加密韋馱|Skanda 🇹🇭
加密韋馱|Skanda 🇹🇭|Oct 13, 2025 09:58
Let me calculate the cost of listing HIP-3 for @ HyperliquidX The cost of opening a market is equal to: -500000 Hype, calculated at $40, is equivalent to a deposit of $2 million U. The opportunity cost of margin funds is calculated at an annualized rate of 10%, which is 2 million US dollars -Then we add the cost of the ticket and assume a 5000-10000 (negligible) -Oracle+gas and other costs, assuming 8-900000 U per year So the overall first-year new ticket cost (most currencies only need to look at the first year) is 2.1 million U (assuming not diluted by multiple tickets) The problem is that there are only three types of new markets that need to be deployed: 1. Popular new coins (such as XPL, PUMP, etc.): The pricing power is likely to be in Binance, with high cost expectations. Large market makers have the incentive to deploy them first, and the cost of deploying multiple markets can also be diluted. The number of such targets is very small, only a few per year. However, this type of target, which Hyperliquid already needs to meet, why do we have to pay for it? 2. RWA: Pricing power is in the off chain, and the on chain part belongs to the discrete market, with attention driven transactions. The quantity is large but the quantity is small 3. Binance Alpha+Futures level small currency: also a discrete market, unless large market makers are allowed to dilute unit market deployment costs If it is the project party or a small active market maker of the project party, then this fee is not cost-effective compared to the listing path of Binance Alpha or Bitget The reason why the vast majority of small currencies can be cashed out through contracts is because of Binance Futures, rather than simply having contracts If you want to climb up the ladder to Binance Futures or are not qualified enough, then big market makers may not accept it. If you deploy it yourself, listing and solving market making cannot pay for this cost What they want is quantity, not just quantity And precisely, the alpha market (retail demand) comes from 3 So in reality, only use cases 1, 2, and 3 are difficult to solve with this method, or in other words, it is difficult to solve with any Orderbook. Discrete market liquidity provides a natural disadvantage for order books Those who are more motivated should be project parties or market makers with sufficient funds, because they know that their projects can generate high transaction fees and they have control over the contracts, which is equivalent to supplementing their project mutual aid pool. They can also recover costs by brushing their own volume But for type 3, why not just create an unlicensed AMM Perp to solve this problem? I will issue it myself, lock in the pool myself, fully recover the handling fees, and I don't have to manage the market making myself. Starting from the volume ladder and climbing up to the top of the contract, isn't this unlicensed PERP enough to serve as a cold start? HL's ambition is to build its own institutional key customer relationship like CEX, but in fact, CEX's T0 VIP relationship revolves around lending conditions (credit), not the issue of listing authority However, HIP-3 does exert some pressure on the PERP of small CEX, but it is not the same blow as Uniswap did to the cryptocurrency business of third tier companies in the past
+3
Mentioned
Share To

Timeline

HotFlash

APP

X

Telegram

Facebook

Reddit

CopyLink

Hot Reads