Rather than make a new thread I’m going to post my contribution here!
Will beat me to this, but I had been working on presenting something similar since I posted this Poll which backed up that sentiment did lean towards a reduction in inflation.
As with Aaron’s disclaimer, these are my own views on where Osmosis tokenomics should go and do not speak for the Osmosis Foundation.
Changes since last Review
Since the previous reviews in proposals 484, 485, and 486, many additional proposals have passed relating to tokenomics.
Proposal 530 and 549 added a Protocol fee to all trades on Osmosis, proceeds of which were split between stakers and the community pool.
Proposals 709 and 710 implemented additional revenue for the community pool and a burn mechanism for OSMO.
Version 25 of Osmosis also added a revenue stream from blockspace auctions.
Current comments on OSMO tokenomics center around noticeable inflation levels and inefficiencies in distribution allocations due to incentive redirection.
This post presents five main strategies for tackling these issues as well as having an overall analysis of inflation.
Proposal A/B - Cut and Stretch Inflation
Proposal D - Inflation to Zero
Allocate Top of Block Auction as Validator Subsidy
Proposal E - Adaptive Inflation
Proposals A and B - Cut and Stretch Inflation
Similar to the mechanism employed in “OSMO 2.0”, these proposals suggest implementing a thirdening to make up for the one skipped this year due to the previous extension to a two-year thirdening period while maintaining the 1 billion max supply.
The two main paths would be:
- Perform a thirdening and return to a yearly inflation cut of a lower adjustment of around -14.5% per year.
- Perform a thirdening and increase the time between the thirdening cuts to around 800 days.
Option A is likely the more resilient of the two since inflation is cut more frequently rather than waiting for over two years for another noticeable change in emissions.
Proposal A generally results in less inflation over the next few years while extending a lower but more consistent level of inflation for longer, while Proposal B moves to the end stage of tokenomics and total reliance on protocol revenue more quickly.
Both of these models would result in lower emissions overall, and a smaller quantity of OSMO would be sent to the community pool as a more significant proportion of the Liquidity Incentives emissions would be needed to maintain the current spend.
However, neither of these models addresses the core problem: inflation is a helpful distribution and bootstrapping model, but arbitrary inflation amounts based on a decreasing issuance are not suitable for all phases in a protocol’s lifetime.
Osmosis will become a very low-inflation chain after the next few years, regardless of any changes, requiring that the multiple facets of Protocol Revenue come together to sustain the validation, continued innovation, and development of new liquidity markets.
The thesis of Proposals A and B is that the changes in OSMO 2.0 left inflation still too high for the current stage of Osmosis tokenomics, and fine-tuning the schedule to approach the final sustainability stage more rapidly is required.
Inflation Analysis
This section examines what we achieve by using inflation for each emission location.
For reference, current inflation is minting 182k OSMO a day, resulting in an inflation of 9.83%.
Chart 1: Overview of OSMO equivalent flow from Inflation and Revenue
Staking Rewards
- 91k OSMO from inflation (50%)
- Receives Transaction fees, averaged over 90 days is $680, ~1,600 OSMO
- Receives all OSMO from Protocol Fees and 33% of non-OSMO, average over 90 days is ~$8500, ~21,500 OSMO
Staking Rewards incentivize stakers to provide OSMO to secure the chain. Currently, there is a Staking APR of around 10.8%, of which around 9.1% comes from inflation and 1.7% comes from revenue sources. 53.7% of all OSMO is currently staked.
Typically, this is referred to in two ways.
Headline APR
Headline APR is often used as an opportunity cost calculation for alternative capital uses. High APRs generate more interest when comparing across chains but, if derived entirely from inflation, will generate poor returns in absolute value.
Chains with well-established staking sets, such as Ethereum and Cardano, return around 3% for staking. More volatile sets, such as Solana and Avalanche, have a rate of around 7-8%. Any step change visible may be due to the limited data set.
High headline APR is initially attractive, but at a certain point, it also likely indicates a market risk to more seasoned staking security providers.
