Volume Inflation Without Wash Trading

In June 2026, Kalshi launched markets in perpetual futures or perps. These are open-ended contracts that allow traders to bet on movements in the prices of other assets, including cryptocurrencies such as Bitcoin and Ether. People with no prior exposure to crypto assets and no ownership of crypto wallets could now speculate on their prices using US dollars on a regulated exchange, without having to use the services of any financial intermediary.

Trading volume on these markets was immediately spectacular, exceeding a billion dollars in the first week alone and rising steadily ever since.

But how seriously ought one to take these volume numbers?

A paper was posted online a couple of days ago that examined trade level data from the exchange and found some strikingly anomalous patterns. The author alleged that these patterns were due to widespread wash trading, which is an unlawful practice that involves movements of assets back and forth between colluding counterparties. Wash trading inflates volume without exposing any party to market risk, and is considered a form of volume manipulation.¹

The allegations were circulated widely on social media and drew a response from Kalshi. They were also covered by the Wall Street Journal, with whom I spoke about the incident. My interpretation of the data is that it is the result of incentives created by the exchange that result in dramatic increases in volume relative to open interest, but not through wash trading.

The most striking evidence presented in the paper in in Figure 3, which shows that almost sixty percent of total volume in the perp referencing Ether was generated by trades of a very particular dollar size, in this case $5,500. Similar clustering of trade sizes can be seen in other Kalshi markets. But no such patterns are to be found in related markets on other exchanges.

This pattern is highly unusual, and I can see why it was suggestive of wash trading at first. But there is an alternative explanation for the data that I think is more plausible.

Kalshi pays market makers to post orders subject to minimum size, spread and duration limits. The idea is to have tight spreads and large resting orders against which other market participants can trade. But the perp market is designed to track the price of the underlying asset very closely. As soon as a gap opens up between the price of an asset in the spot market and the price of the perp referencing it, an arbitrage opportunity is created. A trader can buy where the price is cheap and sell where it is more expensive, confident that the gap will close almost instantly.

This exposes market makers to adverse selection—the orders that they have posted will be snapped up by arbitrageurs as soon as an opportunity arises.² They will lose money on these trades, as their counterparties profit. To protect themselves against this risk, market makers have to keep track of the spot price and repeatedly cancel and replace their orders.

But occasionally they will have orders in place that are not cancelled quickly enough, and these will trade in full in a single transaction. The size of the trade is limited by the amount that is posted, and this is likely to be at the threshold necessary to secure the volume rewards. As a result, much of the volume will occur in blocks of very similar dollar size.

What is happening here is that the volume rewards paid by Kalshi to the market maker are flowing in part to aggressive traders who are able to move before stale orders can be cancelled. It is not even necessary that these traders are engaged in arbitrage—by simply watching the spot market they can predict where the perp price will go. They will close their open positions and secure their profits almost immediately.

Thus the data can be explained without appealing to unlawful wash trading.

But the alternative explanation also raises questions about what the purpose of the reward structure really is. By channeling funds to aggressive low latency traders through the market makers, the exchange is undoing the liquidity provision that the rewards were meant to boost. The problem is made worse by the fact that Kalshi currently refunds fees paid by high volume liquidity takers, which aggravates the adverse selection problem faced by liquidity providers. By eliminating these rebates, Kalshi could lower the costs of retaining market makers, while also providing greater liquidity to the rest of the market.

This would result in dramatically lower trading volume, which in my opinion would be a good thing. The lower volume numbers may be less eye-catching, but at least they would be more meaningful and more organically generated.

The CFTC recently issued an advisory addressing exactly this point, warning that certain kinds of incentive schemes could “encourage participants to trade solely to reach volume targets, heightening risks of wash-trading, pre-arranged trading, or other fraudulent, manipulative, or disruptive trading practices.” The practices described here may not be manipulative in the legal sense, but they certainly generate misleading data.

The regulatory agency has recently been a vigorous defender of prediction markets and highly permissive in its stance. Yet it is apparently investigating these latest allegations. This might be a good time for the exchange to reconsider the incentives it has created, because to an outside observer they seem designed less for the provision of liquidity than the inflation of volume.

Cover page of a recent CFTC advisory

1. The detection of wash trading is much easier with account or wallet identifiers. These were were not available to the author of the paper but are certainly available to the exchange, which also knows the identities of account holders.

2. The formal analysis of market making in the face of adverse selection was pioneered by Glosten and Milgrom and Kyle.