Proposal: Uniswap Liquidity Program v0.1

Authors: Jon Itzler (Variant), Shreyas Hariharan and AG (Llama)

Updates:

  • The multi-sig has been changed to a 4 of 5 multi-sig, with 2 members of LlamaDAO and 3 members from the oversight committee. This was previously a 5 of 6 multi-sig with 3 members from Llama. In order to provide sufficient oversight, it makes sense for the multi-sig to be in the control over the oversight members instead of having equal representation from Llama.
  • We would like to emphasize that the payment of $150/hour for up to 30 hours a week goes to all Llama members. This is not the rate per Llama member. There was some confusion on this point in previous posts so we wanted to clarify it upfront.
  • We plan to post the on-chain proposal on August 31. We hope to get as much community feedback as possible over the next week before we post the on-chain proposal, so that the on-chain vote has already built consensus. So far, we have sought community feedback in our temperature check proposal, consensus check proposal, Snapshot vote, and hosted a Twitter spaces discussion to answer questions about ULP. The Snapshot vote was overwhelmingly positive (99%+ for votes) and we have incorporated feedback from each of our proposal discussions in the final proposal.
  • We have already built a Uniswap v3 Liquidity Program Dune dashboard to track the effectiveness of ULP. (Shoutout to Michael from Llama.)

Summary:

This proposal outlines a framework for establishing a liquidity incentives program (Uniswap Liquidity Program, ULP) with the goals of: accelerating the migration of liquidity to Uniswap v3; encouraging new market participants to experiment with liquidity provision, further distributing ownership of UNI tokens, and supporting ecosystem growth over the long-term.

We propose the program start with an initial maximum budget of 1.662m UNI, to be allocated as liquidity incentives across two quarters. Continuation of the program will require an additional governance vote. Upon renewal of the program, each of the three initiatives can be broken out into separate proposals.

The program will aim to bootstrap liquidity across three different initiatives:

  • Stablecoin pairs
  • Mid-tail pairs (e.g DeFi governance tokens)
  • Deposit receipt tokens (i.e. cTokens, aTokens)
    Similarly to the Uniswap Grants Program (UGP), pair selection is a subjective process that cannot be easily automated: any action requires close monitoring and frequent reporting to the Uniswap community. To this end, we propose a discretionary committee of 8 members — 5 core members from LlamaDAO to actively lead & manage the committee, as well 3 members for oversight. Find LlamaDAO’s full proposal & management plan detailed here.

The committee structure allows for incentives to be allocated efficiently across many pairs without requiring a full governance vote on each pair selection. However, Uniswap governance retains ultimate oversight by granting a budget on a bi-quarterly basis.

Purpose & Background:

While Uniswap v3 has quickly emerged as the market leading DEX, we believe that it has yet to reach its full potential. In particular, large amounts of liquidity remain locked in Uniswap v2, and a majority of v2 LPs have yet to migrate over the v3.

Liquidity incentive programs have seen varying degrees of success. In Uniswap’s case, the initial liquidity incentive program between September—November 2020 was met with considerable interest:

  • Protocol liquidity quickly increased from ~$750M in the week preceding the program to ~$3B at peak.
  • The number of individual addresses holding more than 0 LP tokens increased by ~75% from ~33,500 to ~59,000.
  • The unique UNI token holder base grew by 14% over the course of 2 months.

A contributing factor here is that Uniswap v2 has an organic user-base, which drives organic swap yields to LPs.

We expect and hope that a liquidity incentives program on Uniswap v3 would succeed in increasing wider LP participation, liquidity across sought-after pairs, UNI distribution, and swap volume. Importantly, due to v3’s Concentrated Liquidity feature, Uniswap governance can afford to pay significantly lower reward rates than previous liquidity mining programs to achieve similar market depth.

Quarterly Budget:

  • Max quarterly budget of up to 826,091 UNI across all initiatives to start
  • In addition to the quarterly budget, LlamaDAO (core committee members) is to be collectively compensated at a rate of $150 per hour up to 30 hours a week. Each month, oversight members will be responsible for calculating exchange rate from UNI to USD and appropriate payouts (estimated at ~2917 UNI per quarter), with payments being made in UNI.
  • Oversight members will be compensated at a rate of 300 UNI per quarter.

Pair Selection Committee:

  • 5 core committee members, composed of LlamaDAO team members for the first term, as well as 3 oversight committee members.
  • Each committee has a term of 2 quarters (6 months) after which the program and members need to be renewed by UNI governance.
  • Committee functions as a 4 of 5 multi-sig, with 2 members of LlamaDAO and 3 members from the oversight committee.
  • Committee to share a weekly report detailing the decision making process (using a standard template) and periodic updates detailing the successes and failures of the program.

