Use of Top of Block Auction revenue

Proposal Discussion
1

This forum post is a location to discuss any potential use of the Top of Block Auction revenue.

In Halve OSMO Inflation - #5 by JohnnyWyles I suggested that this be used as a direct, flat payment to support validator operations in the wake of lower inflationary emissions.

I present a proposal for this below, but want to discuss more before going on chain. Particularly:

  • Is this the best use of these funds? Is there another solution to the validator sustainability issue?
  • Should there be any exclusions to the allocation? Should being within the active set be the only requirement?
  • To increase the payment per validator with or without this, should the validator set be lowered from 150?
  • Should this be weighted by more than the 95/5 block proposer split currently in use?

This proposal would allocate the Top of Block Auction rewards directly to validators as an operational subsidy.

Top of Block Auction

The Top of Block auction uses Skip’s Block SDK to add a categorized mempool and was added in V25.

  • Users can submit a transaction bundle to a lane that takes place before all the other transactions in the block.
  • The first position within the block has value, confirming that the submitter will obtain any time-sensitive value, such as cross-chain arbitrage opportunities.
  • The transaction bundle with the highest bid in USDC will be accepted.

Revenue generated by the Top-of-Block Auction is split: 5% goes to the validator that proposed the block, and 95% is accumulated to a module address until governance determines usage.

Revenue

The typical accumulation in the module address for the last month has been around 500 USDC daily.

Revenue over time can be seen on this dashboard: Osmosis MEV Lane by Playwo.

Revenue is proportionate to the liquidity and volume on Osmosis, as well as the number of participants in the auctions, which increases competition. It represents arbitrage opportunities that require off-chain information, such as spot prices on other chains or centralized exchanges.

Distribution

This proposal would signal that revenue that was previously stored in the module address should be distributed evenly to all validators in the active set.

While this is a relatively small payment, equivalent to around 10 OSMO per day per validator, it offsets the 2,700 reduction proposed in Halve OSMO inflation by 1,500 OSMO. This distribution aims to improve the sustainability of validators’ operations further down the active set as the midpoint of the active set would receive around 10.8 OSMO of that 2,700 reduction per day.

This provides a sustainable funding method for validator operations as inflation continues to decrease towards zero. It is in addition to the 5% already allocated as a proposer bonus to the block producer and the commission-based returns from transaction fees and protocol fees.

Implementation

This distribution is currently parameterized with a proposer fee which goes to the block producer. Increasing this would bias the payment towards validators with more voting power.

This parameterization should be maintained, but the proportion accumulating within the module address should be distributed evenly between validators.

This mechanism should be implemented in a future software upgrade. Until then, any USDC accumulated within the module address is destined for the community pool and should be transferred at implementation.

2

Would definitely be in favor of this - as it’s the only way I see of proposing an on-chain incentive which allows validators to continue to pay for infrastructure despite lowering inflation.

Cosmos tying inflation commissions to validator payout has always been a recipe for disaster, as it inherently ties a fiscally infeasible compensation to validator operations. (i.e. the more we inflate the higher validator revenues are).

This would be unique in that it’s a real revenue stream which can grow with the ecosystem and ensure that validator costs are at least covered while inflation lowers sustainably.

It’s my opinion that we would do best at this time to distribute evenly throughout the set. Tying it in to validator position would defeat the purpose currently, and provided that distributions are done daily, churning in and out of the set would naturally cover any discrepancies. If you’re in the set for 1 day at the bottom, you get a piece of top of block to cover costs for that day.

If we even attempt to create a sliding scale or do anything other than an even distribution to all nodes, it gets into more gamification and ultimately means that costs will not be covered for node operators regardless of set position. One way to ensure we keep a healthy node distribution throughout is to have that even distribution and we preserve a decentralized set, lower inflation and ultimately have an alternative revenue stream that can support in the very least the hard costs of running a node - and hopefully one day a bit of profit that can actually assist in other costs.

1 Like
3

Should be lowered to 100 or so regardless. The sheer number of validators doesn’t really matter when the distribution of stake is so heavy on one end.

4

I really do not want to use top of block auction revenue for this.

Top of Block auction revenue captures all black swan events and CEX/DEX arb. Its critical that this is the revenue flow that goes to the community pool imo, not getting flat distributed.

Furthermore flat validator rewards don’t make that much sense at current validator counts, but also without effort in ensuring (even via social policies) that we have some level of non-sybilling going on

1 Like
5

Thanks for proposing this @JohnnyWyles !

We can already see from the few responses that this is a tough subject, with very diverse reasoning why we should or why we should not do this.
We have already tried a flat fee reward for validators in the past already, which didn’t pass at that time. However, with the conversation around lowering inflation it might become an interesting topic again. Did you already look back in that topic regarding pros and cons? if not, then I can try to find it.

@ValarDragon do you mean with capturing all black swan events that the revenue stream can better be utilized as some sort of contingency fund?
I do agree on the flat validator rewards, however… when the compensation is low enough parties might not want to take the effort in general, right? Maybe we can even dynamically cap the payout so that sybilling is not really interesting out of an economical perspective?

@maxpower the bottom 50 validators still account for a 6-7% VP. It is skewed to much, I agree to that. But it is still a relatively sizeable position.

@onivalidator maybe add something that a payout is only done for validators who are active for x days minimum in the active set? That way you can avoid validators switching in and out, which is bad for delegators potentially.

6

The Sybil issue remains present. My main concern about this methodology is whether this direct distribution becomes profitable in itself as time passes, which would be an incentive for sybiling to happen. If it remains at the level of a subsidy for operational costs then it wouldn’t be profitable to sybil.

I think we need to revisit validator payments as a concept. Since the direct inflation reduction may not happen until we work out the development vesting interaction, I feel like we should deal with both of these issues at the same time.

Rather than having arbitrary emissions, have a reserve that is fed by both inflation and the actual protocol revenue, with dedicated outflows to each requirement for the chain to function.

The initial model of inflation is meant to distribute the token to participants, but we are far closer to a situation in which that distribution has mostly happened, and participants need compensating for services provided while the token distribution continues to skew towards service providers.

7

Montagu from Citadel One here.

I like the idea. I think introducing and experimenting with sustainable revenue streams for Validators to substitute inflationary rewards is the way to go as Osmosis moves to reduce issuance. A couple of thoughts/ suggestions:

  • Excluding CEX validators from distribution might be a good idea, if technically possible.
    Note: afaikI, only two CEX validators currently exist ( Binance and Coinbase).
  • Add the Community Pool to the fee switch: This could be set to 0% at the beginning. It’ll allow for quicker implementation if the community wants to update these parameters in favour of the CP in the future (to disincentivise sybil behaviour for ex).
    I also wouldn’t be opposed to setting this parameter to 5% for starters so that the CP still capture some of this value while being tested.
8

But validators will need revenue outside of the inflation subsidy as it keeps going down, which is traditionally some form of MEV on other chains. What flow could be redirected to validators ?

1 Like
9

In essence inflation going down is not a bad thing if token price compensates for it. Look at for example BTC. Every 4 years the new emissions are going down, but the hash rate keeps going up and there are enough parties willing to operate on the chain.

So your question is 100% correct, what revenue streams should be directed purely to stakers, what streams to the CP, what streams to the validators, etc.