Chart 2: Inflation APR by Chain, Source: Staking Rewards
Inflation Adjusted Return
Return against inflation typically governs the amount staked within each chain. Proposal 485 also talked about this, which increased the net yield of staking rewards from negative to positive. Seeing the rapid increase of the percentage of OSMO staked move from 49% to around 56% before the net yield returned to around 2%
Increasing the net yield by allocating a more significant proportion of any emissions to Staking or making the protocol revenue return a higher proportion of rewards will likely increase the amount of OSMO staked by participants.
With the current percentage of OSMO staked being higher than the minting proportion to Staking, there is an inflation-adjusted return of 0.94%. There would be no adjustment if inflation were removed, so it would solely be the protocol revenue return, currently at 1.7%.
Validator Overheads
Staking rewards also pay for validation overheads at the typical 5% commission, which is around 4,500 OSMO a day. Any reduction to staking rewards could make validation unprofitable for smaller validators.
This lack of sustainability for validator operations could be tackled through a direct flat subsidy to the validators without relying on commission rate changes, which may cause a centralizing force as the larger validators could afford a lower commission and attract a greater share of delegations. If this flat subsidy is insufficient from a source such as the Top of Block auctions, then the validator set should be shrunk to compensate.
Recommendation: Lower OSMO Staking APR towards the 7-8% range as a first step towards eventually a target of 3% APR entirely from yield.
Recommendation: Provide Validator sustainability payment in order to ensure that Validators remain profitable.
Development Vesting
- 46k OSMO from inflation (25%), vested with a year cliff from issuance.
Used to fund ongoing chain development and have the most significant impact addition to circulating supply beyond the staking rewards. As this is vested, it encourages the developers to work to increase the value of their vested stake.
As issuance decreases, the quantities issued here are less impactful than the previous vesting development funds, which recipients may stake. As inflation lowers and revenue comes from non-inflationary sources, developers, stakers, and community pool grant recipients have a similar interest in the success of Osmosis.
Recommendation: Remove Dev Vesting Share of inflation and rely on Staker alignment and Community Pool grants (25% cut)
Liquidity Incentives
- 9k OSMO (5%) distributed of a 37k (20%) OSMO budget
The current spending rate only uses a portion of the allocation, with the remainder being transferred to the Community pool for later usage.
Just under half of these incentives (3,820) go to OSMO/STABLE pairings, mainly required to offset the opportunity cost vs staking and compensate for OSMO’s inflation. Reducing inflation overall will likely make this spending more effective at establishing deep Stable paired liquidity with OSMO as there is less Impermanent Loss to offset.
These redirected incentives have generated around 50 million of the 67 million OSMO community pool. It would take around 15 years to distribute fully at the current spend rate.
Recommendation: Remove the Liquidity Incentive share of inflation and establish a working group with a budget from the community pool. (20% cut to minted tokens)
Community Pool
- 9,000 OSMO (5%) from inflation, plus 27,000 OSMO (15%) per day from redirected liquidity incentives
- Taker Fees, averaged over 90 days, ~$3,250 in non-OSMO
- Top of Block Auction, averaged over 30 days, ~$590 in USDC, listed here as no use currently decided
- Pool and Incentive gauge creation fees are minimal at 20 USDC per pool and 50 OSMO per gauge.
The community pool funds initiatives such as the Osmosis Support Lab and Osmosis Grants Program.
The community pool is vast due to receiving redirected incentives for the last two years. Still, it is primarily OSMO-based and unable to be directly spent without diluting the circulating supply. Additional OSMO from inflation has no use due to this.
Taker fees have increasingly been a source of non-OSMO revenue for the community pool, and efforts should be pursued to convert the OSMO in the pool to non-OSMO sustainably.
Recommendation: Remove the Community Pool share of inflation (5% cut)
Summary
In summary, cutting 50% of inflation instantly through changing the mint allocation ratios to become 100% to staking and performing a 50% inflation cut is possible right now.
Further to that, the inflationary staking rewards themselves should be cut by around half, resulting in an inflationary rate of around 2.5% and a Staking APR of around 6.9% at current staking levels, consisting of around 1.7% APR from revenue and around 4.2% return from inflation.
Proposal C - Inflation Cut
This proposal performs the inflation cuts above by:
- Set a thirdening duration of 365 to trigger a thirdening the epoch after this proposal.