Committee Members:

Committee member criteria include:

  • Credibly neutral — need to avoid any sense of conflict of interest
  • Ability to evaluate fundamental strengths of projects and their place in the ecosystem
  • Data driven — capable of assessing quantitative merits/growth signals

Committee members must recuse themselves from any ULP decision related to a project they hold an investment in or are otherwise related to.

Core Committee Members (LlamaDAO):

  • AG: project management and evaluation
  • Shreyas Hariharan: project management, coordinate with Uniswap governance and stakeholders
  • Mason: project management, create progress reports on ULP, coordinate with projects
  • Austin Green: technical and security expertise
  • Michael Silberling: create Dune Analytics dashboards that help us monitor the effectiveness of the liquidity program and enable us make changes when needed

Oversight Committee Members:

Implementation:

UNI is distributed to LPs that provide in-range liquidity, with those that concentrate their liquidity closer to the market price (i.e. higher virtual liquidity) receiving larger allocations.

Initial Proposed Reward Distribution Across Initiatives:

Stablecoin pairs

We suggest an initial stablecoin/stablecoin liquidity mining program over the course of the next three months conservatively targeting $250m in each of the following pools:

  • USDC/DAI [0.05% fee tier]
  • USDC/USDT [0.05% fee tier]
  • DAI/USDT [0.05% fee tier]

A reward rate of 2% APY feels like an appropriate target for stablecoin/stablecoin pools as v3 stablecoin/stablecoin pools are already generating organic yields in-line with money market protocols.

The committee would distribute 741.82 UNI per day to each of the three pools listed above for a total of 186,939 UNI across all three pools over the next quarter.

Mid-tail pairs

We suggest an initial mid-tail pair liquidity mining program over the course of the next quarter. We are conservatively targeting $100m in each of the following pools:

  • LINK/ETH [0.30% fee tier]
  • YFI/ETH [0.30% fee tier]
  • MKR/ETH [0.30% fee tier]
  • AAVE/ETH [0.30% fee tier]
  • COMP/ETH [0.30% fee tier]

A reward rate of 10% APY feels appropriate for mid-tail pair pools. Compared to stablecoin/stablecoin pools, LPs in mid-tail pairs must bear more price risk.

The committee would distribute 1,483.64 UNI per day to each of the five pools listed above for a total of 623,130.61 UNI across all five pools over the next quarter.

Deposit Receipt Tokens (DRT)

We suggest an initial DRT liquidity mining program over the course of the next quarter. Due to Uniswap’s brand recognition and gas optimized contracts, deeply liquid DRT pools may serve as an effective alternative distribution channel for yield-generating platforms. We are conservatively targeting $10m in each of the following pools:

  • DAI/cDAI [0.05% fee tier] [Compound]
  • ETH/wstETH [0.05% fee tier] [Lido]
  • USDC/PcUSDC [0.05% fee tier] [PoolTogether]

As with stablecoin/stablecoin pools, LPs in DRT pairs take on minimal price risk. Additionally, they earn passive yield by maintaining inventories of yield-bearing assets. A reward rate of 4% feels appropriate.

The committee would distribute 59.34 UNI per day to each of the three pools listed above for a total of 16,021.8 UNI across all three pools over the next quarter.

Conclusion:

We recommend the establishment of a Uniswap Liquidity Program (ULP), which will actively incentivize liquidity across three strategic categories:

  • Stablecoin/stablecoin pairs
  • Mid-tail pairs
  • Depositary receipt pairs

ULP will exist as a 8-person committee with 5 core members, where core committee members are tasked with closely monitoring the success of ongoing programs and frequently updating the Uniswap community via regular written reports. The 3 oversight members are responsible for monitoring the core committee and multisig signing.

We believe that the targeted nature of ULP’s proposals coupled with Uniswap’s organic user base will help achieve various objectives: wider LP participation, more liquidity across sought-after pairs, further UNI distribution, and increasing swap volume.

2 Likes

I´d really rather see the UNI being used to pay gas fees as 1INCH did rather than this steaming garbo

The situation is nowhere near comparable to September 2020 btw. People will just farm and dump UNI like they did before because the reason why they are not on V3 is the gas fees. The second the liquidity incentive runs out people will go back to V2 because it still wont be economically viable for them to stay on V3. This proposal is literally just burning resources into thin air.

So who benefits? The farmers and Llama

Who loses? The regular UNI token holders.

I will vote NO

1 Like

What do you mean by targeting 250M per pool? Incentives, volume or TLV?

