Secretly capturing 40% of the DEX market share, revealing the "invisible champion" on Solana.

Original Author: Eric, Foresight News

Reprint: Luke, Mars Finance

At the beginning of this year, the proprietary AMM on Solana sparked a lot of discussion in the English community. Initially, this was because people noticed several unknown AMMs starting to appear frequently in the routing of Jupiter, and these AMM projects had no front-end websites, no liquidity addition entry points, and no promotions. Over time, the trading volume routed to these “unknown” AMMs grew rapidly, and they have now become the largest DEX category on Solana.

According to statistics from Blockworks, on September 6 local time, in the trading of SOL with stablecoins on Solana DEX, prop AMMs including HumidFi, SolFi, Tessera, GoonFi, ZeroFi, and Obric accounted for over 70% of the trading volume, with prop AMMs also ranking first in all transactions on Solana, accounting for over 40%.

The development and growth of Prop AMM is a conquest by professional market makers in the AMM field.

Automated Market Makers (AMM) were originally designed to solve the market maker problem on-chain. In exchanges, user and market maker orders collectively create trading and underlying volatility, but it is difficult to achieve the same level of order book efficiency on-chain. AMMs based on inverse proportional functions have become almost the only best solution to this problem. Participants only need to deposit two types of tokens that make up a trading pair in a liquidity pool at a certain ratio to provide liquidity for trading and earn fee income.

Since its launch, AMM has continuously evolved, from automatically covering all price ranges with liquidity to allowing users to choose their own liquidity range, and now to helping users move their liquidity coverage at the protocol level. The core inverse function of AMM is also being replaced by more complex functions.

These improvements build upon the original foundation and primarily target two goals: reducing impermanent loss and optimizing execution prices. However, as a public AMM protocol, there is always an invisible ceiling: the protocol cannot arbitrarily adjust liquidity, but must adjust according to the protocol's settings. Based on this premise, the idea of establishing a non-public AMM that can flexibly adjust liquidity began to take shape.

Lifinity, launched in January 2022, is one of the earliest projects to introduce this type of mechanism, claiming to be an oracle-based AMM. Since the main prop AMM projects are mostly operated by institutions (such as SolFi operated by Ellipsis Labs and Tessera V operated by Wintermute), the actual pool liquidity and market-making mechanisms are not publicly available. We will explore this through Lifinity, which has disclosed some information.

The biggest features of Prop AMM are twofold: it is based on oracle pricing and allows flexible liquidity adjustments. For a typical AMM, the pricing is given after simulating trades through the corresponding liquidity pool, so the actual price usually includes slippage that might occur. In contrast, Prop AMM directly uses oracle pricing, which in most cases provides weighted real-time prices that do not include slippage from specific liquidity pools, resulting in better prices.

Prop AMM trading is mainly realized through aggregators, and the quotes provided directly to the aggregator by oracles are often better than those from regular AMM pools. After winning the orders with better quotes, the flexible liquidity adjustment capability of prop AMM comes into play. In publicly additive liquidity AMMs, the rules for adding liquidity are fixed, such as in Uniswap. Even though it now supports users choosing the range for adding liquidity, there may still be situations where high trading slippage occurs within a certain price range due to insufficient liquidity.

In the prop AMM, the deployer can concentrate all liquidity near the real-time price during transactions to achieve close to 0 slippage. This is also the liquidity adjustment method revealed by Obric in its documentation.

This method of concentrated liquidity is not the only trick of prop AMM, as theoretically, public AMMs can also be designed to concentrate liquidity when executing trades. However, the “killer move” of prop AMM is actually its undisclosed market-making strategy. The market-making strategy of Lifinity v2 is referred to as “Delta Neutral Market Maker,” which fixes the amount of SOL in the initial SOL/USDC liquidity pool rather than maintaining an equal value between SOL and USDC (as in Lifinity v1). The liquidity pool actively buys or sells SOL to trigger the balance of the liquidity pool, rather than being constrained by oracles and arbitrageurs, achieving proactive low buying and high selling. Unlike ordinary AMMs, which require arbitrage trades to balance the liquidity pool, the flexibility of prop AMM allows market making itself to generate profit rather than incur impermanent loss.