- Set the Reduction Factor to 0.25 to reduce all inflation by three quarters, from 9.83% to 2.46%
- Set the mint allocation ratio to 100% for stakers and 0% for other destinations.
- Perform a community pool spend of 800,000 OSMO to a multisig controlled by the Incentives Working Group to budget for Osmosis incentives for the next three months.
- Unless further proposals are made to adjust the emission rate, this will reduce the maximum supply to around 743 million OSMO.
Chart: Overview of OSMO equivalent flow from Inflation and Revenue, implementing Proposal C with additional Top of Block Auction allocation
Proposal D - Inflation to Zero
Proposal D is an even more drastic step that posits that no inflation is required at this point in Osmosis tokenomics and that the move to a complete reliance on Protocol Revenue is now possible.
This proposal would
- Set a thirdening duration of 365 to trigger a thirdening the epoch after this proposal.
- Set the Reduction Factor to 0 to end inflation, capping the Maximum supply of OSMO at the current supply. (678m at time of writing)
- Perform a community pool spend of 800,000 OSMO to a multisig controlled by the Incentives Working Group to budget for Osmosis incentives for the next three months.
Staking rewards would fall to match protocol revenue, a lower headline return of ~1.7%; however, this would be a higher net yield as no new OSMO would be minted.
Chart: Comparison of Inflation rate over time with the above proposals
Proposal to Allocate Top of Block Auction as Validator Subsidy
The average Validator income from inflation is currently 30 OSMO per day.
This proposal asks for the Top of Block auction to be distributed evenly between all validators as a direct reward for securing the active set. This distribution would offset around 10 OSMO in removal. However, the Top of Block auction is still in the early phases of adoption, and yield may significantly increase as liquidity and volumes increase.
Funds already in the Top of Block Auction Module should be retained in the community pool.
Proposal E - Adaptive Inflation
Proposal E is a middle ground between the previous two proposals, which recognizes that inflation is likely too high but that lowering this far into single figures should be dependent on revenue generated by the chain, with inflation paying for a shortfall in yield compared to the requirement to attract staking security.
This proposal could be implemented after Proposal C as a next step to transition towards Zero Inflation and would require a software upgrade to implement.
This proposal asks that an additional step be inserted in the staking reward distribution process between the minting of tokens and the Staking Pool Allocation and distribution.
This step should be in the form of a module address that collects OSMO both as inflationary minting and as any revenue allocated for distribution to stakers, such as Transaction fees and Protocol fees.
The onward distribution of this OSMO should be controlled by a governance-adjustable parameter that can directly set the staking APR for stake bonded at each epoch to be distributed from this module wallet.
Example
Taking a daily minting of 45,000 OSMO, sources of protocol revenue yielding 25,000 OSMO per day on average and 350m OSMO staked.
Governance may specify a 7% APR for the staking rate, requiring an allocation of ~67,000 OSMO to stakers, leaving ~3,000 OSMO unissued.
This unissued OSMO forms the basis of a buffer that stabilizes the APR at a fixed rate as protocol revenue, the number of OSMO staked, and the parameter itself is modified, leading to a more predictable staking experience compared to the heavily fluctuating returns that are present at the moment.
If there is excessive OSMO in the module address, then this provides a runway for this level of APR for a period of time. As Protocol Revenue to stakers increases and makes a greater share of the total emissions, then this runway becomes infinite.
If staking increases or is static at the new 7% APR, then the APR parameter can also be directly lowered to further decrease inflationary emissions and increase the reliance on protocol revenue while being simple to revert through governance.
If there are not enough OSMO in the module address to reach the target APR then the APR underperforms the intended rate. This may result in Stakers unbonding, reducing the need for as many OSMO to obtain the target APR, but not increasing the inflation above the previously intended level.
Chart showing how spends are handled first as Stake fluctuates, target APR is lowered and protocol revenue increases
Final note:
This is aimed at solving the issue overall going forward, but rather than getting hung up on details, I’ll also support a flat 50% cut—which is kind of halfway between A/B and C but probably far, far simpler to agree on.