Remove LP incentives from non-strategic pools

This is proposal to remove incentives from all pools that contain assets that are not one of the following:

  • USDC
  • USDT
  • DAI
  • WBTC
  • ETH
  • ATOM
  • OSMO
  • stETH
  • stATOM
  • stOSMO
  • IBCX


Currently, there are ~35,500 OSMO budgeted for LP incentives on a daily basis. ~3,900 is redirected to the Community Pool, leaving ~32,600 for incentives. Most of these incentives are directed toward strategic pools that require higher liquidity due to volume/liquidation considerations.

However, there are currently a lot of other pools that receive a small amount of incentives that do not have the same considerations, and while small individually, add up in aggregate, and also require a high level of effort to track and maintain. The removed pools should be able to maintain sufficient levels of liquidity without OSMO incentives with concentrated liquidity. This would also represent a strategic shift in Osmosis with recommending communities that are unable to obtain sufficient liquidity levels to pay their own external incentives to gain sufficient liquidity.

The removed incentives would be redirected to the community pool. The overall incentives would be ratcheted down, and when new strategic pools become available, the incentives would be redistributed from existing pools. There may also be future mechanisms in place to aid in bootstrapping pools before governance passes assets into the incentive redistribution program.


This change would reduce the current list from 60 pools to 14 pools and reduce daily Osmo incentives from ~32,700 OSMO to ~24,500 OSMO. It would eliminate 46 pools and ~8,200 OSMO Incentives.

The 14 remaining pools would be as follows:

These pools remain because they are strategically important to Osmosis, and the community should be hyper-focused on building the liquidity in these pairs.

We reviewed the pools that have been eliminated from incentives and confirmed that there are no other strategic pools that should be incentivized. The 46 removed pools have an average daily Osmo Incentives of ~180 OSMO per day or ~$60 per day. At first glance, it is highly unlikely that LPers are drawn to $60 of revenue per day. It is more likely that they are incentivized by the fees generated from these pools and general support for the coin (see chart below for comparison of APR with and without Osmosis Incentives). Due to the low levels of required incentives, it is likely that the individual communities will be more effective in managing their liquidity requirements through external gauges.

The eliminated pools have a range of issuing 2 OSMO per day (<$1) to 1,300 OSMO per day ($442). Only 5 pools receive more than 400 OSMO per day ($136). They are:


These 5 pools make up the largest portion of the removed incentives, with ~4,100 OSMO per day or roughly half of the ~8,200 incentives. As these 5 pools make the largest portion of the eliminated pools, we performed an analysis of these pools:

Even with removing incentives from these pools, we should maintain 10%+ APR on these pools at current liquidity levels, which is competitive compared to staking rates provided by each chain. Also, it is important to note that with concentrated liquidity, liquidity providers can actively manage their position and earn higher APR for doing so. Providing tighter positions will increase their ROI and reduce slippage for users on Osmosis.

We also reviewed the slippage of current liquidity levels to handle 95% of all users’ swaps, excluding Arbitrage swaps. The chart below shows the dollar value 95% of users swaps are at or below. For example, 95% of user swaps in pool 722 are for $350 or less. If we swapped at the high end of this range, we would have 0.10% price impact.


Historically, there’s an inelastic relationship between an LP’s incentives and its liquidity. Reducing an LP’s incentives by 10% will not reduce its liquidity by 10%. You can observe this by looking at historical GAMM levels of LPs over the last year of incentive cuts. 1% slippage is the default shown on the site when calculating minAmountOut. All pools would have slippage under this threshold for 95% of users swaps. Further, the APR generated from swap fees continues to be at a high enough rate to attract more liquidity as users migrate their liquidity to concentrated liquidity positions.

The remaining 41 pools have an average of 100 OSMO (~$34) in daily incentives split between their liquidity providers. It is likely they follow a similar trend as the 5 aforementioned pools, perhaps even less impacted by incentives since they are incentivized at a lower clip and we will pass further review. These pools will be more effectively managed by their own respective communities who can track their own individual liquidity needs and incentivize through external gauges.

I plan to continue to monitor these pools in the coming months and impacts on liquidity but also other factors such as Taker Fees. This can monitored by the community here: Remove Incentives - Google Sheets

Mechanisms for Bootstrapping Future Pools

The Osmosis Protocol in the future may create programs to bootstrap future pools. An example of what this may look like is having a sub-dao with discretionary budget to bootstrap strategic pools on Osmosis before governance adds them to the recurrent incentives redistribution process. This mechanism will be subject to a future proposal.

Target On-chain Date: Monday, September 25th.