According to data from the Lifinity official website, its SOL/USDC liquidity pool boasts an annualized return of up to 1923.68%, with market-making profits even exceeding trading fee profits, reaching 1049.2%.

More prop AMMs developed by professional market makers may have better liquidity and more complex balancing algorithms, but we do not know for sure as they are not public. Compared to studying how to add liquidity in public AMMs to maximize returns, building your own AMM liquidity pool is clearly a more cost-effective approach.

The largest prop AMM on Solana, HumidFi, recently showcased a transaction on X where it sold nearly 1.5 million dollars worth of SOL for USDC. The transaction sold 7314.41412727 SOL, receiving 1,482,843.781553 USDC. At the time of the transaction, the price of SOL was between 202.67 and 202.87, and even using the highest price of 202.87, the total “loss” from transaction fees plus slippage was less than 0.0007, which is an extremely outrageous figure on-chain.

The possibilities of all the above are also attributed to some special mechanisms of the Solana chain itself.

Helius's report points out that prop AMM utilizes lightweight development frameworks such as Pinocchio and even sBPF Assembly to minimize the computational load required to update parameters, allowing it to refresh quotes at a fee that is almost negligible compared to the value of the transaction itself. In Solana's most widely used HumidiFi, updating a quote only consumes 143 CU, costing less than $0.002. In comparison, trading using Jupiter consumes approximately 150,000 CU.

The auction engine of Jito, the most widely used Solana client, determines transaction priority based on the tip for each CU. Oracle updates that consume lower CU can offer higher tips on each CU, naturally resulting in higher accepted priority. This makes the price update speed of prop AMM superior to that of ordinary AMM. This is not only the key for prop AMM to take business away from ordinary AMM through aggregators but also the premise that prop AMM can use oracles for almost real-time price updates.

In addition, Solana's extremely high transaction confirmation speed and lower costs provide an excellent foundation. If this were on Ethereum, just frequently updating oracle prices could incur significant costs.

In addition to the transaction confirmation mechanism, the high trading volume of DEX aggregators on Solana quickly demonstrates the advantages of prop AMM.

Blockworks data shows that transactions executed on Solana through DEX aggregators approached 50% in August, while transactions directly through aggregators on Ethereum during the same period were just slightly above 15%. Since most prop AMMs generate transactions through aggregator routing, if the usage frequency of aggregators is low and transaction volume is low, the advantages of high-frequency price updates, low slippage, and unique market-making algorithms are relatively difficult to demonstrate.

Blockworks researcher Carlos Gonzalez Campo predicts that in the future, trading of “mature” assets (i.e., assets whose prices do not experience sustained volatility) may increasingly be conducted through prop AMMs, while trading of new or long-tail assets will still be executed through traditional AMMs due to their significant price fluctuations that prevent professional market makers from intervening. In fact, some trading of larger market cap meme tokens on Solana has already begun to be routed to prop AMMs for execution.

Currently, many materials categorize prop AMM as part of dark pools, but there is an essential difference between the two. Although the characteristics of liquidity pools make it difficult for MEV trading to intervene, the transactions executed by prop AMM are still public and transparent; only the mechanism design of the liquidity pool itself is not disclosed. However, this also reflects that blockchain is gradually losing its loyalty to “decentralization” and “permissionlessness”. After the collapse of extreme idealism, sacrificing transparency for efficiency is beginning to take root in the mature track of Web3.

But in any case, we still welcome the challenge that DEX poses to exchanges. As Eugene Chen, the CEO of Ellipsis Labs operating SolFi, said, the lowest cost way to buy Bitcoin on Coinbase now is to transfer USDC to the Solana chain to buy cbBTC, and then transfer cbBTC back to Coinbase.