[Proposal] Developing a sustainable, objective-based liquidity incentive engine for new venues

[paulsengh]

Our Proposal

As a member of the Lido DAO and core contributor to Delta One, I am submitting a proposal to build an objective-based framework for bootstrapping liquidity on new venues, starting with DEXs on Optimism. Delta One looks to improve the liquidity of stETH by crafting a robust, tailor-made incentive engine that is founded upon data–driven research and a first principles analysis. This proposal stems from the research motivated by the Steakhouse Financial team (outlined in their most recent post), as well as a number of discussions with the Lido analytics + reWARDS committee over the past month.

Delta One is a liquidity management platform that offers data-driven and actionable insights for DAOs to cost-effectively source liquidity. Delta One is non-custodial, fully on-chain, and uses state-of-the-art analytics to keep the community informed.

Key Objectives

Delta One aims to accomplish the following objectives:

1. Cut liquidity mining costs by moving towards data-driven models for incentive distribution in contrast to subjective decision making.
2. Improve stETH-ETH and stETH-USDC liquidity on Optimism in a cost-efficient and lasting manner to help the community make stETH the leading base asset across trading pairs.
3. Reduce price impact on Optimism stETH pools to encourage more users to enter positions and participate in the network.
4. Take a first principles approach to quantifying the amount of liquidity needed in stETH pools based on demand, and distill this information into comprehensible liquidity scores.

A Novel Liquidity Incentive Engine

The April rewards budget outlines an allocation of ~700k LDO, which is a substantial expenditure for the DAO. We are glad to see the reductions in expenditure across the past few months, but there is still significant room for growth on optimizing the incentive process. Rewards should be carefully used to bootstrap a network, but the underlying success of stETH should be measured by its velocity within the ecosystem; this velocity should make trading fees high enough for LPs to be profitable without additional rewards. One of the goals of this research is to identify when we have reached escape velocity, and which networks need to be seeded with more incentives.

The reWARDS committee has been granted a large quantity of rewards from Optimism, which will start to be distributed in April. Optimism is one of very few chains that is growing in TVL during this market, and we hope to make the most out of this opportunity for the expansion of stETH. If managed incorrectly, the Optimism deployment can become an expensive liability, as mercenary liquidity providers will abandon the network when rewards dry up––the new pools may never reach escape velocity, and future LPs will pay the price.

Delta One will conduct a deep-dive analysis on the current set of liquidity providers on stETH pools, creating a model for rating an LP based on a variety of factors (see specification below). The research and analysis (Phase 1) will motivate a variety of ways for the reWARDS committee to consume these actionable analytics, which we outline in Phase 2. We believe a tailor-made incentive engine can significantly cut costs by rewarding LPs through well-structured principles, rather than supporting ephemeral yield farmers.

While the community has built an impressive set of liquidity analytics compared to most DAOs, we believe mandating an external team such as ourselves can offer a novel perspective on a well-trodden problem. Delta One is committed to working with the pre-existing data frameworks, and iterating with the reWARDS committee to leverage insights from previous experiments.

Specification

Phase 1: Liquidity assessment and analytics

• Delta One prepares a dynamic and robust report with the following information
  • Liquidity provider distribution of every stETH pool on Uniswap
  • Number of total LPs per pool
  • Create an incentive framework that encourages future LPs to behave well according to the following parameters:
    • Current and historical holding of other liquid staking tokens
    • Current and historical holding of LDO token
    • Historical performance of stETH LP position
    • Liquidity depth of position
    • Price range supported
    • Historical LP positions to detect volatility (i.e. JIT liquidity)
    • Selling of OP and LDO rewards received from incentives
    • Liquidity score that encapsulates the above information with a public leaderboard on LPs quality rank
  • Pool-level data
    • Current and historical sell volume
    • Current and historical add/remove volume
    • Comparative analysis to similar tokens (i.e. cbETH, rETH)
    • First principles calculation of TVL target using data-driven modeling from the above information
    • Liquidity per $1 LDO+OP/month
  • Weekly incentive budget recommendations based on quantitative modeling pool data (built off any pre-existing work done by the analytics + finance team), broken down by how much each pool across various venues should be rewarded. For instance, if it’s more efficient to give LDO + OP on Uniswap’s wstETH-ETH compared to Balancer’s wstETH-ETH, then the report will recommend an increase in the former and reduction in the latter.
  • Portal for existing LPs to receive rewards for answering survey questions about their position

