Kalshi Wash Trading Claims Over Crypto Perp Volume

Have you ever stared at a volume print and felt the number was doing too much work for the size of the book sitting underneath it? That is the uneasy feeling that kicked off the latest Kalshi argument. A trader known as Beni pointed to roughly $538.6 million in 24-hour ETH perpetual turnover against about $3.1 million in open interest and called the gap proof that the venue was dressing up activity. The exchange’s crypto lead pushed back hard. He said two products had been mashed together and that prediction-market charts were being read as if they were perpetual-futures prints. I have covered enough market-structure fights to know this kind of dispute rarely ends in a tidy confession. It usually ends in filings, screenshots, and a lot of people talking past each other.

Why The Kalshi Volume Fight Caught Fire So Fast

Crypto audiences are allergic to inflated prints. That allergy is earned. For years, venues treated reported turnover like a billboard. If the number looked huge, listings followed, market makers followed, and casual traders assumed the tape was real. So when someone posts a ratio near 174 times volume to open interest, the room does not shrug. It leans in.

Beni’s September 20 thread framed the case in blunt language. He wrote that Kalshi fakes its crypto volume and claimed he could prove it. That sentence travels. It is short, angry, and easy to quote. What it is not, at least not yet, is a regulatory finding. No public enforcement action reviewed around September 21 accused the venue of wash trading in its crypto perpetual book. That distinction matters, even if it is less exciting than a dunk thread.

I’ve found that volume scandals tend to live in the gap between what a chart implies and what a contract actually is. Prediction markets count contracts one way. Perpetual futures count notional another way. Interfaces sometimes slap a dollar sign next to a unit that is not dollars. If you skip the glossary, you can convince yourself you have caught a crime when you have mostly caught a labeling habit.

The Screenshot Problem Nobody Wants To Sit With

Beni posted leaderboard captures and argued the largest visible ETH position at that moment was only $17,598. Pair that with hundreds of millions in reported turnover and the optics look terrible. Fair. Optics are not the same thing as a reconstructed tape, though. Trading data on a live venue moves by the minute. By the time a screenshot is quoted, the book has already changed. Independent reconstruction from public pages was not possible after the fact because the prints kept updating.

That does not make the question silly. A thin open-interest stack under a fat volume bar still deserves an explanation. High turnover with low outstanding risk can happen in liquid perps when traders scalp, flip, and recycle the same risk all day. It can also happen when two accounts are bouncing size back and forth. The hard part is telling those stories apart without account-level data.

An allegation can be loud, specific, and still unproven. Volume ratios raise questions. They do not, by themselves, close a case.

In my experience, the cleanest way through a fight like this is to separate three piles: the public numbers, the incentive design, and the surveillance claims. Mix them and you get a food fight. Keep them apart and you can at least see what is actually on the table.

Prediction Markets Are Not Perpetual Futures

Kalshi’s crypto lead, posting as IcoBeast, said the original spark came from an Artemis chart about prediction-market share, not perpetual futures. He argued the rebate conversation had been dragged onto the wrong product. According to the exchange’s own glossary, prediction-market volume is a contract count over a period. Perpetual futures are a different animal. They use margin, leverage, funding, and they do not expire on a fixed date the way a classic event contract does.

That product split is not a cute talking point. It changes how you should read every chart in the thread. If you take a prediction-market share graphic and treat it as a perp notional print, you will invent a scandal that the graphic was never built to support. IcoBeast also said the venue does not run the cited rebate program on its crypto prediction markets. The perpetual incentives, he said, sit in a separate bucket.

Beni separately argued that the interface parks prediction-market contract volume next to a dollar sign, which can make a contract tally look like dollars traded. The glossary still defines the unit as contract count. Whether older screenshots matched that definition pixel for pixel was not independently verified. Interface sloppiness is annoying. It is still not the same as a wash-trade ledger.

Perhaps the most interesting aspect is how often crypto readers treat every number on a screen as if it were mark-to-market cash. It isn’t always. Units drift. Labels drift. People then fight about morality when they should first fight about definitions.

What The September Rebate Filing Actually Says

Here is where the story gets less theatrical and more useful. Kalshi sent a program update to the Commodity Futures Trading Commission on September 2. Certification followed on September 16, according to the filing database. The program covers all of the venue’s perpetual markets, including cryptocurrency and metals, and is slated to run through December 31 unless it is changed or pulled earlier.

