Please note: PowerPool Research is not competing with any ‘boutique’ research agencies, Research DAOs, and other similar entities. We created it for solving issues related to PowerPool/protocols we are building on top and for collaboration with other researchers.
Today we are excited to publicly announce a research group funded by PowerPool DAO and a brand new publication blog for PowerPool research materials on mirror.xyz.
We aim to adopt various DeSci (Decentralized Science) tools over time and make our research web3-native and compatible with modern research principles.
PowerPool Research was created inside PowerPool DAO more than a year ago to solve problems that every protocol faces: on-chain data collection & visualization, building simulation models, product prototyping, stress-testing, exploring, and understanding novel Defi primitives.
Why do we announce it publicly? What’s the goal of this announcement?
We stand for open-source collaboration for solving mechanism design, simulations, and protocol economic security problems.
This announcement is the start of our collaboration journey with other protocols and research DAOs
We aim to make public the models’ source code, input data, dashboards/queries for its collection, methodology, report articles, and other materials. In our opinion, collaboration and the opportunity to review results together is a great added value for researchers, protocols, and the community.
Aims, goals, and philosophy of operation
The main aim of PowerPool Research is to deliver quality research results, simulations, and data-backed answers to important questions for PowerPool and other related/partner projects.
Our goals:
- Establish collaboration with relevant projects and research DAOs in the space to solve problems related to protocols PowerPool is building on
- Carry out simulations, product prototyping, and data analysis
- Publish models, input data, and research results
- Contribute to robustness and safety of protocols PowerPool builds on top of and our/partner products
Talking about web3 engineering philosophy, here is a good article from Michael Zagrham on this topic. It comes down to Engineering Values, such as focusing on safety, forecasting side effects, deep involvement, putting best effort, and responsibility.
PowerPool more and more focuses on the PowerAgent on-chain automation network rather than building Defi products for the end-customers. According to it, we have an aim to contribute to products automated by PowerAgent from the research/data analytics/stress-testing side.
Principles of collaboration
The main idea of collaboration is to generate value together. According to this, we have the following rules of collaboration:
- PowerPool Research selects joint research topics with a focus to improve/prototyping protocols or products PowerPool has integration or related to
- PowerPool Research is a public open-source research group that collaborates and contributes to research with other projects free of charge. PowerPool DAO covers all expenses - such as the time and efforts of PowerPool contributors.
Thus, we publicly announce each research case we start to explore, with a problem statement, and other important facts prior to starting the job.
We ask for reciprocity from our collaborators, which includes (besides participation in the research process) at least a Twitter announcement and some other activity when results are delivered, listed below:
-> Participation in a joint article summarizing up results, and helping with finding and attracting candidates for peer review
-> A repost of the PowerPool Twitter storm with the results with a comment from the collaborator's side + original tweets covering research results
-> A community call to discuss results
The team
PowerPool Research team consists of several contributors right now:
-
Vasily Sumanov, the group leader
Head of Research, Ph.D. joined crypto space in 2013 as an enthusiast, contributing full-time to the space since 2017.
Contributing to Token Engineering, simulations, mechanism design, and Defi. In recent years Vasily had been researching AMMs and cryptoeconomic patterns. Besides PowerPool, recent contributions include work with Balancer Labs on Python AMM models and data collection for several Defi projects.
-
Princess Luna - “one line of code to simulate them all”
Python expert with physics academic background that decided to contribute his skills to web3 research problems. -
Sky Strifer. An intelligence agency for on-chain data.
Dune Analytics and Bigquery ninja, Data Engineer, crypto, and venture enthusiast. MSc. in Applied Mathematics & Physics. -
Adelie Penguin. Junior Researcher.
Master's student in physics, engineer, and full-time crypto nerd. Adelie builds simulation models and explores novel cryptoeconomic primitives and systems.
Recent research cases and results
PowerPool Research was initially launched to support PowerPool products from the research side. Here are our latest research topics:
Exploring and simulating approaches for determining the optimal composition of several Yearn.finance yield-generation vault indices based on on-chain data.
The main goal of the experiment was to understand is there an option to create an algorithm efficiently updating the basket of yVaults to maximize APY for yield-generating investment strategies (Vaults) with constantly changing yield due to environmental factors.
We made an attempt to find patterns in Vault LP share growth and ‘catch’ harvests and yield-compounding events for the hypothetical algorithm of basket rebalancing. For this purpose we collected on-chain data, visualized it, and backtested several algorithms:
The price velocity chart indicates harvesting & compounding events. The outcome of this research is that there is no patterns or algorithm for efficient capital re-allocation to capture harvesting events for yVaults basket. The model code & input data is published on Github.
Developing a capital-efficient StableSwap LUSD pool on top of Balancer v2, Liquity, and B.Protocol using automated asset managers.
PowerPool historically builds on top and contributes to Balancer Protocol. When Curve Wars started in autumn 2021, we decided to build StableSwap pool on top of Balancer v2 that can fund itself without hugely relying on ‘thin-air-printed’ token rewards.
Liquity protocol LUSD stablecoin has a unique feature to re-direct all liquidation cashflows to LUSD token holders. Using this option and some other composability tricks, we developed an idea of the staBAL-LUSD pool that evolves to LUSD-a-bb-USD pool, built on top of Balancer v2, Liquity, and B.Protocol:
We modeled the pool as if it actually operated on the mainnet during the experiment and tracked the evolution of balances and trading volumes. It also included liquidations in Liquity Stability Pool, liquidity mining rewards in third-party protocols (BAL, LQTY) that we can bring to the pool utilizing asset managers, and many more.
The results were outstanding - this pool could generate 12% APY in stable market conditions and up to 40% APY (25% monthly) while the market is going downside without additional token rewards:
We published two articles[1],[2] with results and a Github repository.
Determining the parameters for optimal $TORN autocompounding in an automated governance staking vault built on top of tornado.cash.
Since one of PowerPool products is PowerAgent on-chain automation network we decided to build a smart vault for staking $TORN into tornado.cash governance staking contract.
The main issue that we need to solve was the stochastic nature of rewards which means that if we want to maximize compounding interest by spending a reasonable amount of gas we need to explore data and create a harvesting & compounding strategy.
For this purpose, we collected data since the new $TORN tokenomics and governance contract was launched, explored it, and propose optimal vault operation conditions.
We determined optimal operational conditions for vaults with various amounts of TORN staked taking into account real network conditions such as gas pricing.
Results are published on PowerPool resources[3],[4], and Github.
Several other cases are now in progress, including the very important topic of on-chain automation (on-chain networks, signer selection algorithms, and cryptoeconomic models of operation).
Let’s research together
We are open to research collaboration with other protocols, persons, and entities. Note, that we focused on research cases contributing to solving problems related to protocols PowerPool builds on or related to. Please, reach out to PowerPool Research's Twitter or research@powerpool.finance.