Pre-IPO perpetual futures are already trying to price Anthropic before it lists: useful gauge or recreational bet?

By Bloomberg (Money Stuff column, by Matt Levine) · November 3, 2026.

In his newsletter, Levine poses a capital-markets question under the headline "How Much Should Anthropic Pop?", that is, how much Anthropic's stock should jump on its first day of trading. The text we had access to is cut off: it lays out the problem but not the specific figures on Anthropic. We therefore stick to what is actually in the material.

The starting point is pre-IPO perpetual futures. These are contracts that some offshore platforms, mainly Hyperliquid/Trade[XYZ], list on large private companies nearing a public listing. A perpetual future is a bet on a price that never expires. According to Levine, these contracts confer no rights over the underlying shares and, in theory, are off-limits to U.S. investors.

It is also unclear whether current shareholders can sell them to hedge their exposure, or whether large institutional investors use them to signal their demand in the IPO. Levine concludes that they are, in a sense, a recreational prediction market, not a derivative tied to the stock through arbitrage. Without that link, nothing forces the contract's price to resemble the future trading price.

The nuance he adds is that prediction markets have proven surprisingly accurate. According to Levine, Kalshi publishes research showing that its markets predict well. On the Money Stuff podcast, Jeremy Maletz, a prediction-markets trader at Susquehanna International Group, noted that probabilities come out well calibrated even with modest volumes. The wisdom of the crowd, the columnist sums up, largely works.

Our take: Anthropic's IPO is starting to be "priced" outside the regulated system before an official price exists. That does not validate anyone's price, but it is a sign of how value is discovered when a frontier AI company approaches the public market. The tempting thesis is that, if the crowd gets it right, these contracts will be a good preview of the debut. The caveat is that getting elections or games right does not mean getting right an IPO with restricted access and thin liquidity, where narrative weighs more than arbitrage.

There is a second layer. A big "pop" benefits those who buy in the offering and existing shareholders, but it also signals that the company left money on the table when setting the price. It is a classic IPO tension, magnified here by the scale of capital these labs need. Broadly, frontier AI demands investments in compute and energy that only the public market can sustain over the long term.

The optimistic long-term view is that financing that infrastructure opens the door to advances in health and greater abundance. The harsh short-term view is the risk of a bubble and speculative volatility. Since Jim Cramer, according to related coverage, is talking about "bears" who would like to derail the IPO, it is worth separating market noise from what actually happens with the business. Until we see the columnist's figures, the honest conclusion is a modest one: perpetuals are an interesting indicator, not a valuation.