Renegotiate the deal with Stride

Hub Proposals Proposal Ideas
1

Stride is using the Hub’s security for a pittance. For example, Coinbase Custody and its 20% commission yielded less than 150$ in Stride-derived commission between Aug 1 and Sep 1. That’s the Hub’s biggest validator with a way above average commission that isn’t even making enough commission to pay for equipment, let alone man power. Small validators are essentially indentured servants of Stride getting effectively nothing.

I propose that Stride should at least match Neutron’s payments to validators, which isn’t that much either, but can at least cover equipment costs, or we should stop providing ICS to Stride.

Big and small validators please chime in on this. Do you like having your expertise/time/effort valued so little?

8 Likes
2

This is an interesting commentary. I would love to know the extra cost associated with validating stride (or any ICS chain) in addition to cosmoshub. If it is a predictable increase, this should probably factor into ICS discussions, and I wonder how that would integrate with permissionless ICS.

3

Last I checked, Stride was paying the Cosmos Hub ~$1k/day, while Neutron was paying the Hub $10/day.

Did the economics of Neutron’s deal change?

1 Like
4

This is correct Aidan, what @pooch might be referring to is the idea that we suggested last year to use part of the unclaimed NTRN from the airdrop and sent to the Cosmos Hub community pool as reward for validators which was finally implemented. Stride has some of the smartest and most talented people in the Cosmos ecosystem, they already identified for example the opportunity for LSTs for the upcoming Namada MASP with the shielded set rewards. Stride has achieved a lot and the Cosmos Hub should feel proud and thankful to have Stride as a consumer chain

5

Ah that’s interesting I missed this - do you have a link to this proposal / the implementation? Would like to read up on it

6

Sure here is the proposal: https://forum.cosmos.network/t/approved-cosmos-hub-neutron-validator-alignment

7

Neutron has a profitability issue, but is different in 2 aspects:

  • indeed they “airdropped” NTRN to the validators to help them cover their expenses
  • they aren’t really generating any sizeable revenue yet, but once they do it will be distributed to everyone and not only to a subset of the validators.

We have zero ATOM delegated from Stride (if the only wallet is cosmos10uxaa5gkxpeungu2c9qswx035v6t3r24w6v2r6dxd858rq2mzknqj8ru28, which I believe it is), and we get approximately 0.30 STRD per week in rewards plus dust in other stTokens.

And yet we are compelled to run a node, with all the responsibilities and risks that it entails and without seeing much of the $1k/day that are paid to the Hub – it certainly does not cover our infrastructure expenses.

Stride has had a very positive impact on the ecosystem, that’s not debatable. But perhaps a change in the economic model would be in order so that it is more fair and balanced?

For example, delegating more widely (if a validator must run a node because they’re in the top n-95, then they should also benefit from the liquid staking delegations); alternatively switching to PSS could be an option so that validators can opt out as they see fit.

6 Likes
8

On behalf of the PRO Delegators’ validator, we fully support the idea of renegotiating the Interchain Security (ICS) agreement. There are clear inefficiencies in the current arrangement, and both Stride and the Hub’s validators are struggling as a result. For example, we noted some operators in the set are failing to push Stride upgrades in a timely manner, potentially due to the lack of profitability as opposed to infrastructure costs. This suggests that there may be systemic issues with the agreement between Stride and the Hub that need to be addressed.

Additionally, we have been granted by the AADAO the opportunity to conduct a thorough analysis of Partial Set Security (PSS), a key upgrade in ICS 2.0. You can find the full details of our analysis here:

Stride, as one of the existing consumer chains from the earlier ICS 1.0 iteration, has been migrated to a Top-N 95% model, maintaining its previous position. However, we believe that this could be significantly improved by renegotiating the terms under the more flexible ICS 2.0 framework. While we recommend retaining the Top-N model for the core public good mission delivered by Stride, the chain would gain massively from making use of the exclusion lists for certain validators, and use a votePower cap could offer significantly better reward distribution. These are just one of many examples that can be explored depending on the desired vision. We are eager to share our insights with the @Stride team and @Riley or any other strategic representatives.

Please feel free to reach out to us at contact@pro-delegators.com to schedule a meeting and explore potential paths to better align both parties.

Thanks for reading,
Govmos.
pro-delegators-sign

7 Likes
9

Hi, @aidan. Thanks for the reply.

Please show your work on the 1k$/day.

Please don’t get me wrong: Stride is great for the ecosystem, there is no doubt about that.

