Maintaining a high level of AMM liquidity is important for keeping bridging costs low and rates competitive. Higher AMM liquidity allows users to bridge large amounts with less slippage.
This is especially important for arbitrageurs. Bridge arbitrageurs make tight margins and need to offset their gas costs. Being able to transact in larger volumes with less slippage allows arbitrageurs to keep price differences across networks closer. Smaller price differences mean better rates for all users.
We believe a liquidity mining program starting at 2,200,000 HOP/month spread across the ETH, USDC, DAI, and USDT bridges may reduce bridging costs and drive higher volumes for Hop’s core assets.
Rewards can be split amongst networks and their AMMs based on approximate relative volumes as follows:
Network
Network %
Network HOP/month
Asset
Asset % of Network
HOP/month
Arbitrum
45%
990,000
ETH
58.00%
574,200
USDC
30.00%
297,000
DAI
6.00%
59,400
USDT
6.00%
59,400
Optimism
30%
660,000
ETH
58.00%
382,800
USDC
30.00%
198,000
DAI
6.50%
42,900
USDT
5.50%
36,300
Polygon
20.50%
451,000
ETH
43.00%
193,930
USDC
42.00%
189,420
DAI
7.00%
31,570
USDT
8.00%
36,080
Gnosis
4.50%
99,000
ETH
23.00%
22,770
USDC
44.00%
43,560
DAI
19.00%
18,810
USDT
14.00%
13,860
Towards the end of each month-long period, adjustments to the rewards distribution can be discussed and voted on, a different liquidity mining program such as the one described in this post can be adopted, or the liquidity mining program can be discontinued.
Looking forward to hearing your thoughts and feedback
What is the proposed infrastructure for the program? Is this staking HOP within the protocol itself akin to other bridge tokens, or through liquidity pools like Uni?
The infrastructure is the standard Synthetix StakingRewards contract which will be used to reward holders of the bridge AMM LP tokens (e.g., hETH/ETH LP).
I heard Game Theory of Liquidity mining during the community call and it made me think of the “tit-for-tat” optimal strategy. (Game Theory, Evolutionary Stable Strategies and the Evolution of Biological Interactions | Learn Science at Scitable)
I have never before thought about how we would integrate this basic type of logic into the liquidity incentives, but I believe that there is likely some connection. The largest issue I would think is that if we suppose that selling the token is “bad” that centralized exchanges enable users to short the token in a non-transparent way. If everything were on chain that would provide a real opportunity to track all actions on chain to evaluate users.
If we are to entertain $hop rewards for LPs, I think there should first be an emissions gauge with hop votes. Then the LPs should only be eligible for rewards if they lock the LP for a given amount of time. For instance, in incentivised Frax LPs if you lock your LP for 3 years you get 3x the rewards.
At best, liquidity mining kicks off a sustained growth that continues after the program ends. At worst, liquidity mining is a temporary bounty for mercenary capital that dumps tokens for dollars and leaves after the program ends. We should focus on learning until we are confident that it’s more the former than the latter. The latter is especially painful today when there are no protocol fees.
A simple experiment could be incentivizing the ETH pool on Arbi and the USDC pool on Optimism. In doing so, we could compare how the incentivized pools compare with the three unincentivized pools with the same asset on other chains as well as the three unincentivized pools on the same chain.
You could also flip it and incent most pools but withold from two markets and use those as controls.
A simple experiment could be incentivizing the ETH pool on Arbi and the USDC pool on Optimism. In doing so, we could compare how the incentivized pools compare with the three unincentivized pools with the same asset on other chains as well as the three unincentivized pools on the same chain.
I do love the idea of launching something, assessing the impact, and iterating from there.
For this experiment, how do you think success or failure could be measured? Is the intent to test how much liquidity the incentives attract or to test if the benefits of liquidity mining outweigh the costs?
I’d be in favor of compensating somebody to come up with a data driven approach to assess the ROI (including designing the experiment itself) but off the cuff:
How much liquidity is attracted by each dollar of HOP incentive
LP stickiness both in dollars and addresses 1 day, 1 wk, 1 month, 1 year after LM ends
Ultimately looking for how much “long term liquidity” we get per $ of HOP incentive - gets to a CAC/LTV type of view
Use the controls to try to isolate the variable of LM is on as much as possible.
To @defi_dad point, other structuring like vesting or kpi options can be interesting as well to combat mercenary capital.
I wanted to note an analysis which is alternative to the experiment placed above. It comes from the observation that when the reward token value goes down, so does the TVL. By plotting $HOP/$USDC against bond yield, $HOP/$ETH against staking yield, we can see the relative attractiveness of farming $HOP. Wallets with TVL which are correlated to this are mercenary wallets / deposits whose timing corresponds to high returns in are mercenary deposits that would be expected to increase in size if monetary value increases and decrease in size if monetary premium decreases. I can begin coding up a DUNE dashboard to get at this. Initially, the Hop DAO could target a premium similar to these two opportunities and then make dynamic decisions about how much Hop to allocate given the analysis of deposits. Similarly, if there weren’t a 0.4% fee on LPs there could be temporal experiments, which fluctuate the amount of $HOP from week to week to see the responsitivity of LPers.
Ultimately looking for how much “long term liquidity” we get per $ of HOP incentive - gets to a CAC/LTV type of view
Wallets with TVL which are correlated to this are mercenary wallets / deposits whose timing corresponds to high returns in are mercenary deposits that would be expected to increase in size if monetary value increases and decrease in size if monetary premium decreases.
I think it might be worth taking a step back to define what makes an LP “mercenary”. To me, a mercenary LP is one that dumps all of the protocol token rewards instead of holding them. They don’t care about building alignment with the project they are farming and have no stake in the project. This is a different from an LP that is sensitive and responsive to yields and may or may not have a stake in the project.
I don’t think we should think of liquidity mining as a way use high yields to attract LPs that will then stick around in a low yield environment either out of apathy or altruism. Instead, we should think of these subsidies as a way to accelerate growth across all of Hop’s markets (especially new ones) and support the markets that are not yet self-sufficient.
The cross-chain bridge market will likely grow orders of magnitudes over the next 5 years and HOP subsidies can help Hop maintain its market position during periods of high market growth.