How did you get to the selected pools?

Interesting proposal. I love seeing new ideas about how to grow the protocol and utilizing the treasury. I have a few questions/thoughts about the proposal.

  1. Do you have docs or a write-up for the staker contract?

  2. Has the staking contract undergone a review or audit?

  3. How did you choose LINK, YFI, MKR, AAVE, and COMP as your mid-tail markets? Should there be fewer or more markets?

  4. Any particular reason to choose 0.30% for Mid-tail pairs vs 0.05%?

  5. Is the goal to increase v3 TVL, number of trades, number of traders, market depth, market breadth, or something else? The summary says the ULP has the “goals of accelerating the migration of liquidity to Uniswap v3; encouraging new market participants to experiment with liquidity provision, further distributing ownership of UNI tokens, and supporting ecosystem growth over the long-term.” Maybe it would help to narrow it down to one goal so the community can try to measure progress objectively?

  6. Do you need one year of funding upfront, or will one or two quarters suffice to show progress?

  7. After the initial liquidity program launches, how will addition pairs get rewards?

2 Likes

Hey guys - thank you for your proposal. You clearly have put a lot of thought into it and think it is a great idea to leverage the treasury to drive growth to the protocol.

  • Given V3 leverages concentrated liquidity, should we be focusing on driving more liquidity or more integrations / volume? In other words, what’s the relative value of liquidity mining at this stage versus allocating more budget to grants for integrations / co-marketing?
  • It feels obvious that we should wait to do any liquidity incentives until Uniswap deploys on Arbitrum and / or Optimism to drive liquidity to those pools.
  • How will we measure success over time? Would we be better off running smaller experiments and seeing performance on retention?
2 Likes

Thanks for the proposal, I think liquidity incentives are definitely something the community should consider. However, I’m left wondering if a centralized committee structure is actually the ideal setup.

It seems we’re taking it as a given that centralized decision making is the optimal way to distribute a considerable amount of treasury funds, which seems at least a wasted opportunity in governance experimentation.

ULP and other programs would be setting an important precedent for how we want to decide on the use of the treasury, we should see these as opportunities to consider what tools could be built to make governance voting and decentralized decision making more efficient, and not just surrender to centralized committees as the default choice.

I found more info on the staking contract.

I still want to understand in detail how they plan to use it but at least the general structure is there.

What do you mean by targeting 250M per pool? Incentives, volume or TLV?

We are targeting $250m in liquidity in each of the stablecoin pools (USDC/DAI, USDC/USDT, DAI/USDT), $100m in liquidity in each of the mid-tail pools, and $10m in liquidity in the DRT pools.

Appreciate the thoughtful questions, @Getty.

Do you have docs or a write-up for the staker contract?

Staking contract docs: https://docs.uniswap.org/protocol/reference/periphery/staker/UniswapV3Staker
Staking contract source code: https://github.com/Uniswap/uniswap-v3-staker

Has the staking contract undergone a review or audit?

Yes, it has gone through an audit. There’s no report available but you can look at this: https://github.com/Uniswap/uniswap-v3-staker/issues?q=is%3Aissue+is%3Aclosed+label%3Atrailofbits.

Is the goal to increase v3 TVL, number of trades, number of traders, market depth, market breadth, or something else? The summary says the ULP has the “goals of accelerating the migration of liquidity to Uniswap v3; encouraging new market participants to experiment with liquidity provision, further distributing ownership of UNI tokens, and supporting ecosystem growth over the long-term.” Maybe it would help to narrow it down to one goal so the community can try to measure progress objectively?

The primary goal of ULP is to migrate liquidity to v3. We will monitor whether this is happening closely and report back to UNI governance periodically on progress. Michael’s Dune dashboards will help us track the effectiveness of ULP.

Do you need one year of funding upfront, or will one or two quarters suffice to show progress?

I think two quarters of funding is sufficient to evaluate the progress of ULP. We can consider adding additional pairs or tweaking incentives after assessing progress.


I will get to your questions on pool selection. But one thing to note is that this is the beginning of ULP and the program will evolve based on feedback from the market, community, and needs of Uniswap.

1 Like

Thank you for your insightful questions, @Unisocks.

Given V3 leverages concentrated liquidity, should we be focusing on driving more liquidity or more integrations / volume? In other words, what’s the relative value of liquidity mining at this stage versus allocating more budget to grants for integrations / co-marketing?

Both the liquidity program and integrations are complementary, not mutually exclusive. Getting each of the pools to the target liquidity we have set ($250m in stable pools, $100m in mid-tail pools, and $10m in DRT pools) would be a good first step for other initiatives.

