Automatic Staking of Osmosis Incentive Rewards

Proposal Discussion
1

This proposal would approve a change to Osmosis tokenomics during a future software upgrade that automatically stakes internal OSMO liquidity rewards to a user’s preferred validator set.

Background

Under the current tokenomic design, OSMO liquidity rewards are mainly allocated to Supercharged pools, as these cater for the majority of volume, accumulating block by block, requiring a claiming transaction.

With the increased swap fees generated by the efficiency of concentrated liquidity positions, the fee subsidy rate on Osmosis has fallen below one, meaning that more fees are being generated for liquidity providers than they are receiving from OSMO incentives.

image
image1810×720 48.3 KB

This has changed the purpose of OSMO incentives from being the main reason to provide liquidity to being a guide to which pools should have additional liquidity deployed to optimize trading routes.

Since the loss of the OSMO multihop discount with the implementation of taker fees, this will likely be pairings directly between two non-OSMO assets. Incentivizing these pools to attract liquidity will maximize volume by minimizing trader fees per trade. Therefore, the purpose of OSMO incentives is to bootstrap these pools initially, followed by supplementing any impermanent loss over time by Liquidity Providers.

Historically, Osmosis governance has been reluctant to allocate OSMO incentives to non-OSMO pools as these liquidity providers have no reason to keep the emissions by compounding back into the pool. This led to governance proposals being passed, such as:

  • Proposal 128: Add a bias factor to external incentive matching
  • Proposal 264: External Incentive Matching Reduction within non-OSMO categories:
  • Proposal 273: Phase out Incentives on smaller Other Category pools
  • Proposal 389: Remove Non-OSMO Pool incentives

Suppose OSMO incentives are to resume being allocated to non-OSMO pools. In that case, alignment must be encouraged between Liquidity Providers for these pools and Osmosis. Since these liquidity providers are mainly after the trading fees in their pairings, they will likely sell any OSMO rewards for bonus yield rather than engaging with its usage.

This proposed tokenomic change adds some friction to an undesirable user flow for Osmosis while lowering it for a more desirable one, making the choice of OSMO reward usage a conscious one rather than routine. In the event of losses by liquidity providers, they may unstake these rewards and sell to recoup their losses. Until then, the minority of their rewards will further secure their liquidity on chain and earn them staking rewards, partially composed of the taker fees their liquidity helps generate.

This awareness of the governance process also encourages active liquidity providers on Osmosis to engage with Osmosis governance to request developmental improvements that improve the platform and attract further liquidity providers.

Validator Set Preferences

Implemented in v15 of Osmosis, users can choose their preferred validator set. This allows easy distribution of staked tokens amongst a wider variety of validators while maintaining the desired weighting for new delegations by signing a single transaction that delegates to this weighted validator set. Validator sets can also be recommended by third parties to encourage the distribution of stake amongst validators who match criteria that the third party determines, e.g., active governors, relayers, or core stack contributors.

If a user has not chosen a validator set, the delegation will default to a user’s current staking distribution as the validator set. This would ensure that rewards are still staked to provide security and earn staking rewards, even if the user has not explicitly chosen a validator set.

For users who have not staked any tokens, the proposal would cause liquidity rewards to require claiming with a vesting period equivalent to the unbonding period. Liquidity providers should choose a preferred validator set so that their liquidity rewards earn yield rather than being merely vested and does not allow liquidity providers to bypass the auto-stake system by not choosing a validator.

Target On-chain date: 30th October 2023

Note: Previous discussion was had 9 months ago on Commonwealth here: Commonwealth
This proposal was previously abandoned due to the bonding period already present in liquidity pools as well as Osmosis rewards being the main return from liquidity providing. As neither of these are the case and we may move towards non-OSMO pools being primarily incentivised I am re-raising this proposal.

1 Like
2

I think you mean that the vesting period is equal to the unstaking period (the text now says "staking period, which can in theory be infinite).

Furthermore, is there a guide how to set up such a preferred set? I can’t recall seeing any front-end being able to support this feature until now, so it might be that some steps are needed before this step is actually taken.

And lastly; why not use the OSMO where no preference is set to automatically stake to the bottom 10 validators to strengthen the network and improve decentralisation? A lot of people might set and forget, which will actively benefit the network and the validators.

3

I think you mean that the vesting period is equal to the unstaking period (the text now says "staking period, which can in theory be infinite).

I’ll change to Unbonding time since thats the param in use, I see how that could be interpreted in both ways.

Furthermore, is there a guide how to set up such a preferred set? I can’t recall seeing any front-end being able to support this feature until now, so it might be that some steps are needed before this step is actually taken.

There are technical docs at https://github.com/osmosis-labs/osmosis/tree/main/x/valset-pref and the frontend feature should be out pretty soon.

And lastly; why not use the OSMO where no preference is set to automatically stake to the bottom 10 validators to strengthen the network and improve decentralisation? A lot of people might set and forget, which will actively benefit the network and the validators.

That idea feels like it would bring more contention by making the validator minimum stake higher.

4

Oooh, reviving this proposal from the dead, very interesting. I think it’s a lot more relevant now than it was last time this was proposed, given concentrated liquidity and the shift toward trading fees as a revenue source.

I’m still a bit wary of pulling too many major levers at once, and as I’ve said before I feel like we’re moving too quickly with massive protocol changes without letting the dust settle from other changes. However, given that we’ll be launching large incentives campaigns for dydx and celestia I figure that it makes sense to implement something like this now (or relatively soon).

The only question I have is how will this impact integrators that rely on LP rewards being liquid, like Quasar and Mars? Are they aware of this? If not, it’d be good to get their feedback so that they can adjust accordingly if this passes.

1 Like
5

Since this would be something in a software upgrade they would have time to adjust - last time that was also a major source of complaints. However, I think this should be far less of a case because of the fee subsidy rate being so low.

They will have to decide what to do with the OSMO rewards themselves, I think it could be interesting if they decided to keep the staked OSMO for the protocol itself and potentially subsidize users in return.
I think they’ll likely just implement a daily unbonding event and so there will just be a slight lag in their compounding though. I took it to Twitter recently and didn’t have any major feedback against but haven’t heard directly from them.

1 Like