I wholeheartedly agree with @Govmos in that Stride itself could make things right simply by revisiting their delegation strategy (across all chains). Here’s a model that could work just via a script and an account that’s authorized to delegate and redelegate:

  • Never delegate to a validator that has been slashed, ever.
  • Only delegate if a validator has been up for 3 months or more.
  • Use a quadratic function to determine the delegations to the top 95% excluding those eliminated by the constraints above.
    • y = ax^2 + bx + c; quadratic function
    • b = 0; keep it simple
    • c = amount of ATOM to delegate evenly to eligible validators
    • z = total ATOM to delegate
    • z = integral y dx from 0 to 0.95 = a * 0.285792 + c * 0.95
    • c = 0.1z; evenly spread 10% of total ATOM to delegate
    • a = ( total ATOM to delegate * ( 1 - 0.1*0.95 ) ) / 0.285792
  • Rebalance all the delegations 24 hours after the last rebalance.

Notice that there is no subjectivity in the delegations and they’re derived from on-chain data, other than the parameter c. c is arbitrary and can be tweaked to Stride’s content; the lower c is the more the distribution favors small validators.

With a script and cron there’s no need for a smart contract or manual action or a multisig. It’s simple. Plus, the quadratic delegation function would massively help with decentralization of voting power. Using a script and cron would also get around the 32 validator limitation of a smart contract, if that still exists.

I’ll happily take responsibility for making sure that the script is executed each day on The Hub and Osmosis while our validator remains active in both sets.

10

I really think ICS 2.0/PSS solve a lot of these issues. There’s no optionality currently, and I think it creates a lot of animosity for validators of both Neutron and Stride. The sentiment i’ve heard from some validators and people involved in these projects seems to be, for lack of a better word, rough due to incentive misalignment.

Who wants to do complex upgrades on a chain that’s barely paying you for work?

ICS V1 is the culprit here. Let validators opt in and out, and explore other revenue models.

11

quadratic
quadratic923×633 22.6 KB

The delegations would be recalculated on every rebalance, ie every day, so DragonStake would get a big bump up on day 1, then would drift back down day after day as long as they remain in the 95%, and then get a big bump again when they’re the last of the 95%. Coinbase would remain largely unaffected since they’re #1 and would just have their portion of the evenly distributed ATOM staked to them.

12

ICS V1 is the culprit here. Let validators opt in and out, and explore other revenue models.

Totally. The back-of-the-envelope calculations leading to the 95% cut-off had very bad assumptions in them. (No offense the author of the calculations, whom I greatly respect. At the time, I didn’t pay attention to the forum enough to voice my real-world validator economic experience.)

1 Like
13

if i understand you correctly there would be one issue with this approach - re-delegations are still subject to unbonding period cool-down
but anyway, even having that executed every 21 day is much better that what we see now, thumbs up for raising that discussion

1 Like
14

Redelegations are not subject to the unbonding period, they’re instantaneous. They are potentially constrained by the max entries parameter, but I think that just requires a little business logic in the script.

15

@aidan , does the 32 validator limit for Stride still exist?

16

Hey all! Interesting discussion. I want to give a few counterpoints from the consumer-chain perspective. (Note: This is my personal opinion as a long-time Cosmos community member, and not as an employee of Stride Labs. My employers may / probably do have a different opinion)

I think looking at ways for Stride and the Hub’s valset to be more economically-aligned makes a ton of sense but, as I’ve said many times in the past, this needs to be handled both from the consumer-side and the provider-side of the equation, not just from the consumer-side.

Over the last year Stride has averaged between $60,000 and $600,000 in ARR for the Cosmos Hub per Defillama, with an average cost to run the Stride chain estimated at $500,000 annually (node costs). Stride has been profitable for the Hub in the past, although it is not profitable currently due to the abysmal price action of tokens throughout the ecosystem that make up Stride’s revenue. This revenue is distributed to all validators that have opted in to secure Stride, contrary to what the following message states:

What Can the Hub Do to Support its Consumers?

Despite being at an all time revenue low due to ecosystem price action Stride is still, by far, the largest source of revenue for the Cosmos Hub. Stride has a very clear path toward profitability for the Cosmos Hub from ICS fee revenue (and has achieved this in the past). As a consumer chain and from a sustainability perspective, it’s doing what it’s supposed to be doing.

So I’d also ask if there are any things that the Cosmos Hub can do to be a more sustainable security provisioner and avoid placing an undue economic strain on both consumer chains and on validators. IMO there are two things that can be done:

  1. Distribute ICS fee revenues equally to every validator that has opted in to secure a given consumer chain rather than proportionally according to stake-weight. If all validators are paying an equal cost to secure a Top-N% chain, those validators should share equally in the revenue so long as a consumer chain isn’t profitable.

  2. Reduce the size of the validator set. Obviously this option is a bit more controversial. The reason that smaller validators are struggling so much to be profitable validating the Hub is that the set is too large. It’s unreasonable to ask consumer chains to bear the economic weight of the Hub’s unsustainable cost / revenue structure. IMO the set could be decreased by as much as 30-50 validators (or more), which would alleviate a lot of the unsustainable costs here.