Hey Marlin! Thanks for putting this up.

While I’m supportive of the general idea behind this, I think the proposed execution puts Osmosis at significant and unnecessary risk of a liquidity death spiral. I propose as an alternative removing incentives slowly over 2 months, which would allow us time to gauge the impact of an incentives removal on Osmosis liquidity. Here’s why:

The Timing of this Proposal Couldn’t Be Worse

Osmosis is in a period of significant change. Liquidity is currently migrating to concentrated liquidity pools and, in the process of doing so, is unbonding at one of the highest rates in Osmosis history.

At the same time, Neutron is launching a massive incentives initiative for pools on Astroport and, soon, Duality. They’re very well-positioned to vampire liquidity from Osmosis at exactly the time when we’re proposing to remove incentives all at once while people are already unbonding. This is a very bad time to be proposing a sudden and dramatic removal of incentives.

There’s an argument that with concentrated liquidity, fees on these pools will be a sufficient incentive to generate APR that keeps liquidity relatively stable. However, this argument assumes that volume remains the same when often this isn’t the case. In reality as liquidity falls, so do volumes, which reduces APR, which causes more liquidity to be removed, etc. There’s potential danger for a death spiral for liquidity here.

If this weren’t the case we’d be comfortable removing incentives from all pools permanently, including the ones not covered by this proposal.

The Risk of Pulling all Incentives at Once Far Outweighs the Cost Savings

Running some quick numbers, if we assume incentives get removed in 60 days vs now, this would allow an additional 492,000 OSMO to be released into circulation (less than 0.1% of the circulating supply). 492k is calculated by the 8200 OSMO in cost savings from this proposal * 60 days.

If we phase a removal out on a weekly basis over that 60 days, that number decreases even further. The new supply put into circulation over that time is simply to small to have a measurable impact on token price.

On the other hand, if incentives are pulled immediately and a mass liquidity exit occurs, most of these tokens are paired with OSMO. What do we think will happen to those OSMO tokens as this liquidity migrates over to Neutron (where no OSMO pairs exist)? These will naturally be sold into the pools.

Removing incentives slowly solves for both problems by allowing us to monitor liquidity changes in pools for a disproportionate impact, while also offsetting any sell pressure over that 60 day period.

Re: This Proposal

I want to stress that I’m not advocating that we don’t remove incentives, but rather that we do so with intention and with caution n a way that allows us to gauge the impact / protect against any potential liquidity death spiral or vampire attack from Neutron. There’s literally no reason not to.

There’s very little harm in doing so from a supply dilution perspective, whereas the harm from pulling them all at once could be quite significant.

1 Like

I would really like to hear what @HathorNodes and the liquidity incentive working group has to say about this.

If I am not mistaken they are having a meeting next Friday and can add this as a new business agenda item.

Thanks @RoboMcGobo.

Also don’t forget the addition of the taker fee which can add to the balance and speed up the death spiral. I am certainly not a fan of doing too many changes at once. We somehow seem to think that a loooot needs to be changed in a very short timespan, mainly because no changes have been made since the top of the bull and now. I find that very risky and indeed ill-timed.

Furthermore; some of the assets are also weird choices. For example NCT where in the past the community expressed to have Osmosis as the potential main location for trading carbon credits. And also the removal of qOSMO/OSMO is something I don’t like, because we go putting all our eggs in the Stride basket. We apparently didn’t learn from the UST-scenario where that is very unwise choice to make and I will certainly not vote in favor of it.

Some other choices I fully understand though. So with a small reduction in scope it can be something where we can monitor the evolution of TVL and volume on these pools to build some history we can use to make choices on when things quiet down a bit.

1 Like

Due to the inelasticity of the supply I do not think there would be much change regardless how long we take to ease out of rewards. These pools have eased out of rewards for the past couple years through standard decrease in emissions and are at a point where it is negligible as evidenced by only the 5 pools above having more than 400 OSMO per day in incentives (~$125).

To further demonstrate the inelasticity of the pools I have looked at the rewards and supply over time, along with tracking token price, and noted that liquidity is more correlated to token price than incentives:

From the chart above we can note that incentives from AXL went up from July to August by nearly double. The liquidity in this time period however went DOWN. This is because the price of OSMO went down in the same time period. No new AXL was being put into these liquidity pools. The incentives at this point have very little impact which tracks with the analysis done in the original post where there is sufficient swap fee incentives to incentivize providing liquidity in these pools. Additional incentives are overpaying.

