Decentralize Cosmos Hub with pSTAKE’s new validator delegation strategy for stkATOM

Decentralize Cosmos Hub with pSTAKE’s new validator delegation strategy for stkATOM

Overview

Decentralization is a founding principle of crypto and one of the guiding values behind pSTAKE. stkToken liquid staking solutions are built to bring a net positive to the ecosystem and its community. The basic idea behind stkATOM is to drive wide-scale adoption and increase ATOM use cases, but at the same time, help ensure sufficient decentralization and continual strengthening of the larger Cosmos ecosystem.

The basic idea behind stkATOM is to drive wide-scale adoption and increase use cases for pSTAKE’s ATOM liquid staking solution. At the same time, it helps to ensure sufficient decentralization and continual strengthening of the larger Cosmos ecosystem.

Currently, 62 validators (the biggest validator set for ATOM liquid staking) receive delegations through a community-driven approach to select validators based on criteria such as commission, governance participation, uptime, etc., powered by PSTAKE governance. Read more about it here. The selection process of validators was kept simple at launch to keep delegations simple at launch with a view of progressive decentralization.

This forum discussion proposes an updated automated validator delegation and rebalancing strategy for stkATOM to decentralize the Cosmos Hub.

Importance of Liquid Staking Providers

Validators assume a crucial role in PoS-based blockchain networks. There are many barriers currently that prevent small retail holders from becoming validators. Moreover, a validator has to play an active role and cannot afford to remain a passive entity. Validators must take responsibility for many aspects, including proposing and validating blocks, casting votes, coordinating committees, and managing stakes. For their continual and active participation, validators are rewarded adequately.

However, average users who represent a majority of the network can participate by delegating their stake. A fundamental question then arises: how do you select a validator?

Should a user choose a validator with the lowest commission fees, a validator with a large user base or large existing delegations, or one with a famous or familiar history? A decision based on the abovementioned requirements will concentrate power and authority in the hands of a few older validators and reduce decentralization in the network. With this in mind, pSTAKE is guided (but not limited to) by some basic principles outlined below.

As a liquid staking provider, pSTAKE should not be a validator gatekeeper and should deselect bad actors/validators rather than trying to select the right validators.

The current methods for selecting validators include the following:

  1. Governance-driven criteria (pSTAKE)
  2. Council-driven criteria (Stride)
  3. Signal intent (Quicksilver)

The governance-driven approach adopted by pSTAKE allows for the onboarding/removal of validators through proposals made by users/validators, who can then add/remove a validator from a network. The overall delegation process should be easy and convenient. It needs to be simple and automated to ensure scalability in the long term.

The original approach always intended to develop a mechanism in the long term to score validators’ performance based on parameters like uptime, latency, data center decentralization, commission rate, and existing delegation. An automated validator delegation and rebalancing strategy based on a scoring mechanism will help complement the overall and long-term objective of creating a scalable approach to validator selection and rebalancing across all Cosmos chains enabled on pSTAKE.

Proposed new validator delegation strategy

pSTAKE’s new delegation strategy uses a weighted scoring mechanism with decentralization parameters to suggest the final validator set and automate rebalancing across the chain. It will lead to further decentralization of ATOM delegations and lead to the following benefits:

  1. Constant liquid staking alignment with the Cosmos Hub validator set
  2. Increase transparency through on-chain and data-driven automated validator delegation and rebalancing strategy
  3. Reduce protocol and users’ slashing risk
  4. Promote validator behavior that is aligned with Cosmos Hub

The proposed solution will comprise the following steps:

  1. Collect on-chain validator data, including all decentralization parameters, in a transparent manner
  2. Apply the pSTAKE Delegation Strategy filtering criteria to rule out specific validators like CEXs or high commission
  3. Apply the pSTAKE Delegation Strategy weighted formula to calculate individual weights for the shortlisted validators
  4. Delegate tokens (including new deposits) to all validators according to their target weights daily
  5. Rebalance delegations regularly to meet the target weights

The pSTAKE Delegation Strategy for Cosmos Hub will consider the following decentralization parameters to filter and calculate validator weights:

  • Voting power: 0.05% to 5%
  • Commission: 5% to 10%
  • Average 90-day uptime: 95% to 100%
  • Governance participation in the last 180 days: 60% to 100%
  • Time in the active set in the last 180 days
  • Slashing events in the last 180 days

After running scripts based on the above metrics, this is the validator list with the corresponding weights that the code suggests to decentralize the Cosmos Hub. Please note that the list (including new validators and individual weights) will vary on a daily basis.

Participate

pSTAKE’s new Automated Validator Delegation and Rebalancing Strategy for stkATOM will help to improve Cosmos Hub decentralization and maintain constant liquid staking alignment.

