Proposal For Introducing a Programmatic, Market-Based Emission Mechanism Based on Staking Participation Rate

[Ben.Hawkins]

Inflation exists primarily to bootstrap network security in Proof-of-Stake systems. However, Solana’s original inflation parameters were set arbitrarily high at a time when the network required substantial incentives to attract stakers. Today, given Solana’s significant growth and robust economic activity, these incentives have become excessive and are no longer economically optimal.

Current inflation imposes an unnecessary cost on SOL holders and creates selling pressure because rewards often must be liquidated to cover tax obligations. This persistent inflation acts as a leaky bucket, unnecessarily diluting SOL’s economic value.

SIMD-0228 introduces a new emission schedule that dynamically adjusts SOL issuance based on staking participation rates. Specifically, this proposal suggests:
• Increasing inflation modestly if staking participation drops below 33%, thereby safeguarding the network in the worst-case scenario.
• Reducing inflation below the current rate when staking exceeds 33%, aligning rewards more closely with the actual economic needs and security requirements of the network.

At approximately a 42% staking participation rate, staking returns would remain exactly as they are today, but inflation would fall significantly to just 2.7%. This change clearly benefits Solana by drastically reducing unnecessary dilution of SOL’s economic value.

Regarding concerns that lower inflation could negatively impact smaller validators by reducing their profitability, it’s important to clarify that validator profitability and sustainability are better addressed through targeted solutions such as reducing vote transaction costs and other efficiency measures. Inflation primarily serves as an incentive for stakers rather than as a direct subsidy for validators. Thus, inflation adjustments should be considered separately from validator profitability.

In summary:
• Strengthening security through automatically increasing staking incentives during periods of lower participation.
• Reducing overall inflation, allowing SOL’s market dynamics to become healthier, more attractive to long-term investors, and less distorted by artificial selling pressure.
• Validator sustainability should be addressed through more effective, targeted mechanisms instead of unnecessary high inflation.

[Ben.Hawkins]

For those that haven’t seen it, here’s a helpful calculator to see the proposed emission values at differing staking rates: https://solana-inflation-calculator.vercel.app/

[OscarGF]

First of all, I want to thank @kankanivishal for co-creating this proposal alongside @tsharkamble7780 and Max.

I’d like to share my perspective not only as a staker but also as an independent validator actively bootstrapping an upcoming product within the Solana ecosystem and joining the builder community.

From a validator standpoint, a key concern within the community is the potential impact on the validator set. To be direct and assertive, although this could also affect Starke’s position, this discussion should not be driven by emotions unless the reduction in validators poses a risk to network security, which is ultimately the primary role of a validator. That said, it is undeniable that today’s validators are not truly independent entities, given their strong reliance on SFDP and other stake pools. From this perspective, I think the outcome of this SIMD alone does not significantly influence true network security, even if it affects the number of validators.

That being said, my biggest concern is not about whether inflation should be adjusted but about the timeline it is being proposed. We are discussing fundamentally altering the emission policy of a multibillion dollar system, yet considering a drastic shift in its issuance schedule by a large percentage. From an institutional perspective, such an aggressive change in a foundational monetary characteristic introduces substantial risks, both in terms of economic predictability and broader market sentiment.

One cannot underestimate the collateral impact of altering a monetary policy so abruptly. Such a move risks shaking confidence in the system’s long-term stability, as it might set a precedent for more abrupt changes in the future. Historically, whether in crypto or traditional finance, monetary policies require careful calibration to avoid unintended consequences. Investors, validators, and builders rely on consistency and predictability to make informed decisions. A drastic shift like this could trigger unintended market reactions, including a potential loss of confidence in the governance process itself.

To mitigate these risks, I strongly encourage:

Progressive Implementation

If the goal is reducing inflation, let’s achieve the same outcome over an extended period. A gradual approach allows the market to adjust smoothly and prevents unnecessary volatility.

Rollback Mechanisms

If the expected results do not align with initial projections, there should be a rollback option to mitigate an undesired and unpredicted end state that might force another drastic change as iteration to a different direction not drafted at this point and also different to the initial one. This would also reduce the potential uncertainty in a worst case scenario.

If no changes are made, we will vote NO.

[Umberto]

Hey all, we at Chorus One are working on a detailed article analyzing different aspects of SIMD228. The article will be out soon, in the meantime, I believe it might be useful to check out our dashboards

1. Security from current curve based on growth assumption - you can tweak parameters based on your own assumption and make conclusions
2. Overview on several topics affected by current inflation and proposed inflation - e.g. historical stake rate, APR, TVL ecc

[alphagamma]

If this proposal is approved, say staking rate remains above 65% regardless, how many epochs before inflation rate goes to 0 assuming the current inflation rate? 4.690%

[OscarGF]

How it has been described, the new curve would be applied in only 10 epochs and current projection based on current stake is ~1.7%

@kankanivishal

To ensure that the transition from the old static issuance schedule to this new schedule is smooth, we will interpolate between the old issuance rate and the new issuance rate over 10 epochs using the formula:

where is a parameter that controls the speed of the transition, taking the values 1/10, 2/10, …, 9/10, 1, over the first 10 epochs before settling to the new issuance rate at = 1.

[ghostlabs]

I just want to start by noting from an economic point of view a dynamic market based emissions schedule is a sensible mechanism to drive toward, so no debate on the premise.

