PIP-2: Proposal to distribute and decentralize staking power across the Polygon chain

Status: Draft

Abstract: Polygon has seen an explosion in staking in the over one year mainnet has been active. All 100 validator slots are now filled, but the distribution of delegated MATIC is clustered around large validators with low or 0% commission. Unfortunately, this has caused centralization and puts the network at risk. By increasing the minimum commission, the validators, delegators, and the network can benefit:

  • Validators will be able to afford more performant hardware, increasing their uptime and decreasing latency between nodes on the network.
  • Delegators who stake with more performant validators will receive more rewards compared to those staking with validators that miss many checkpoints (and therefore do not receive that reward).
  • It can encourage delegators to look at more than just commission when choosing a validator, thus making the network more decentralized.
  • With a higher average uptime among validators, the network stability will improve, giving all users a better experience when sending transactions over Polygon.

Motivation: The top 10 validators have received 82% of the total delegation. Out of those validators, only the Binance node has a commission greater than 4%, with 7 out of 10 having a 0% commission. These low commissions incentivize additional delegations, which further centralizes the network. 0% commission also removes incentive for validators to keep a high uptime, which further threatens network security.
Increasing the minimum commission charged will give validators more incentive to utilize professional hardware and maintain the highest possible uptime. Increasing the minimum commission will also give delegators more nodes to choose from at the same reward level. This will certainly help to decentralize the network. A decentralized network is more secure, leading to safety of funds for both delegators and the Polygon network as a whole. Properly funding validators will allow nodes to invest in performant hardware, which will allow a greater uptime percentage. With greater uptime percentage, fewer checkpoints will be missed and that will translate into more block rewards for delegators.

Specification: Currently, there is only a MAX_COMMISSION_RATE set to 100 in the Matic Stake Manager contract. A new contract would be required to include a MIN_COMMISSION_RATE, possibly as a configurable variable for easy update in the future. In the below code, this MIN_COMMISSION_RATE is created, along with the code enforcing the new minimum.

uint256 constant MIN_COMMISSION_RATE = 5;

uint256 constant MAX_COMMISSION_RATE = 100;

...

function updateCommissionRate(uint256 validatorId, uint256 newCommissionRate) external {

...

require(newCommissionRate >= MIN_COMMISSION_RATE, "Incorrect value");

require(newCommissionRate <= MAX_COMMISSION_RATE, "Incorrect value");

validators[validatorId].commissionRate = newCommissionRate;

}

A minimum of a 5% commission was selected as enough room for the average validator to maintain their servers, but not too much that it starts to cut into delegators rewards. This is also in line with many other PoS blockchains, where 0% commission is not as common. Tezos’ bakers average 8-15%, Harmony One validators average 2.5-10%, Persistence One validators average 5-10%, and Solana validators average 10%, to name a few.

Rationale: A common misconception about the Polygon network is that staking rewards are distributed when a validator creates a block or sends a checkpoint, and delegators will get more rewards by staking with larger validators. This isn’t completely accurate, as any validator that is online to validate a checkpoint will receive rewards. Rewards are sent to validators and delegators based on their total stake in the network. Therefore, outside of commission, the only variables in rewards are how much a delegator stake, and the uptime of the validator. Not only does staking with smaller validators provide the same rewards, it encourages better decentralization of the network.

Other options explored:

  • Putting a max cap on each validator’s total stake
  • Decreasing rewards if a validator’s total stake surpasses a threshold
  • Requiring a validator to stake a percentage of the total stake on their node, which incentivizes good behavior among larger validators

Backwards Compatibility: There are no backwards compatibility issues in terms of functionality, as having a validator change their commission is already built in. However, the current contract would not be usable as there is no way to enforce a minimum commission, so a contract would need to be deployed.

Security Considerations: Although there is not much risk in implementing a minimum commission, any modification or new deployment of a contract inherits risks of bugs. Issues with the updateCommissionRate should only be limited to allowing validators to choose any number for their commission, and should not extend into other parts of the contract or Polygon ecosystem.

FAQ:

1. Any alternatives considered? Yes, we looked at other blockchains to see how they handle the decentralized approach; if we could rebalance delegators across the validators, commission would have a higher impact for lower-staked validators.

a. Tezos requires a minimum amount of self-stake that corresponds to the total-stake (a validator would need 1 token for every 10 total-staked tokens). Validators would need to put up extra capital or start turning delegators away.
b. Cardano starts reducing rewards to pools over a ‘saturation point’, thus punishing pools with too much stake.
c. Harmony caps the maximum stake of the validator.

Will this take away too much from the delegators? It shouldn’t, as this means delegators will still receive 95% of their current rewards. For example, using today’s reward of 13.7% APR, a user delegated to a 0% commission validator with 10,000 Matic and receives 1,370 Matic per year. If the validator changes to 5% commission, they will then receive 1,301.5 Matic per year; a difference of 68.5 Matic, or 0.685% of their total delegation. To note, many delegators have chosen a 5% commission or higher validator, in which case this PIP will not have any impact on their rewards.

7 Likes

Great post @delroy, this is a super interesting topic to be solved. For all PoS chains, including of course ETH 2.0. We know this has been an issue also for PoW networks where large miners have leveraged their income to take more and more of the hash power. For a while there was a real risk of chinese miners to controll the network fully (if coordinated of course).

I feel this centralization is almost more built into the PoS infrastructure since it leverages the onramps (CEX) centralization of new users. At least until we onboard enough users ‘on chain’ and even then most will gravitate towards the largest actors. Given that Binance is taking a fee and still users won’t move their funds to other higher yielding validators.

