[Urgent] Aurora Validator, Decentralisation Concerns

[satojandro]

tldr;
Aurora (network) is launching the Aurora Validator with an aggressive $Aurora rewards scheme that threatens to destabilise, weaken and potentially take over the network. These concerns have been acknowledged by the Aurora team and dismissed as it is their position that “it is the NEAR Foundation who has to ensure decentralisation, not us.” This is an urgent thread to get community input on the issue and explore alternatives.

Overview

• NEAR just released an update that enables individual validators to offer Staking Rewards (additional rewards on top of the standard ±10% $NEAR). Learn more
• Aurora is the first validator to offer additional staking rewards. Initial Blog Post
• The original post on Aurora’s governance forum contains a lot of details about how the validator will work and calculations about income fees, etc. However, they notably do not include ANY calculation or risk assessment to show they properly considered the risk of decentralisation.
• @alex.shevchenko recently posted a Twitter Thread clarifying some of the most common questions. Again, zero mention to any risks or considerations in regards to the effect Aurora validators can have on decentralisation.

Key Figures - Aurora Validator

• The Aurora validator is being deliberately set up with an aggressive rewards schedule that designed to reach 80m NEAR

Key Figures - Staking Ecosystem

Validators Stats Page on explorer

• There are currently 324m NEAR being staked across 73 validators
• The top validator has 42m NEAR
• Top 8 validators have 33% of delegated NEAR (107m NEAR combined)
• We explicitly tell stakers that delegating to delegators other than the top 8 improves the decentralisation of the network

Existential Threat to NEAR

I’m going to be blunt as not many people seem to be paying attention: the setup described above is a recipe for disaster which creates many risks, challenges and outright existential threats to the network.

Given the current setup, I believe it is reasonable to assume Aurora validator reaches at least 80M NEAR. This has many potential implications that all need to be considered carefully:

• 80M NEAR (Aurora’s target) represents close to 25% of the Network, dangerously close to being able to overpower every single other validator.
• Where are the 80M NEAR coming from? It is safe to assume that most of the NEAR will simply be reallocated from existing validators. This in turn means that there is a real risk that many validators will not have enough NEAR to keep their seat - now we have one Master Node with 25% of the Network AND fewer and fewer validators.
• Having ALL the incentives of the ONE Aurora Validator (‘the Master Node’) with 25% of the stake tied to a different coins presents all sort of nightmarish scenarios. What happens if the price of Aurora dumps? (Very few coins in circulation now, very little use for the coin). Or what happens if the price of Aurora gigapumps (EVM growing much faster) - then the existing rewards would be enough to take over more than 33% of network. Normally, the ups and downs in the market are shared across all validators and their stakes. But creating a divergence between the economic reality of NEAR node stakers and Aurora rewards stakers creates all sorts of problematic scenarios. I strongly believe that the Validator Staking program should be on top of NEAR rewards (baseline + bonus) to mitigate these scenarios.
• Even if we take corrective measures after we realise that the Aurora validator is disproportionally large, I fear some of the damage is permanent and hard to compensate for (including repetitional damage). Why should we wait until after the entirely predictable problem eventuates to think of solutions and alternatives.
• Market Perception - decentralisation, specifically lack thereof, is the knife that will be used to attack us. We have seen how many chains have come and gone, blockchain project’s do not last if they are controlled by a handful of people or if the network is not secure.
• I hate to say this, but there is also the reality of Aurora and NEAR pathways and interests diverging at some point. Even though common sense would suggest Aurora needs NEAR as a security layer, some of the behaviour and messaging I’ve seen on the Aurora side sets off the hostile takeover vibes.
• Finally, and the reason why I am bringing this debate to the NEAR Community, the Aurora team do not currently think it is their responsibility to ensure the decentralisation of NEAR validators (even though they are the ones creating the threat).

Proposals
While there may be some disagreement on some of the points above and the assumptions made, I believe that there is a range of changes and simple initiatives that could be implemented to prevent any of the risks above while enabling Aurora to operate their validators and attain their desired goals.

Suggestions to Aurora Team

1. Reconsider the ways in which the $Aurora tokens that have been committed to the NEAR Ecosystem are distributed. Rather than having 5M solely to the Aurora Node:
   1.1 Allocate $Aurora tokens to DEXs (I also note that $Aurora rewards on REF have run out and the Aurora team has not replied to the request to extend them yet)
   1.2 Consider allocating a chunk of $Aurora tokens to other validators. This could be done through Meta Pool; have Metapool distribute $Aurora to stakers in the same way they are currently distributing $META.
2. Recalculate the Rewards Structure so that the upper limit at which $Aurora rewards are still higher than traditional validators is lower (i.e 60M, perhaps 70m max., down from the current 80m). This could be set as a % of total NEAR staked and could be promoted as a standard - no single node should have more than x%

To the NEAR Foundation

1. Increase the amount of NEAR allocated to Metapool from the current 1.5m, this seems to me like the easiest and fastest way to ensure that the network remains decentralised (Metapool rebalances to all validators outside of top 8, based on performance).
2. Consider hard-coding a delegation limit per validator to ensure no one validators can grow enough to overtake the network.
3. Liaise with Aurora team and be vocal about the need to preserve decentralisation. Collaboration over cannibalisation.