The delegation strategy’s active lookout on decentralization parameters will ensure that ATOM stake distribution does not reach validators acting against the network’s best interest. It will lead to more decentralization and help achieve delegation scalability without needing constant manual intervention.

Everyone is invited to share their feedback and thoughts on the proposed delegation strategy, specifically on:

  1. What do you think of the various validator delegation methods? Do you think the proposed pSTAKE method significantly improves delegation transparency?
  2. What is your take on the various decentralization parameters considered?
  3. Do you think pSTAKE’s delegation strategy can help decentralize ATOM liquid staking and the Cosmos Hub?
  4. What are some cons or drawbacks of the proposed strategy?

Upon in-depth discussion, a proposal will be put on the PSTAKE Snapshot for voting, and the new delegation strategy will be implemented via a Persistence One chain upgrade.

1 Like

Hey @dneorej,

its great that pSTAKE is working on improving decentralization through liquid staking and i think the path you are taking is the right way since it significantly improves the transparency of delegations and also includes at least most of the validators.

I also like that the parameters are not too complicated but suffiecient for an evaluation. I have some questions about them though.
For parameters with ranges: Will these linearly affect the delegation amount? From 0 to 100%?
Does eligibility start after being in the active set for 180days?
And are no slashing events allowed in the last 180days to be eligible.

I guess the most difficult part is to balance these parameters propertly.

Yes correct, that’s linear between the min and max parameter. Being outside of the parameters makes a validator ineligible for any delegation.

Yes, a validator has to be active for 180days or more

Correct, if a validator has been jailed in the last 180 days, that means he would have been inactive at some point, hence ineligible.

There will be a front-end to make all of these things clear, similar as how it is already live for stkOSMO. Hope this clarifies

Seems sensible :+1:

What happens when a validator falls out of the criteria ? Is the delegation removed instantly ?

Based on these rules, if a validator gets slashed he would become ineligible for 180 days ; similarly if he doesn’t vote on a proposal he would need to wait for 20 new proposals to be posted and vote on each of them before becoming eligible again (which could take a while then) - correct ?

The validator will indeed be removed, but not instantly. Currently the rebalancing mechanism is still ‘passive’, meaning that with the next user who stakes/unstakes, the protocol will try to get closer to the target weights by staking/unstaking more from certain validators. This rebalancing mechanism will become ‘active’ after the next chain upgrade (currently under audit), where the protocol will actively try to rebalance stake to the target weights frequently.

That’s correct

No, so governance participation doesn’t need to be 100%, as long as the validator votes on 60% or more of the proposals in the last 180 days. The higher the participation, the higher the score and delegation. Saying this, I realise this wasn’t written correctly in the original post (I will edit it) - apologies for the confusion.

Hope these things help to clarify how it all works.

Got it, thanks for clarifying. Indeed the 100% governance participation criteria sounded a bit harsh, but this is much more achievable.

The paramters i would agree on are:

  • Time in the active set in the last 180 days
  • Slashing events in the last 180 days
  • Voting power: 0.05% to 5%

For these two parameters i think it would make sense to max out a little bit earlier, like 99% and 95%. Validators should have some room to priorize security over uptime/voting. This could also encourage validators to force behaviors like this (GAME Validator is messing with other validators "uptime" - Validation - Cosmos Hub Forum)

  • Average 90-day uptime: 95% to 100%
  • Governance participation in the last 180 days: 60% to 100%

For

  • Commission: 5% to 10%
    i would prefer a hard cutoff of about 20%.
    Teams need some leeway due to different infrastructure and tax regulations.
  • Commission: 5% to 10%

This could be changed. Cosmos Hub is getting minimum commission of 5% in month or two as per:

Commission: 1% to 10% would be better, if this will be active before that happens.

Max 10% is good though. pSTAKE users shouldn’t lose rewards because of extremely high fees.

ICF also uses 1-10% for their delegations.

Currently, 62 validators (the biggest validator set for ATOM liquid staking) receive delegations through a community-driven approach

Considering the Hub is providing abundant LST liquidity for different LSTs, it is beneficial imo to have a larger validator set. Otherwise voting power will continue to be concentrated into the hands of a smaller set of validators which could have genuine ramifications for Hub governance.

Also looking forward to the upcoming dashboard like the one already live for stkOSMO.

Working towards a 100% governance will be tough but exciting to see where it goes.

A commendable aspect of this proposal is its focus on transparency and data-driven decision-making. This not only mitigates the risk of concentration of power among a few validators but also promotes alignment with the network’s best interests.

Altho I have one thing in mind that may fall under:

Have we considered the potential impact on smaller validators or new entrants to the ecosystem? Excessively strict rules for selecting validators might unintentionally hurt some validators, limiting the variety and new ideas in the network.

Overall, I believe that this new delegation strategy aims to distribute delegations more evenly among validators, reducing the risk of centralization and enhancing network resilience, so, I’m all up for it.