It feels obvious that we should wait to do any liquidity incentives until Uniswap deploys on Arbitrum and / or Optimism to drive liquidity to those pools.

Including Optimism/Arbitrum is definitely an idea we are interested in as the program evolves. These ecosystems are young and it’s best to let Uniswap grow on L2s before adding liquidity incentives. This program focuses on bringing v3 liquidity on L1, which will be important for L2 liquidity as well.

How will we measure success over time?

We are primarily evaluating migration of liquidity to v3. Here are some items we are tracking on our Dune dashboard:

  • Is incentivizing v3 liquidity providing a better trader experience on v3?
    • Swap Performance: Share of trades where v3 has the best price (DEX aggregators as a proxy), and price impact trends for incentivized pairs.
    • Benchmark Growth: Share of trades and trading volume for incentivized pairs on v3 vs v2 and other DEXs
  • Is the liquidity program creating deeper v3 liquidity versus v2?
    • Liquidity Depth: Monitor total TVL and in-range liquidity for incentivized pairs, and volume of active LPs.
    • New v3 Liquidity: Track v2 liquidity migration to v3 in incentivized pairs, and compare new liquidity going to v3 vs v2.
  • LP Retention: Track how many users are LPing, how many are new vs returning LPs, and LP cohort retention.
    • Do we have the right rewards structure in place? What can we learn?
    • UNI Reward Structure: Change in TVL and in-range liquidity by UNI incentive tier.
    • External Rewards: Change in TVL and in-range liquidity for pairs with multiple incentives (i.e. native token liquidity mining, staking), and/or incentives on other DEXs.

Appreciate your input, @wario. Curious if you have thoughts on what an alternative system could look like? It might be difficult to implement for an initial version of ULP but would be great to explore going forward.

1 Like

Thanks @HelloShreyas, here’s an idea that should be trivial to implement for an initial version of the ULP.

Instead of a “Pair Selection committee” with an initial proposed distribution selected by some criteria never explained, we make the criteria for selecting pools explicit and codify the rules in english, like we would for the construction of an index fund. An index of pools is created following the specified rules, and the index is updated every determined period. The quarterly budget is distributed among the pools in this index also following some explicitly defined rules.

Surely you can see how the distribution of several million dollars from the treasury can be problematic when done arbitrarily by any selected committee, and that we should strive to minimize this when possible.

4 Likes

Will the liquidity providing program be through the offical uniswap app page?

Or will Llama be providing their own UI/UX as a third party?

How do you envision the program developing forward? I think a good layout would be for Llama to provide insight and guidance to recommend pairs based on data, while providing a UI/UX through the offical UI/UX that allows the UNI community to vote to activate recommended pools as batches.

1 Like

Hi Naught! Yes the Uniswap Labs team is building an interface in app.uniswap.org. If the community approves this proposal in on-chain vote, we’ll release the interface at the LM program start time.

2 Likes

Looking at v2.

  • $4.4b in tvl

  • Top ten pairs account for $1.6b in tvl. (excluding WISE & HKMT)

  • FEI related pools account for ~$700m of assets on v2.

  • Mirror related pools account for ~$420M of assets on v2.

The most affordable way to get liquidity to shift from v2 to v3 is to get FEI and Mirror to move their incentive programs onto v3. We should be asking why they haven’t moved already and what we can do to help them make the move.

Next, lets look at the top pairs (by TVL) on v2 (not Fei/Mirror).

  • USDC-ETH: $244m

  • ETH-USDT: $190m

  • WBTC-ETH: $152m

  • UNI-ETH: $112m

  • DAI-ETH: $81m

  • USDC-USDT: $68m

  • LINK-ETH: $63m

  • DAI-USDC: $55m

It is clear the money is in the stables and the mainstays of Ethereum. These are the pairs we should be incentivizing to get liquidity to migrate.

However, we should also help demystify v3 and make it easier for protocols and users to use and understand v3.

7 Likes

We at a16z would like to thank LlamaDAO for their hard work on this proposal.

After analyzing the proposal, we don’t think that the benefits of the proposal outweigh the cost of ~$50M in payments to LPs for the following reasons:

  • Uniswap v3 already has dominant DEX market share on Ethereum layer 1 without incentives.
  • A better use for liquidity mining would be incentivizing liquidity provision on layer 2 chains.
  • There are issues with the pairs chosen for incentivization.

Given these reasons, a16z does not support the proposal in its current form.

We explore each of the aforementioned points in greater detail below:

Uniswap v3 already has dominant DEX market share on Ethereum layer 1 without incentives.