Phase 2: Execute recommendations

• Delta One creates a set of actionable strategies, and works with the reWARDS committee to safely and seamlessly execute the liquidity recommendations
  • Based on the analytics from Phase 1, Delta One will propose a governance vote to distribute rewards in a customized fashion, rather than using the AMM’s generic staking mechanism. The LP would not necessarily need to come to a frontend, as rewards can be airdropped to them based on the rubric above (time in range, time in pool, etc.); the implementation details will be finalized after polling community members on a variety of options presented by Delta One.
  • If appropriate, Delta One will create an interface to disperse incentives on both a per-pool and per-LP basis. The interface may connect to an optional smart contract (haltable by the Lido DAO community through a vote) to make distribution of rewards more nuanced (locked rewards, variable per hour/day/week based on historical volume). This frontend, hosted and owned by Delta One, will also allow LPs to analyze their historical rewards performance, and receive general guidelines on how to be a better LP.

Phase 3: Continuous optimizations

• Delta One provides continuous incentive management support to maximize growth in a capital-efficient manner to create long-term, sustainable growth across new venues.
• Regular monitoring based on key metrics and market conditions, with consistently updated recommendations.

To cover the costs of leveraging 2-3 data scientists, 1 frontend engineer, 1 designer, and 1-2 quantitative analysts, Delta One requests a 25k LDO and 50k OP grant for a duration of 3 months.

About Delta One

Delta One is a decentralized market making protocol that specializes in set-and-forget liquidity provisioning strategies. We work with a number of leading institutions, trading firms, and protail DeFi users to keep crypto markets liquid and confident. Given our proximity with leading market making participants, Delta One’s network is immensely valuable to DAOs that are looking to take their liquidity management to the next level. One of the major contributors to the protocol is OpenBlock Labs, which is backed by notable figures in the crypto space, including: Foundation Capital, Electric Capital, Circle Ventures, AlleyCorp, and others.

Since its inception in 2021, Delta One has been working with a number of prominent DAOs and institutional players for liquidity management. It has also been audited by top-tier firms, including Halborn, OtterSec, and Neodyme.

The team has backgrounds from Stanford, a16z, Carnegie Mellon, Meta, Palantir, and other top-tier institutions; the highly technical background of our team makes us confident that Delta One is uniquely positioned to tackle a problem of this nature.

Final Thoughts

Delta One looks forward to assisting in the design of a sustainable liquidity model for stETH across Ethereum and Optimism. This grant will be a great first step towards solidifying a productive relationship for both communities. We look forward to hearing the community’s thoughts, and answering any questions. Thank you for your consideration.

[jbeezy]

I am in support of this proposal. It is super interesting to see how we can eventually extend this to include lending or eventually build the PID controller proposed by @steakhouse.

I realize crawl, walk, run and all. Would encourage the outcomes to be open to public review if possible.

[carvas]

Thank you @paulsengh for the proposal and for all the exchange of ideas in private so far.

What’s needed in my opinion if we proceed (all these comments being conditional on approval ofc):

1. An extensive sync on which of these analytics (under “phase 1”) are already developed and which aren’t. A meaningful number are already so let’s avoid duplicating work. I am fully available to do this review with your team if this grant is approved.

2. There are other pertinent considerations to how much stETH pools liquidity is needed beyond “demand” (which I assume means “demand [for swapping]” in your sentence). So, totally agree that this should be arrived at from first principles but it’ll be vital to coordinate the goals with the DAO and reWARDs committee (like you do mention elsewhere).

@paulsengh

Take a first principles approach to quantifying the amount of liquidity needed in stETH pools based on demand, and distill this information into comprehensible liquidity scores.

3. Would this be a grant paid to Delta One contributors, to OpenBlock Labs, or to a Delta One protocol multisig/else? (not sure what’s your structure atm, help me)

4. Currently, most rewards cycles are monthly, and most distributions are triggered by the reWARDs multisigs (after being approved by the DAO in that month’s reWARDs budget). Weekly distributions under that regime would not work operationally. As such, this would require some sort of dev lift to set up rewarder contracts that could do do that variability somehow (extra input needed).

@paulsengh

Weekly incentive budget recommendations

Regarding the price of the grant proposed, I don’t have insightful comments and defer to both LEGO and Stakehouse.
Regarding Delta One, all my experience so far interfacing with the team has been great, and think they can add value contributing to Lido.

[steakhouse]

This proposal seems perfectly reasonable and exactly in line with our liquidity management goals. We’d be pleased to support Delta One in their analysis and we’d, similarly to @jbeezy, appreciate seeing public release of many of these findings.

Echo the ask from jbeezy to extend the analysis to include lending platforms, but just starting with liquidity would put us in a great place.