For cryptocurrency perpetuals, eligible taker fees are rebated down to 0.3 basis points, or 0.003%. Eligible makers get rebates that leave them with a net 0.3-basis-point payment. Eligible participants are defined as all Kalshi Self-Clearing Members. That is the structure Beni pointed at when he said matched flow could face little combined friction.

The filing does not pretend the design is costless. It says payments must be cut when overlapping incentive programs would produce net-negative combined maker and taker fees on an individual trade. In plain English, the venue told the regulator it does not want a trade that pays both sides to print for free in a way that turns the fee schedule into a money machine.

More important for the wash-trading charge, the same document excludes fees from trades that result from, or are being investigated for, self-matching, wash trading, pre-arranged trading, or other abusive practices. The Chief Regulatory Officer can yank program status and start disciplinary work when needed. So the filing confirms the rebate rates in the argument. It does not confirm that rebates were paid on fake prints.

Look at that table long enough and you see why both sides can feel righteous. One side sees a fee schedule that makes circular trading cheap. The other side sees written exclusions, a compliance officer with revocation power, and a product mix-up at the start of the thread. Both observations can be true at once. Cheap flow is a risk factor. A risk factor is not a verdict.

How Incentive Programs Quietly Shape The Tape

Maker rebates are not a Kalshi invention. They are a standard tool in electronic markets. You pay the passive side, or at least you make standing quotes less expensive, because empty books are useless. The trouble starts when the rebate is rich enough that the quote is no longer about price discovery. It becomes about farming the schedule.

Staff at the CFTC put out an advisory on August 12, more than a month before this fight. The note said well-built incentive programs can help liquidity and price discovery. It also said some designs can invite bad behavior if controls are weak. The advisory spends most of its pages on prediction-market programs, then reminds venues that core-principle duties still apply to other derivatives on designated contract markets.

Staff specifically warned that steep volume thresholds can raise wash-trading and pre-arranged-trade risk. Market-maker deals that guarantee net profits or cover losses through rebates can push people toward artificial strategies. That language is not an accusation against Kalshi. It is a map of where programs go wrong in general.

• Steep volume hurdles can reward circular printing instead of genuine two-sided flow.
• Guaranteed net profits can turn a market-making desk into a rebate harvester.
• Weak real-time surveillance leaves the ugly patterns sitting in the tape until someone screenshots them.
• Public, nondiscriminatory access still requires financial and operational filters, which is not the same as secretly picking winners.

Kalshi’s September 2 filing tries to answer several of those points. The company told the regulator it had reviewed manipulation risk, would watch participating Self-Clearing Members more closely, and would keep suspicious trades out of the rebate pool. It also said the program is public and offered on nondiscriminatory terms. IcoBeast pushed back on the idea that the venue hand-picks who gets paid. Guidance on designated contract markets stresses impartial access, while still letting venues demand that members meet capital and ops standards.

In February, the same regulator reminded exchanges that wash sales, pre-arranged trades, and noncompetitive trading can violate the Commodity Exchange Act. Venues have their own duty to keep audit trails, run surveillance, and enforce house rules. That reminder is industry-wide. It is not a caption under the ETH-PERP chart Beni posted.

Other Venues Pay Makers Too, And That Cuts Both Ways

IcoBeast compared Kalshi’s setup with programs elsewhere in derivatives. Public fee cards back the broad point. One major on-chain perp venue publishes maker rebates that reach negative 0.003% at the top displayed maker tier, while saying it does not run a private special-fee market-maker desk. A large centralized futures venue publishes negative maker fees for qualifying USD-margined tiers. Liquidity providers get paid to stand there. That is normal market plumbing.

Normal is not harmless. If everyone uses rebates, everyone also inherits rebate risk. The presence of a maker payment on Kalshi does not prove wash trading. The absence of a unique fee design does not prove the book is clean either. You still need trade-level evidence: matching accounts, timing clusters, inventory that never takes risk, cancels that look choreographed.