To the Community

1. Restrain from delegating to Aurora Validator until some of the issues above have been addressed
2. Join the conversation here, on the Aurora Governance forum, and anywhere you feel you can communicate your views and concerns to others
3. If you do want to capitalise on early APY of Aurora validator, consider distributing your stake across at least two validators (Aurora + 1 out of the top 8, Metapool probably easiest).

@marieke.flament @illia

[SanketN81]

@satojandro

Consider allocating a chunk of $Aurora tokens to other validators. This could be done through Meta Pool; have Metapool distribute $Aurora to stakers in the same way they are currently distributing $META.

This is really interesting, but not sure how it’ll be implemented as Aurora validator takes 100% of Fees earned in NEAR & instead provide users with Aurora tokens, it won’t be possible in the case of metapool as every validator will have different fees.
So, unless Aurora isn’t earning the majority portion of earned near in fees I don’t see the point in giving out Aurora tokens.

Solutions could be as u suggested to lower the rewards & make it as aurora validator won’t get x%+ of total staked.

or if Aurora Sets multiple validators & distribute rewards among those, but anyways those will be controlled by aurora only so not sure if that makes sense.

I Hope they find a solution🤞

[erik.near]

The direction of this concern seems reasonable since the landscape of validator competition is changed by the introduction of these new reward types on the new contracts, which can create some more defi-like incentive-driven dynamics for attracting delegation to validators.

The use of smart contract staking is a cool feature of NEAR which is going to keep creating innovative new staking reward types so we have to accept overall that this will keep evolving. Validators will need to acknowledge that their competitive environment will change more rapidly than in other protocols as they keep innovating to attract stake.

On the other side, it is great as a delegator to have more options for how to delegate, which will help bring more people into the NEAR ecosystem and raise its overall robustness and security.

From a systemic perspective, it’s right we should make sure one contract doesn’t absorb too much of the stake since it can have undesirable voting power and, in the extreme, can create governance attack vectors by using short term incentives to harvest stake.

In the short term, I don’t expect this will happen with the Aurora contract because there are some additional risks people take by joining this contract (iirc it’s not yet audited, and it has price risk between AUR and NEAR), so it’s not a “slam dunk” to participate as a delegator.

In the longer term, competition is the best thing to help reduce the market power of any one validator using these incentives. So the more validators we can help to build innovative new mechanisms, the more choice delegators have and the less market power any one will have. So IMO the best thing we can do overall is help attract viable alternatives and to highlight what tradeoffs delegators are making (eg in UIs for staking) between those options.

A core challenge is that some validators don’t have the same access to additional token incentives (since Aurora has their own token, they can ratchet up the rewards rate quite attractively) while others might have to work with third party projects to provide additional incentives through their validators (eg via BD to find projects wanting to distribute their tokens to holders). IDK what innovation will be needed to innovate and compete, but I suspect they will find it.

I’m not sure the perfect solution here, but competition and clarity seem like the most important ingredients.

[hodlr]

I get the concern for decentralization, but Aurora is the first to put innovation forward into the community. I think the only thing they may have done “wrong” is too much hype for such a new system. Maybe they could do better with more of a Beta period.

Anyway, my thoughts are this:

1. Aurora concept of dual rewards is great, within reasonable output
2. Other validators should be able to contribute to the setup such that:
   A. Other validators can assist aurora nodes by validating NEAR & Aurora
   B. Provide further decentralization that encourages participation in multichain
   C. Help Metapool get exposure to multichain rewards
   D. In the future, allow all validators to configure strategic reward allocations for the ecosystem projects

My main point here is: I want to see validators get competitive to further decentralize just as you hope, and want to see that happen because their strategy helps reward the right behaviours of the network & validation. Since DeFi APY is basically the only thing majority of users care about, it will always win at attention grab, why not help aurora tweak their strategy such that the entire validator network can help & capitalize on?

[Ragingbunyip]

Some further suggestions

• integrate Metapool with the web wallet and provide new users with Metapool staking as a visible default option. This will help Metapool grow and improve decentralization. It also eliminates a point in the onboarding experience where first time users are confused (they aren’t sure who to delegate to)

• diminish the marginal rewards for validators as they get more people staking with them. This would mean a single validators APY falls towards zero as it reaches a scale where it impacts network security. If not this, then some kind of mechanism baked into the game theory that stops any validator becoming too powerful.

• increase rewards for new/small (but high performing) validators. Effectively an infant entities subsidy? We need many many more, can we drop the entry barrier for becoming a validator?

[jack]

As a NEAR holder, I support innovations from Aurora that make NEAR even more valuable.