@paulsengh

• First principles calculation of TVL target using data-driven modeling from the above information
• Liquidity per $1 LDO+OP/month

Especially keen to see the methodology behind this research and the subsequent conclusions.

[paulsengh]

Thank you @carvas for the comment and all the insightful dialogue over the past 1-2 months.

1. Leveraging existing frameworks and previous experiments is a top priority, and we are committed to working with the reWARDs committee to avoid duplicate work.

2. Yes, it is important to clearly define the type of demand that incentives should be used to promote and, in general, the key performance indicators (KPIs) that the community seeks to measure. For instance, an increase of stETH as collateral on lending protocols motivates the use of incentives for liquid stETH DEX markets that can withstand sizable liquidations. Another KPI may be the volume on pools with stETH as the base asset, which entails a different incentive rubric. Each new venue and ecosystem will have a unique landscape to navigate, so we look forward to coordinating with the DAO and reWARDs committee to align on these goals.

3. This grant will be paid to a Delta One multisig and dispersed to contributors. One of the core contributors is OpenBlock Labs.

4. We kept the implementation details of the distribution open-ended for now, as the exact direction will be more clearly motivated once we have done a deeper dive from Phase 1 on creating the incentive framework.

[deepatalwar]

Hi all - Deepa from Gauntlet. We’re quite familiar with this space - we build infrastructure to dynamically optimize incentives for Immutable and AMMs.

@paulsengh

Delta One will conduct a deep-dive analysis on the current set of liquidity providers on stETH pools, creating a model for rating an LP based on a variety of factors

What is your proposed research methodology for analyzing LP behavior in order to achieve those 4 objectives?

@paulsengh

Weekly incentive budget recommendations based on quantitative modeling pool data

What is your proposed framework / system for making dynamic recommendations?

@jbeezy

I am in support of this proposal. It is super interesting to see how we can eventually extend this to include lending or eventually build the PID controller proposed by @steakhouse

We’re happy to support here as well. We build PID controllers like this to dynamically optimize incentives:

[paulsengh]

Thanks for your comment, Deepa! We will be sure to engage community members like yourself to obtain feedback on our methodology. We look forward to hearing your informed opinions on those posts as we dive into the research. This proposal is primarily to motivate the research, and the implementation details will vary based on discussions with the reWARDs/analytics committee, and the broader community. We do outline a number of key LP attributes that we plan to analyze (see Specification section), but we also want to leverage as many of the existing frameworks and models as possible.

If you’d like to collaborate with our team on specific parts of the proposal, feel free to also reach out to me on Twitter (https://twitter.com/paulsengh); we are working with a number of contributors across the community, and engaging with a number of experienced committee members on a recurring basis. There is a lot of context from conversations over the past 1-2 months that we’d be happy to catch you up on.

[vsh]

Hey! I’m on LEGO council, and I don’t agree completely with the objectives here. I would vote yay for a proposal for liquidity incentivization that would explicitly include these two goals:

• bringing extra stake in Lido
• increasing the organic trading volume vs. stablecoins,

with metrics that gauge the efficiency of achieving these two goals per $ spent. As this is a coincentivization effort, Optimism-related goals should also be present.

E.g. out of your goals:

1. Cut liquidity mining costs by moving towards data-driven models for incentive distribution in contrast to subjective decision making.

Can be achieved extremely easily by not incentivizing; it’s not a goal I would support. I think Lido DAO would want to spend money with efficiency, not stop spending money. I think that the thought behind this point is healthy, but it should not be a goal of its own. Rather would prefer to incorporate it in most of the metrics of success you use.

2. Improve stETH-ETH and stETH-USDC liquidity on Optimism in a cost-efficient and lasting manner to help the community make stETH the leading base asset across trading pairs.

Here making stETH base asset I support wholeheartedly - it supports both the goals I outlined, but I’d like to understand the metrics of success here you might have in mind.

3. Reduce price impact on Optimism stETH pools to encourage more users to enter positions and participate in the network.

Here I would focus on attracting organic arbitrage rather than reducing price impact. Growing an efficient market would have a way more lasting impact than pumping up straight liquidity (which IMO might be a good first step in the chicken an egg problem, but we can’t be stuck at Phase 1 forever).

4. Take a first principles approach to quantifying the amount of liquidity needed in stETH pools based on demand, and distill this information into comprehensible liquidity scores.

Needed for what exactly? This is also a secondary goal, I think. E.g. to achieve great organic volume Lido DAO could want to have a pool in protocol X with depth of Y and fee Z; growing it beyond this size would have diminishing returns.

I’m just one person on the council tho.