I’ve sat with enough fee schedules to say this out loud: a 0.3 basis point net on both sides is tight. Tight fees make genuine scalping cheaper. They also make circular matching cheaper. Designers know that. That is why exclusion language exists. Whether the exclusion is enforced is the part outsiders cannot see from a leaderboard JPEG.

Surveillance Promises Versus The Trade In Front Of You

Kalshi has talked up a multi-year arrangement with Nasdaq Market Surveillance. In an August 10 company note, it said the system would cover event contracts and perpetual futures and would watch across markets for manipulation, insider trading, and other abuse. That is a stated setup. It is not an independent audit of the specific ETH-PERP session Beni questioned.

This is where readers get sloppy. A brand-name surveillance vendor is comforting. Comfort is not a reconstructable print. Cross-market monitoring can catch patterns a single book misses. It can also miss a wash that stays inside one product and one member pair if the rules engine is tuned poorly. I would rather see the alert logic than the press line. We do not have the alert logic.

Still, pretending surveillance does not exist is just as lazy as pretending it settles the argument. The filing says heightened attention on Self-Clearing Members. The company says abusive trades are rebate-ineligible. The critic says the tape looks synthetic. Those three sentences can coexist until someone publishes account maps or a regulator does.

The Jump Angle Sounds Juicy. It Does Not Close The Case

Beni pulled Jump Trading into the story by pointing at a previously reported commercial link. A February report, citing people familiar with the matter, said Jump was set to receive a small equity stake in return for liquidity. The companies were not quoted confirming the exact terms. A separate public relationship exists around liquidity for Kalshi’s first bespoke prediction-market block trade on a carbon allowance contract.

None of the official papers reviewed for this dispute name Jump as the source of the ETH-PERP volume under fire. A commercial liquidity deal, even if real, does not equal wash trading. Market makers exist to warehouse risk and print two-sided markets. That is the job. The job can be done cleanly or it can be gamed. Equity sweetener plus rebate math is a reason to watch the book. It is not a signed confession.

I’ll be blunt. Name-dropping a well-known trading firm is an efficient way to make a thread feel bigger. Efficiency is not evidence. If the next drop includes identifiable accounts, timing clusters, and inventory that never leaves home, the conversation changes. Until then, Jump is atmosphere.

How ETH Perps Arrived On A Prediction-Market Venue

Kalshi’s ETH perpetual market has been live since June. The venue rolled Ethereum futures shortly after a regulated Bitcoin perpetual started trading in the United States, then expanded into Bitcoin and a wider set of altcoin perps, seventeen by the latest count in the source material. Early on, the company said perpetual volume topped $5.5 billion inside the first two weeks of the rollout. That figure came through the company and predates this fight.

That launch context matters because it explains why people care. A regulated event-contract venue stepping into crypto perps is a category collision. Prediction-market fans talk about probabilities. Perp traders talk about funding, basis, and liquidation maps. When those cultures share a brand, chart literacy gets messy. A contract-count widget can be read as cash volume. A rebate memo can be read as a cheat code. A liquidity partner can be read as a volume printer.

I do not think that collision is going away. If anything, more event venues will try adjacent derivatives because the customer overlap is obvious. The customer overlap is also how labeling fights become integrity fights overnight.

What Wash Trading Actually Looks Like When It Is Real

People throw the phrase around like confetti. Real wash flow has a shape. Two accounts, often under common control, trade with each other so ownership of risk barely changes. Prices may print, volume may swell, and the economic exposure stays put. Sometimes the point is leaderboard rank. Sometimes it is rebate harvest. Sometimes it is a listing narrative.

High volume against low open interest can fit that shape. It can also fit a busy scalp tape where risk is opened and closed inside minutes. Day-session futures pits have shown that pattern for decades without anyone calling the whole session fake. Crypto inherited the suspicion because too many offshore books really were padded. That history is why Beni’s ratio landed. History is not a substitute for this book’s audit trail.

1. Start with the unit. Is the print contracts, notional dollars, or something the interface never bothers to define?
2. Compare volume with open interest and with the visible top-of-book inventory, not with vibes.
3. Read the incentive memo. Cheap combined fees raise risk. Exclusion clauses show the venue at least knows the risk exists.
4. Ask whether the same product is being mixed with a different product’s chart.
5. Treat partner names as context until account-level evidence shows up.

That checklist is unglamorous. It also keeps you from writing a courtroom scene on top of a ratio.