Note that both of these are actions that could be taken by the Cosmos Hub to make ICS a more sustainable offering without placing an undue burden on consumer chains, which may have the effect of deterring potential consumers or driving them to competing security solutions like Symbiotic or Eigenlayer / Ethos. Asking every prospective consumer chain to airdrop tokens to validators in an incredibly competitive security market in which (to be frank) the Hub is an underdog, is simply not going to be a sustainable path forward.

The first of these feels more reasonable / less controversial, and I’d invite someone in the Cosmos Hub community to put a proposal on the forums to this effect. I’d be happy to support as needed, as I feel this is a long overdue change. Though I would also strongly encourage the Hub to consider some version of the second option as well.

What Can Stride Do to Support The Hub?

It’s been mentioned here already, but Stride has another path to driving revenue toward Cosmos Hub validators via its delegation program. In the past, Stride’s validator set has been limited to 32 validators because of limitations on the gas that can be accepted by the Hub for Stride’s delegation messages. This limitation is sensitive, and changes in gas required for transactions on the Hub have resulted in issues in the past.

Thankfully, with Stride’s latest chain upgrade, executed just last week, this gas limitation is no longer an issue and Stride can now have arbitrarily large delegation sets, meaning we can delegate to pretty much the entire set.

An update to the delegation program is in process, but there are a few things that need to be worked out, namely:

  • LSM / Validator bond: Stride won’t be able to delegate to validators that don’t validator bond enough ATOM to accept LSM delegations. This is a Cosmos Hub limitation, not a Stride limitation.
  • Opt-in PSS: If Stride delegates to validators that don’t validate all prospective PSS chains, Stride LST holders will have a lower yield, making it a less attractive product. A good delegation program that includes this consideration could actually act as an incentive for more validators to opt-in to securing PSS chains, which feels like an ideal activity to incentivize from the Hub’s perspective.
  • Upcoming changes like the proposed Vote Power Tax: Less of an issue now but could become relevant in the future

What’s currently being considered is an interim solution that will put stake in the hands of as many validators as is technically feasible, allowing for more validators to earn commissions from stATOM inflows. This can then potentially be followed by a more comprehensive delegation program adjustment that takes into account the above considerations. This will all of course need to be approved by Stride governance first, and I’d expect that we’ll see a proposal on the Stride forums in the near future addressing this.

3 Likes
17

My message was admittedly a bit unclear, but I was referring to the revenue stemming from the LSM delegations, which are going to the whitelisted validators.

As I mentioned just below the part you quoted, what we get from that revenue sharing is extremely low and even when the markets pick up it will not cover our operating costs, by far (and we’re #49 in the active set right now, so not a top validator but not insignificant either).

Apart from that I agree with your position and ideas.

That latest upgrade allowing Stride to delegate to more validators might be a path to mitigating the issue, but indeed it would probably need to have other conditions than just opting in to avoid spreading the LSM delegations too thin.

2 Likes
18

The opposite actually. Stride attracted a certain amount of ATOM for liquid staking but still a small %. If Stride would spread these delegations across many validators rather than a smaller group then the revenue for each would be meaningless given the tiny delegations, as is the case with pstake for example. In addition, there are some complaints as if Stride chose a few validators randomly themselves when actually there was a robust selection process to which any validator could apply and then an evaluation that lasted months with governance proposals on Cosmos Hub and Stride. In summary, Stride chose the top validators of the Cosmos Hub by performance, contributions and other metrics guaranteeing that the ATOM deposited on Stride for liquid staking have the lowest possible risk for slashing since they are spread within the top subset of Cosmos Hub validators. Given than Stride is a consumer chain of the Cosmos hub it is its responsibility and job to select the best subset of validators with a robust selection process rather than a lazy spread to the whole set. This is also because for other chains like dYdX Stride constantly used the argument of being a consumer chain benefitting from the security of the Cosmos Hub as a reason to allow Stride to grow its TVL significantly so for Stride amongst all chains in which they provide liquid staking, the Cosmos Hub is the most important since Stride’s TVL growth 100% depends on having the security backing of the Cosmos Hub and hence Stride has maximum responsibility towards the Cosmos Hub

1 Like
19

stride stakes to atom validators per cahortas per time. its the bonus youre talking about imo

20

I disagree.
Stride stakes over 4M $ATOM. That isn’t a small amount, and it could be spread more evenly.
The slashing risk… exists anyway. It’s kind of a foundational element of PoS.
A penalty would probably be in order when validators get jailed.

(Disclaimer: we have been jailed on Cosmos in October 2023, not because of a downtime but we got into the top n-95, didn’t notice, and we hadn’t opted into Neutron :unamused:
Still, that was our mistake and we own it and its consequences.)

On a side note, I see that Stride now delegates 28k atoms to us, and different amounts to 93 validators in total.
It’s something and I appreciate the gesture.

2 Likes