Therefor I wonder if this will have a large impact on the distribution of funds or if we would simply be putting more money in the large validators pockets?

I haven’t found any data on how this looks on the Beacon Chain, I assume it would have the same problems, many validators but a large differance in staked amount?

Looking at your other examples, the Cardano approach is the one I like the most since it should have the least impact on market mechanics, it would just be lowering yield for the actors going over that amount. Of course given that Binance users should have lower yield than the 0% comission ones maybe its not enough? I guess it depends on how steep the reduction of rewards are. The ADA staking pools looks to all keep clear of that maximum amount so it might be quite a large slash.

3 Likes

I like this approach and I’m gonna add something more.
My opinion is that I believe that looking to push decentralization, the first objective should be “uncap” the 100 validators limit and set a goal to reach, like 500 in 6 months?
Second one and related to the first would be move from the current 26% to 35/40% tokens staked.

How do we go there, if changing minimum rate may be a way with little impact on the validators business, why not looking also at something that could impact in short term but will be beneficial for everyone on long term? Maybe longering the minimum lock, reducing the % of rewards but stack up a % incentive up to a capped %,on keeping the tokens locked?

Just thinking out loud but would love to discuss this.

3 Likes

The biggest reservation I have when it comes to promoting Polygon is the total number of validators. This is a core metric used to compare different blockchains (even if it is not the best metric). If we can increase the number of validators and at the same time incentivise higher performance validators nodes, then this is a net positive for the network.

4 Likes

@AlexDeFiCoach Yeah this is also a promising approach

2 Likes

Matrix Stake is bringing more than a decade of cloud experience and combining the knowledge and experience to build a robust validator node ecosystem at cloud scale and :100: up to support to empower decentralization.

Currently providing professional staking service for Polygon (Half a Million+ $MATIC), Harmony ( 23 Million+ $ONE) and Velas (5 Million+ $VLX).

We have already seen this great initiative on Harmony and Velas and seen tremendous growth in terms of validator counts with evenly stake thus empowering decentralization and would like to fully support the proposal.

Let’s Go! Polygon to the moon. Thanks @delroy for sharing.

4 Likes

Hi all, just want introduce myself as you’ll see me commenting and replying from time to time. I am one of the validators that helped write the PIP, in coordination with other validators, community members, and the team. Looking forward to everyones’ feedback!

Hi Tommy! I’m glad you find it interesting, I certainly did as well. It’s certainly not unique for Polygon; all PoS chains have to tackle the centralization problem. While it will certainly give the large validators who have built a following off 0% commission the bonus of now having a higher commission, I think it stops the bleeding of these validators being the magnets of delegators going forward. The goal being to give delegators a reason to look towards smaller validators as they won’t be losing rewards by not picking a larger 0% validator.

I have not found much on the Beacon Chain as well, but we share the same assumption. I’d be interested if others have better data.

I like Cardano’s approach as well, but I don’t know if it works as well on a chain with a capped number of validators with no replacement mechanism at the moment. Any number of scenarios could happen, from Binance redelegating their stake above the cap across multiple validators, or cartels/bidding to start to gain favor with Binance for their stake. I figured just by leveling the playing field and implementing a minimum commission, it doesn’t disrupt the economics too radically to cause uncertainty.

Thanks for the feedback!

1 Like

At first, I was leary about increasing the validators as I know how expensive it can be, but the more I think about it, the more I support it. WIth a minimum commission and maybe some delegation incentives from the team, there’s no reason we can’t add more. My thought would be to increase the cap about 25 at a time, and see what effects (positive or negative) it has on the network. As long as it’s not negative, keep upping the cap. Just my 2 cents.

Neat thoughts. I wonder how hard it would be to implement a locking mechanism or a variable APY. I see more networks going towards liquid staking rather than locked staking, so while locked staking has it’s place, I’d argue a more liquid approach could incentivize people parking their Matic in delegations for even short-term gains before they decide to part with their Matic. Incorporating the staking mechanism on the Polygon chain might also go along way.

Great thoughts!

2 Likes

I agree, removing the cap of validators (or increasing it) would be almost needed to use the Cardano approach. I’m however not sure why you think people would move away from the large validators if they increased their fee to the minimum fee? I think the data is showing people stake where convinient first, usually this would be at the large validators that can afford a smooth experience and if its also a CEX a natural on-ramp.

1 Like

Sorry, probably didn’t word it correctly. I don’t think many will switch from a larger validator to a smaller one. I think leveling the playing field by having both large and small validators having the same commission, along with the new sorting algorithm on the website that puts smaller validators on top, will influence new delegators to stake with lower validators.

1 Like

I agree with Tommy that increasing the minimum wouldnt actually make people move away from large validators. And alternate could be minimum fees if you have more than some X% of staking power.
Then people are now incentivized to delegate to nodes that do not have large staking power and by playing around what is X , we can improve decentalization

I care a lot about decentralization and this is what I have been proposing to a few folks. So while saying we want more validators is easy, if we do it we become slow and transactions become costly. I think we can increase the validator pool. And then say every Y days 100 validators get picked randomly. So total can be 1000 or 10000 validators. and as the total validators go up, the Y can come down giving everyone frequent opportunities to become validators. At that point, even the delegation becomes more even as now every Y days delegators need to delegate to new people. I believe this is how polkadot does it as well. So the idea has been tested elsewhere.

And back to raising minimum- don’t think that helps

4 Likes