原文：《Using Liquidity Mining 2.0 (LM2) to Distribute Rewards》by Henry He

Compiled by: ChinaDeFi

The DeFi craze in 2020 was triggered by the launch of Compound's governance token, which was entirely driven by the concept of liquidity mining. There is no doubt that liquidity mining has pushed DeFi into the spotlight, allowing more people to see the potential of DeFi relative to CeFi and TradFi. On the other hand, the abuse of liquidity mining and its many immature token buyers have indeed damaged the reputation of DeFi. For DeFi as a whole, the net benefits of liquidity mining are still up in the air.

There is something novel about liquidity mining now, but in general, most projects are simple imitations or direct forks. Even more sadly, there are still many scams in the market.

Obviously, the current liquidity mining design is not optimal, and it is also one of the main reasons why the project's liquidity mining plan is unsustainable. What's worse is that in the past two years, there has not even been any effort to fix the design flaws of liquidity mining.

Liquidity Mining 1.0 (LM1) — rewards are distributed based on the size of LP positions

Liquidity mining, in simple terms, is a token incentive program designed to attract liquidity providers (LPs) to provide liquidity for specific trading pairs/pools on an AMM.

Synthetix was the first to distribute reward tokens to LPs in its sETH Uniswap pool. To be more specific, in order to receive rewards, LPs need to first provide liquidity to the sETH pool on Uniswap, and then stake their Uniswap liquidity tokens to the staking reward contract created in 2019. Reward tokens will be fairly distributed to LPs based on the percentage of liquidity tokens staked by LPs relative to the total LPs’ staked tokens.

From the perspective of token economics incentive design, the liquidity mining method pioneered by Synthetix is ​​to distribute reward tokens based on the size of liquidity positions. We can define this method as Liquidity Mining 1.0 (LM1). From the results, such an incentive plan helps Synthetix achieve the goal of attracting more users to mint more sETH.

LM1 became the de facto design and implementation of liquidity mining. It enabled many projects to solve the liquidity problem to some extent, at least in the beginning. However, there were many problems in LM1 that made it unsustainable.

First, the fact that reward tokens are distributed to LPs even though there may be no or very few transactions actually means that the liquidity is not really well utilized. From a token economics perspective, using project tokens to incentivize liquidity is expensive for most projects because such incentives do not contribute much to the growth of the protocol economy. When liquidity is not utilized, the incentive scheme becomes worse.

Second, in many cases, multiple pools need to be incentivized. The existing approach is to allocate a certain number of reward tokens to each pool without considering the contribution of each pool, such as how many transactions were executed and how much trading volume was completed in each pool. Reward allocation decisions are either made by governance voting like Curve and Balancer, or by the team like Sushiswap, which is sometimes very arbitrary.

Liquidity Mining 2.0 (LM2) — Rewards are distributed based on fees earned by LP positions

LM1 can be improved upon, and a better liquidity mining incentive design would be to allocate reward tokens based on the AMM transaction fees earned by liquidity positions. This design is fundamentally different from allocating tokens based on the size of liquidity positions, and let’s define this approach as Liquidity Mining 2.0 (LM2). Clearly, LM2 solves the two major flaws in LM1 presented in the previous section.

First, within a fixed token distribution interval, if there is no transaction, then LP will not earn AMM transaction fees. Without fees, no reward tokens are distributed. In addition, it also discourages LPs from providing liquidity beyond the project's needs. With LM2, projects will not waste their precious tokens on unused liquidity, thereby reducing token inflation and downward pressure on token prices caused by liquidity mining.

Secondly, there is no need to manually allocate reward tokens to multiple pools through governance token voting or team decisions. These manual methods give LPs the wrong incentives and can also lead to unfair treatment of different pool liquidity. With LM2, if a certain LP position in the pool earns more AMM transaction fees, more reward tokens will be allocated to that LP. Simple and fair!

LM2 implementation: a difficult problem

Typically, a project issues an ERC20 token (mostly used as a governance token) and allocates a portion of these ERC20 tokens to a liquidity mining program. During liquidity mining, a fixed number of tokens are distributed at fixed time intervals.

In LM1, a fixed number of tokens per time interval is evenly distributed across the total number of all LP tokens used for liquidity mining. Each staked LP will receive the number of reward tokens based on the number of LP tokens they staked. During this time interval, whenever the number of LP tokens changes, the ratio is updated accordingly and the rewards are updated accordingly. This implementation ensures a fair distribution of reward tokens among all LPs participating in liquidity mining.

Unfortunately, distributing a fixed number of tokens at each time interval based on the trading fees collected by LP positions is practically very difficult to implement. In a fixed time interval, trading fees are driven by two dynamic and unpredictable parameters: 1) the time when LP positions generate and earn trading fees is dynamic and unpredictable, because no one can predict when traders will trade; 2) the trading fees generated and earned by LP positions are also dynamic and unpredictable, because the size of trades is also dynamic and unpredictable. Therefore, distributing a fixed number of tokens based on two dynamic and unpredictable parameters will produce an unfair distribution of reward tokens among all LPs participating in the liquidity mining program.

One solution is to adjust the relevant data distribution model for two dynamic parameters - transaction time and transaction size. Then develop an on-chain implementation that dynamically updates the model based on each new transaction and distributes reward tokens accordingly. This solution will be closer to a fair distribution of reward tokens among all LPs participating in liquidity mining.

There are certainly other ways to implement LM2. A better approach would be to adopt a new token model that not only has better token economics but also allows for easy implementation of LM2.

in conclusion

Liquidity mining has pushed DeFi into the spotlight. There are some flaws in the current liquidity mining design and implementation that make the liquidity mining scheme unsustainable. We can make improvements, and the reward tokens should be distributed based on the transaction fees earned by the LP position, rather than based on the size of the LP position. Due to the current token model and reward distribution schedule, it is a difficult thing to distribute reward tokens based on transaction fees. We believe that innovative solutions are coming.