Akash Proposal #26

Disputing the merits of a 5% commission.
September 12, 2022

Intro

On September 10, akash1ajdqcckwn2493jdlpqqde5uerq39z00a0yweqj (a new wallet), submitted Proposal #26 on Akash. The proposal aims to increase the minimum validator commission rate to 5%, meaning all validators charging less than 5% would have to raise their commissions to at least 5%. This is a text proposal, and therefore does not effect on-chain parameters, but the spirit of the proposal should be executed if it passes. This style of proposal is live on other Cosmos SDK chains, like Osmosis and Juno, though was recently rejected by the Cosmos Hub. Moultrie Audits will vote No on this proposal. We've attached the full proposal below and will respond line by line.

Full Proposal

"0% or low commissions are known to be bad for the health of the network. By setting up a minimum commission rate, it will help to build a healthier network environment and promote decentralization of Akash network. 5% has been accepted by a wide range of Cosmos community in Osmosis, Juno, Chihuahua, to name a few. If this proposal is passed, the software of Akash network will be needed to be updated so that commission of all validators lower than 5% will be increased to the minimum automatically. Validators with commission rate lower than 5% should increase their commission rate to 5% as a gentlemen agreemnt before the required network upgrade."

Refutation

This is a disappointing, lazy proposal. If the effect is to reduce validator sovereignty, surely the issue demands more than 5 sentences. Regardless, we'll proceed.

• Claim: "0% or low commissions are known to be bad for the health of the network." Response: 0% commissions are not *known* to be bad for network health. Significant discussion exists around this topic, of which the proposer presents none, and there are undoubtedly theoretical positives and negatives about 0% commissions. To blanketly characterize them as bad is entirely disingenuous.
• Claim: "By setting up a minimum commission rate, it will help to build a healthier network environment and promote decentralization of Akash network." Response: Here are the number of validators with a 0% commission broken down by groups of 10, according to voting power (greatest to least)
• 0-10: 1 (10%)
• 11-20: 0 (0%)
• 21-30: 0 (0%)
• 31-40: 0 (0%)
• 41-50: 2 (20%)
• 51-60: 0 (0%)
• 61-70: 0 (0%)
• 71-80: 0 (0%)
• 81-90: 4 (40%)
• 91-100: 0 (0%)
• As you can see, there are only 7 (7%) of active validators with a 0% commission, the majority of them are in the bottom 50, and 6/7 (85%) of them are outside of the Nakamoto coefficient. This indicates that the majority of 0% commission users are smaller validators attempting to incentivize delegations to increase decentralization. Now let's examine median (to account for 100% commission rates) commission rates, along with the associate ranking from least (lowest median commission) to greatest (highest median commission).
• 0-10: 3% (2)
• 11-20: 6.5% (5)
• 21-30: 6.85% (7)
• 31-40: 9.9% (10)
• 41-50: 6.94% (8)
• 51-60: 4% (3)
• 61-70: 4.5% (4)
• 71-80: 9.25% (9)
• 81-90: 1.5% (1)
• 91-100: 6.75% (6)
• The group with the lowest median commission is also 81-100. More importantly, the composite ranking score of the bottom half of the validator set is 26, while the composite score of the upper half is 32. This means that the bottom half of validators, based on deciles, offer lower median commissions. While this data is certainly flawed, it provides evidence against the theory that raising minimum commission rates is known to increase decentralization. One must ask, by what metric? There are many other reasons to delegate to validators outside of their commission rates, such as: security, uptime, contributions, responsiveness, marketing, etc. If the commission rate was the most important metric, we would expect those 4 validators in the 81-90 section to shoot to the top of the validator set, capturing redelegations from the 9 validators in the top 10 with a commission rate above 0%. Clutch Staking was created 14 months ago with a 0% commission, and they are still only ranked 50. There simply is not substantial evidence indicating that 0% commission rates harm decentralization or are "bad for the health of the network."
• Claim: "5% has been accepted by a wide range of Cosmos community in Osmosis, Juno, Chihuahua, to name a few." Response: While a few chains have implemented it, it is by no means ubiquitous, as evidenced by the Hub's rejection.
• Claim: "If this proposal is passed, the software of Akash network will be needed to be updated so that commission of all validators lower than 5% will be increased to the minimum automatically. Validators with commission rate lower than 5% should increase their commission rate to 5% as a gentlemen agreemnt before the required network upgrade." Response: This is more evidence of the exceptional laziness of this proposal. There is no offering of a repo or PR where this is completed by the submitter and contains basic spelling errors.