Uniswap v3 alone already has >50% DEX market share on Ethereum layer 1. Furthermore, it already has deep liquidity with over $2.9B in value locked.

While there is more capital currently locked in v2, it’s not clear that the benefits of adding more liquidity to v3 will exceed the cost of the program, especially since v3 allows for much greater capital efficiency than v2 by design.

A better use for liquidity mining would be incentivizing liquidity provision on layer 2 chains.

The most pressing problem for Uniswap v3 on Ethereum layer 1 is not the depth of its pools but rather the cost of interacting with it. The median gas cost of trading through the v3 swapRouter contract over the last 7 days is 160K, which is ~$60 per swap at prevailing gas prices of ~100 gwei.

The most effective thing the Uniswap community could do in the short term is migrate to more scalable layer 2 blockchains like Arbitrum and Optimism.**

Liquidity mining is most effective when it’s used to “jump-start” network effects in a nascent ecosystem. In Uniswap’s case, a jump-start for liquidity could encourage trading and integrations from applications that require external liquidity. Taker demand and integrations could encourage more folks to provide liquidity to earn trading fees, which in turn could encourage more integrations, and so on. This “flywheel” effect has played out to some extent on Ethereum layer 1 already, but has yet to be sparked for layer 2 deployments.

Liquidity mining incentives on these chains could be a helpful jump-start.

There are issues with the pairs chosen for incentivization.

“Deposit Receipt Tokens” are not a promising DEX market. Redemption of these tokens from the protocols themselves (e.g. Compound) would potentially be more straightforward and less costly (avoiding slippage and a 5 bp fee). Furthermore, COMP rewards accruing to the cDAI in the DAI-cDAI market would be forfeited.

It’s worth noting, however, that a relatively small portion of incentives (~2%) will be going toward these pairs.

Incentivizing 5 bps stablecoin pairs won’t necessarily make Uniswap v3 more competitive. Offering low slippage on stablecoin<>stablecoin pairs is easier due to their low price volatility. As such, the determining factor is trading fees. Pouring more capital into a 5 bps fee pool won’t necessarily make Uniswap v3’s pricing more attractive, as lower cost pools exist such as Curve’s 3pool (3 bps fees) and DODO’s USDC-USDT pool (1 bps fees). Indeed, most USDC-USDT volume from 1inch is routed to DODO.

There isn’t a clear quantitative methodology behind identifying the mid tail pairs. The reasoning isn’t clear behind the 5 DeFi governance tokens chosen for incentivization as “mid tail pairs.” One of these tokens, AAVE, is not a great candidate given that liquidity is dominated by a balancer pool that provides users with LP tokens that can be used in AAVE staking. 1inch AAVE-ETH routing is dominated by this pool.


These shortcomings, in conjunction with the high price tag, should encourage the community to look for other ways to address the core purposes fueling the proposal. In light of this analysis we offer a few potential solutions below and welcome the community’s feedback in continuing this dialogue.

Looking forward:

There could be some lower cost ways for the treasury to incentivize LP migration to v3:

  • If the problem is high gas cost, the treasury could subsidize a % of gas costs for new LPs.
  • If the problem is lack of awareness and tools, there could be a grant for building better tools and documentation around v3 LPing.
  • There could be a grant to build easy tools for other teams to incentivize liquidity on v3. For example, Fei incentivizes a large pool on Uniswap v2. Understanding these uses and working with teams to incentivize v3 liquidity instead of v2 could be high leverage for UNI holders.

Furthermore, we think using liquidity mining to bootstrap layer 2 deployments could also be a high leverage use of funds.

**a16z is an investor in Optimism.

8 Likes

I was initially in support of this proposal by llama. I do see a need to attract more governance participants and active community members. I have noticed recently with this proposal, and flipside’s; a gravitation towards creating committee’s and salaries built around the proposal (businesses being built using the uniswap treasury as a bootstrap).

I think uniswap governance and participation would increase with direct incentive proposals, instead of creating small bureaucratic salaried teams; allowing community members to vote a direct seen benefit (layer 2 incentives, with a clear block start, and end date).

I think using liquidity mining for bootstraping layer 2 makes plenty of sense. As a uniswap user, I have not tried layer 2. My hesistation is from the unknown of how it works (switching networks) and potential asset lockup times. Uniswap community and engagement will grow if it is seen as a place to go and be the “first time user” onramp to layer 2.

1 Like

Update: we have decided to not proceed with this proposal. We will share our rationale over the next day.

Subsidizing gas fees would make a lot of sense to be absolutely honest…this would be the way to get most bang for our bucks.