ANT Arbitrage: Is the 10% Profit Still There? Anthropic hasn't even gone public yet, but discussions about arbitrage surrounding ANT are already heating up. Some are studying the prospectus, some are comparing exchange rules, and some are directly checking order books to calculate how much room there is left to enter the market. Combining community ideas and on-chain data, we'll look at this in several parts: I. What kind of money is being made in this arbitrage? Trading Anthropic, the number of shares corresponding to contracts on different platforms is different. Hyperliquid's xyz:ANT is quoted directly at the price per share; Binance's ANTHROPICUSDT uses 1 billion shares as the initial estimated base; OKX currently uses 10 billion shares as the estimated base. Before listing, everyone saw quotes in different units; afterwards, the relevant contracts will convert to the price system of listed stocks according to their respective rules. The market discussion structure is to go long on the relatively cheaper side and short on the more expensive side after conversion, waiting for the price difference to converge. Rebase will adjust both the price and the position size simultaneously, while maintaining the nominal value of the position itself. Profit margins come from the price difference between the two sides. II. Why is everyone focusing on "12.7 billion shares"? This is the most important parameter in the entire transaction. Based on item-by-item estimations in the prospectus, including preferred stock conversion, existing common stock, charitable foundation shares, shares converted and vested at listing, and the expected new issuance, we arrive at approximately 12.7 billion shares of post-IPO share capital. A fully diluted estimate would also consider options, unvested RSUs, and some unconverted securities, resulting in another estimate of approximately 16.7 billion shares. The prospectus does indeed list both outstanding shares and fully diluted shares separately, and clarifies that some Amazon convertible bonds will be reserved for conversion after listing due to shareholding caps. The 12.7 billion shares figure is a research estimate; figures for newly issued shares in the IPO, etc., still need to be supplemented in subsequent prospectuses. Based on 12.7 billion shares, shorting one unit of Binance ANTHROPIC requires a long position of approximately 12.7 shares equivalent in xyz. Both sides refer to the number of shares allocated. III. How many are left today? Based on the order book data at 17:43 on November 4th, the gross price difference for the Coinbase/Deribit combination was approximately 10.5%, and for the Binance combination, approximately 9.85%. I re-captured the order book data from 09:14 to 09:16 Beijing time on November 5th and calculated it step-by-step by active trade direction: - Hyperliquid Long: Bought 12.7 shares equivalent in xyz, with an average transaction price of approximately $116.30 and a notional amount of approximately $1476.98. - Binance Short: Sold 1 unit, with an average price of approximately $1595.13, resulting in a difference of approximately $118.15 and a gross profit margin of approximately 8.00%. - Coinbase/Deribit Short: Selling 1 unit at an average price of approximately $1600.95 yields a profit of approximately $123.97, resulting in a gross profit of approximately 8.39%. The above calculations assume 12.7 billion shares, with the percentage calculated using the notional amount of long positions as the denominator. Transaction and funding fees are not deducted, and this is a snapshot of the order book. The opportunity shown in the screenshot has changed; the profit calculated today is smaller than yesterday's. IV. Small and large funds face different order books. @SUOHA_AI suggests that small funds can compare Coinbase/Deribit short selling quotes, while large funds need to consider... Considering Binance's liquidity: The data we gathered today explains this difference: Calculated within a 0.5% range below their respective optimal bid prices, Binance has approximately $515,100 in visible buy orders, while Deribit has approximately $31,300. The former is about 16 times the latter. For those preparing to actively sell and short, these buy orders represent the available counterparties. With small trades, higher bids may offer an advantage; as the amount increases, the average transaction price becomes increasingly dependent on the subsequent liquidity. The same arbitrage logic will yield different price differences depending on the size of the capital. V. Where have the on-chain funds accumulated? This morning's Hyperliquid API snapshot shows: xyz:ANT saw approximately $29.8 million in trading volume over the past 24 hours, with an open notional value of approximately $59.13 million. Entropy's io:ANTH saw approximately $64.44 million in trading volume over the past 24 hours, with an open notional value of approximately $117 million. Both markets have reached considerable scale, but when choosing a long position, funding costs must be factored in. I have summarized the funding cost settlements over the past 24 hours: - xyz cumulative approximately 0.0016% - Entropy cumulative approximately 0.2345% Based on a fixed notional long position of $10,000, this corresponds to funding cost expenditures of approximately $0.16 and $23.45 respectively. Entropy's post-IPO conversion parameters are determined based on fully diluted equity, potentially resulting in a lower per-share equivalent price; simultaneously, funding costs during the holding period are continuously eroding this advantage. VI. What's most worth monitoring next? Three figures: the final converted number of shares, the available price difference for the trade, and the accumulated funding fees paid during the waiting period. Based on today's small-scale trades on Binance and XYZ, ignoring fees, the break-even share capital is approximately 13.72 billion shares. This means that the closer the final number of shares used is to this level, the thinner the original approximately 8% profit margin becomes. Before engaging in arbitrage, you need to clearly define which part of the profit you intend to make. Arbitrage across multiple platforms involving information and mechanisms carries a direct correlation between risk and reward; any change in one platform may necessitate a complete overhaul of the entire system. ANT Anthropic Hyperliquid Arbitrage