Source1 Incentives:

Source2 Supply:

Additionally since the other assets are incentivized at a very low amount (less than 400 osmo daily), there is very little risk of pulling too many incentives from them at once. They have already have had a time period of decreasing incentives.

Also note, this is a signaling proposal. If passed, the routine prop implementing these changes will occur at the start of October

If you don’t mind, can you please share the results of your correlation analysis and the data used.

If you don’t mind, can you please share the results of your correlation analysis and the data used.

See the Table listed above. I viewed the assets that receive the most OSMO incentives that be removed and viewed Liquidity in the middle of each month, as well as Incentives received. Incentives increased/decreased during during the period viewed and the liquidity didnt seem to necessarily move in the same direction. I then noted the asset price and compared the price movements of these assets and looked at the liquidity in the chart as well and noted that increase/decrease in prices led to increase/decrease in liquidity. Therefore, for these 5 assets, it appears coins aren’t moving in any meaningful way from incentives and the price impact has the largest impact on liquidity.

So when I run a simple OLS regression using the data from your table I get these results:



  • asset 2 = INJ
  • asset 3 = AKT
  • asset 4 = AXL
  • asset 5 = SCRT


  • month 6 = June
  • month 7 = July
  • month 8 = August
  • month 9 = September

As you can see, the number of OSMO liquidity incentives is correlated with liquidity, but neither is the price of OSMO nor the price of the corresponding asset.

All else being equal, the model shows that we get $307.30 in liquidity for every OSMO we spend on liquidity incentives, regardless of the price of OSMO or the price of the other asset.

What is interesting is that liquidity is statistically correlated with two assets, AKT and AXL. The model suggest that the AKT/OSMO pool would have $848,272.40 in liquidity while the AXL/OSMO pool would have $707,219.50 in liquidity without any incentives. The Mesh Security partnership Osmosis has with AXL and AKT, the community pool holding more than 1M AXL (~$360K currently) and nearly 35K AKT (~$28K), less than 39% of liquidity in the AXL/OSMO pool is bonded, and AXL is a leverage asset on the Mars Outpost while both AXL and AKT are on Umee, are perhaps driving this correlation.

1 Like

So, concluding, the OSMO incentives are not ill spend on those pools?

I am mostly curious where the sell pressure comes from. Do they come from the “strategic” important pools mentioned above or these smaller pools. The amount of sell pressure we see can’t come from these pools only.

So the narrative that this might reduce sell pressure might even be completely wrong.

1 Like

interesting discussions on the liquidity and incentive mechanics, but let’s not forget the end-user here. A lot of folks use Osmosis as their go-to DEX for swaps within the Cosmos ecosystem. Dramatically reducing or removing incentives might indeed risk a liquidity drop, and you know what that means—worse slippage, higher costs, and slower trades. That’s a straight-up hit on UX, which, let’s be honest, can make or break a platform. So, while we’re crunching numbers and projecting outcomes, let’s make sure we don’t alienate the very users that make Osmosis tick. Just my two cents


At least for the SCRT/OSMO, EVMOS/OSMO, and INJ/OSMO pools. Because of the statistical significance of AXL and AKT in and of themselves, and both the AXL/OSMO and AKT/OSMO pools are SFS pools, it is likely that liquidity incentives can be removed without there being a significant drop in either AXL or AKT liquidity on Osmosis from their current levels.

The model does suggests that Osmosis could loose $231K in SCRT liquidity, $155K in EVMOS liquidity, and $286K in INJ liquidity. This totals to $672K in liquidity. On the low end the model suggest that $99K in liquidity would leave Osmosis across these three assets while on the high end it suggest that up to $1.2M in liquidity could. Because SCRT/OSMO, EVMOS/OSMO, and INJ/OSMO are also all SFS pools, SFS could blunt the bleed so to speak. There are DEXs on SCRT and EVMOS though, and INJ is a DEX too.

While WhiteMarlin is likely correct that reducing a LP’s incentives by 10% won’t result in a 10% loss in liquidity, the model strongly suggests the elasticity of liquidity varies by token. As such, certain liquidity pools will be effected more by the loss of incentives than others.

I am also concerned about the accuracy of some of WhiteMarlins conclusions about how inelastic liquidity is given the statistical evidence I have presented that contradicts his statement “that liquidity is more correlated to token price than incentives”, given his use of an “eyeball” test rather than any actual statistical tests for correlation to make such a claim.

1 Like

For me the most worrying is that it seems that every idea @sunnya97 shoots from the hip is converted into a proposal.

I have already seen Sunny putting nuance in his post, but that is already past the point of making it into the proposal. So this era is quite scary where people are following such ideas blindly…