The Late Twist: Non-Public Information And A Pause

Late in the exchange, Beni said he had received new non-public information and was sitting on another thread for roughly 24 to 48 hours while talking to lawyers. He said he would update after deciding what he could publish. At the time of writing, no regulator filing and no independently checkable packet supporting that newer claim had appeared in public.

That pause is either responsible or theatrical, depending on your mood. I lean responsible until the packet lands. Threatening a sequel is easy. Producing trade identifiers is hard. If the next chapter is real, it should survive contact with the rebate exclusions, the product glossary, and the absence of an enforcement caption. If it does not survive that contact, we are back to screenshots and ratios.

Why Readers Keep Mixing Integrity With Marketing

Crypto still sells itself with league tables. Biggest volume. Fastest listing. Deepest book. Those tables reward venues that look busy. Prediction markets added a second scoreboard: share of event flow. Put both scoreboards under one brand and you get a perfect environment for crossed wires.

IcoBeast wrote, in an earlier back-and-forth, that he knows the activity is real. Belief from a house lead is expected. Skepticism from a trader staring at a 174-times ratio is also expected. The market does not have to pick a hero. It has to pick a standard. The standard, for a designated contract market, is surveillance plus audit trail plus the willingness to kick rebate eligibility when the pattern looks circular.

Liquidity programs are not automatically corrupt. They are automatically tempting. Temptation is why the exclusion language exists.

Maybe that sounds too even-handed. Fine. I would rather be even-handed than useful to whichever side is louder this afternoon. The filing is checkable. The product split is checkable. The enforcement docket, as of September 21, does not carry a matching ETH-PERP wash case. Those are the load-bearing facts. Everything else is interpretation.

A Practical Read For Anyone Using The Book

If you trade these perps, treat reported volume as a popularity contest until you can see depth that actually absorbs size. Watch funding, spreads, and how quickly a modest clip moves the mid. Those tells are ruder and more honest than a 24-hour turnover headline.

If you care about market structure, read incentive memos the way you read a pitch deck. Look for net-negative combined fees, volume cliffs, and the sentence that says abusive flow is ineligible. Then ask who decides eligibility in practice. A Chief Regulatory Officer with revocation power is only as scary as the last revocation.

If you write about this space, stop collapsing prediction-market contract counts into perp notional. The glossary is sitting there. Use it. Dollar signs on a widget are not a philosophy. They are a design choice that can mislead people who already arrived suspicious.

Quick filter I keep coming back to:
  1. Unit first, narrative second.
  2. Ratio as a question, not a verdict.
  3. Rebate text as risk design, not proof of crime.
  4. Partner rumors as atmosphere until accounts appear.
  5. Enforcement docket as the adult in the room.

None of that will satisfy someone who already decided the venue is running a carnival. It will keep you from embarrassing yourself when the next screenshot drops and the units are wrong again.

Where This Leaves The Wash-Trading Charge

So where are we, after all the heat? A trader published a sharp ratio, a small top position, and a rebate critique. A house lead said the first spark was a prediction-market chart, not a perp tape, and that crypto prediction markets do not carry the cited rebate program. A certified filing confirms cheap crypto-perp fees and also excludes wash, self-match, and pre-arranged flow from those fees. Surveillance branding exists. A liquidity-partner rumor exists. A promised follow-up thread exists. A public enforcement case matching the allegation does not, at least not in the material reviewed through September 21.

That is an unfinished story, not a cleared one. Unfinished is allowed. Markets generate unfinished stories every week. The honest move is to keep the claim labeled as a claim, keep the filing labeled as a filing, and refuse to let a dollar sign on a widget do the work of an audit.

I keep thinking about how little patience this audience has for ambiguity. Volume is supposed to mean something. When it might mean two different somethings on two different products under one logo, people assume bad faith. Sometimes it is bad faith. Sometimes it is a glossary fight wearing a crime story’s clothes. This one still looks like the second thing, with just enough rebate math to keep the first thing plausible.

Watch the next 48 hours the critic flagged. Watch whether anyone publishes trade identifiers instead of adjectives. And if you only remember one line from this piece, make it this: cheap maker payments can invite circular flow, but a loud ratio is still only a loud ratio until the accounts show their hands.