uallate

x.com/i/article/2021…

x.com/i/article/1995…

Perps have only a limited set of dimensions across which one can innovate and over the last 2 years we saw innovation across all of them except Collateral. 1. Liquidity Provisioning. @HyperliquidX was the first to launch the community vault, providing APR for depositors, reinforcing exchange liquidity and trading activity, which is now adopted by most players in the space. 2. Fees Structure. @Lighter_xyz was the first to introduce the zero fees model. As discussed many times, there are both pros and cons to this approach for the business and for end users, but it was an innovation in the market. 3. Ways of trading. @HyperliquidX came up with builders code (which we launched as well), allowing external teams to access underlying exchange infrastructure and liquidity and enable different frontends and use cases for trading perps (wallets, trading terminals, Telegram bots, etc). I believe we are yet to see innovation in the UX of perps trading, as the main trade screen of exchanges has not significantly changed over the years. 4. Markets one can trade. In 2025 we saw multiple perp DEXs, including @extendedapp, offering TradFi perps and I believe that over the coming years this trend will strengthen and we will see growth in OI, volumes and exchanges offering them. Earlier last year we saw pre-launch markets, which were an interesting concept that never really took off due to the complexities of providing stable and deep liquidity. Generally, adding new market types is the least important innovation, as any perp DEX can offer any market as long as there is a mark price and a market maker or vault ready to quote those assets. 5. Collateral. I do not think we saw any innovation on this front, with USDC being the sole collateral on perp DEXs while cross asset collateral and unified margin have been available on CEXs for years. Next Monday we will be releasing tokenisation of vault shares (a new type of collateral) and we believe it stands a good chance of getting traction among our users and potentially being adopted by other perp DEXs in the longer run, similar to how every perp DEX adopted community vaults. We will be releasing it in stages, with the full release plan to be communicated tomorrow. In a nutshell the idea is very simple. Allow users to post vault deposits or shares as yield bearing collateral for perps trading. In the target state: 1. eXtended Vault Shares (XVS) will contribute 90 percent of their value to user equity, essentially doubling overall collateral available for trading on the exchange. 2. Yield that vault depositors receive will depend on the trading activity of the user. There will be a base yield (currently ~15% all-time APR) available to all vault depositors and an extra yield that will depend on the user activity. The most active users will receive the highest yield. The fair question that many would ask is how liquidations will work given that XVS represents a claim on the vault equity, which includes open positions. The logic will work as follows: 1. When a user with an XVS balance becomes liquidatable, the first step of the liquidation process is to liquidate their XVS balance, meaning withdraw it from the vault. 2. When performing a liquidation or withdrawal of vault shares, the vault closes its open positions in proportion to the liquidated or withdrawn XVS. 3. If the vault for some reason cannot close its positions via the order book (e.g., due to lack of liquidity), a new type of operation will be enforced, called Force Close. The vault will close positions against the most profitable and highly leveraged user on the opposite side at the Mark Price. Force Close is similar to ADL but unlike ADL, where positions are closed at the bankruptcy price of the liquidated user, in Force Close positions are closed at Mark Price without incurring direct losses to the user on the opposite side. Given the existing risk limits of the vault (i.e. max position it can take on on the given market), we expect force close to be used only in the extreme scenarios. 4. Given the above, the XVS balance of the liquidated user can always be withdrawn from the vault. 5. Once the XVS balance of the liquidated user is withdrawn, we proceed with the regular liquidation process. The other valid concern we heard is what happens if the vault suffers significant losses while liquidating unhealthy users (the JELLY case), driving XVS price down and leading to a cascade of liquidations. Extended Vault processes liquidations deterministically, meaning that before taking over a liquidated position it knows: 1. that the position can be closed (vault will never take over the position it can't close) 2. that the realised loss will not be above its maximum acceptable loss per trade, per market or per day Hope the above clarifies the intended logic of the upcoming XVS release.