# Time to bring back decentralization to NEAR validators ## Problem - There are only `9` validators that control `>33%` of the stake. - Delegators are not incentivized to delegate to other validators. There is no difference for delegators to stake into top pool with `1%` fee or the bottom validator with `1%` pool. - Bottom validators don't receive enough staking rewards to cover hardware costs. ### Background NEAR Protocol has `5%` annual inflation. `10%` of this inflation goes to the treasury currently controlled by NF. Validators split the remaining `90%` of the inflation. The reward per validator is based on the their performance in the current epoch and the stake amount. If all validators perform at `100%`, then the inflation for an epoch is split proportionally to the validators stake. Currently, the bottom validator has a stake of `26,571` NEAR. With the APY at `8.35%` resulting in `2,218` NEAR in total rewards. With `10%` fee it's only `221` NEAR or `$230` a year. This is not enough to cover infrastructure cost. Which ranges from `$100` per month and hight (including backup nodes). So a validator needs to have at least `$1200` a year not considering working hours. ### Solution There is a very simple protocol change to validators rewards. We can take `5%-10%` from the current `5%` inflation (about `2.6M` NEAR a year) and distributed it equally divided across all active validators (can be scaled down for some based on the epoch performance). With the current amount of active validators `213`. Each validator will receive about `12.2K` NEAR per year in additional rewards. That will result in additional `46%` APY for the bottom validator. And with `10%` fee should be enough to cover infra costs. Now the validators pools will have variable APY. The bottom validators APY will be higher and it will incentivize delegators to delegate to validators on the bottom. Over time it will shift stake to be more uniformly distributed. ### Caveats - Top validators will lose about `4%` of the rewards (at `10M` stake), resulting in `0.334%` loss of APY. - The `5%-10%` coefficient can later be updated to by dynamic based on the number validators in the `33%` ### Links - Previous discussion: https://gov.near.org/t/research-alternative-rewards-models-for-validators/32508 - Near Staking: https://near-staking.com/ - Explorer Stats: https://explorer.near.org/stats