only existed because a pool was paying for it.
Where This Leaves The Next Few Weeks
The earliest termination date is October 13. A different incentive filing is already in a short review window. Perpetual listings are still expanding. September is already larger than August with a day still on the calendar when the snapshot was taken. Fundraising conversations are live. A federal look at repeated Ether tickets has been described, not confirmed as a full enforcement matter. Hold all of those facts at once and the picture is less a scandal cartoon than a growth company hitting the first real stress test of its crypto chapter.
Stress tests are useful. They show which volume was rented and which volume was owned. They show whether a denial holds up when the rebate ends. They show whether a 40 billion dollar conversation can survive a month of ugly charts. I do not know the ending. I do know the next useful chapter is not another viral thread. It is the first quiet Tuesday after the program is gone, when nobody is performing for a reward pool and the book has to speak for itself.
That Tuesday is the test. Everything before it is noise with a filing number attached.