For the sake of discussion and to avoid repeating what has already been laid out above, I’ll make a few succinct points below that I would be interested to see community views or debate on as I am still trying to formulate my own informed view on this one:

1. If a transition to a dynamic model occurs, is the proposed method of anchoring ‘Staking Participation’ the optimal approach? Part of the rationale provided for this change hinges on substantial increases in REV as the source of income to Validators to operate a globally distributed transaction network. Over the course of 2024 average REV contributed only 22% of the network TEV. Q4 was an impactful operating period contributing to a 5.6% increase alone (an outlier in the dataset) and seemingly a heavily weighted assumption of the rationale being REV will continue to sustain or maintain a relative growth rate. Should REV factor in to this model as a more intrinsic input toward bottom-line network operator revenue growth? If “MEV” = “REV” perhaps its a terminology alignment, but the point remains - should the ‘Issuance Rate’ be tied to a function of actual network revenue growth?
   [Referenced Blockworks data. For consideration, ‘Operator Payouts’ do not reflect the variable expenses of running this global network, i.e. is it profitable/sustainable? What is the economic timing to reduce subsidization of the network? Separate discussion points below]
2. What is the definition of “Securing the Network”? Number of Operators? Enough diversified stake to represent the common interest? Given blockchain networks have other dimensions beyond ‘continued operability’, should these metrics be evaluated within the economic equation. In theory, if only 3 independent equally staked validators remained the network would still be secure and operational, until…?
3. Is there any analysis to assess the elasticity of the current stake distribution? {Price (growth) vs Rate of Return (income)} + [some coefficient of decentralization] - assuming that’s important in this new model
4. A big theme of discussion around this proposal as some have defined, “Death of the Small Validator”. Markets are dynamic, volatile and chase returns. Generally capital seeks to optimize return, and in liquid markets will move quickly. Running a successful validator has been posed in many cases as a ‘Full-Time Job’. Part of that is technical, another part is attracting “inelastic” stake. The magnitude of short-term or drawn out shocks to markets can be impactful. This is more of a fundamental consideration pointing back to some of the factors to the equation in #1-3. How important is this one? A random example that never leaves me, 2009 financial collapse led to much greater centralization of the ‘global financial transaction network’ - blockchains and market dynamics solve this (in this case if it is profitable to be an operator in the network).

From my position I raise these points as additional considerations for what the appropriate inputs of this new economic ‘staking/inflation’ model might be. This is still unchartered territory, overall structure and timing are important to ensure a stable foundation for sustainable long term network growth. Does this formula consider all of the variables in alignment with the overall goals of this new ‘Global Network State’?

[zantetsu]

I will not vote in favor of SIMD-228. I will outline my reasoning in a few posts, because there are several very distinct issues to address. First is the most fundamental issue: I do not feel that it is in the interest of my stakers to have their staking rewards reduced, so as a validator whose vote represents their interest I cannot in good conscience vote yes on this SIMD. The argument that this vote would be in their interest because it would create a better Solana whose benefits would outweigh their loss of rewards is interesting, but given that it is based on conjecture, and with the knowledge that all economic policy decisions have unintended consequences, I think that taking a chance that the outcome would be good for my stakers would be irresponsible. So on their behalf, I must vote no.

[zantetsu]

I will not vote in favor of SIMD-228. To support my position, I will address the argument that Solana is “overpaying for security”. I haven’t seen clear evidence for this. Some have proposed that if the inflation rate were vastly reduced, either stakers would devote more of their SOL to defi (meaning that it was only the high staking APY preventing their defi participation on the first place) or stay staked and accept the lower staking APY (meaning that the higher APY was never necessary to satisfy the security goal of incentivizing staking). I think what will actually happen is that the shock loss of staking APY will cause stakers to seek more returns from validators’ profits, which will lower validator profitability across the board. Validator profitability aligns validator actions with network security. Reduced profitability means reduced security. So what’s being offered by SIMD-228 is not “the same security for less cost”, it is “less security for less cost”. Whereas the former would be a clear win, the latter is not. And since the proponents of this proposal only offer the former argument, and provide no evidence nor even theory for what the impacts on Solana will be in the inevitable “less security” case of SIMD-228, I don’t think the “overpaying for security” argument is convincing as a reason to vote for SIMD-228.

[zantetsu]

I will not vote in favor of SIMD-228. To support my position, I will address the argument that reduced staking APY will lead to increased participation in defi. I think the most common reason, by far, that stake does not participate in defi is because of the perception of risk involved in utilizing defi. A large amount of stake likes the fact that vanilla staking is as secure as the Solana block chain itself, with no additional risks beyond that of holding SOL outside of a stake account. Reducing the staking APY will not cause a significant fraction of this stake to choose higher risk investment options (defi); instead this stake will do two things: one, seek more block reward returns from validators to offset the loss of native staking APY, and two, flee Solana altogether to find other more appealing low risk investments. Neither of these will be good for Solana. Validators taking less profits will harm the validator community, which will harm the security and performance of Solana. And stake fleeing will cause a price crash that will reduce confidence in the long term viability of Solana, in addition to recycling stake to a different new set of SOL holders who probably will also just stake their SOL (even at a reduced inflation rate) and for the same reason that the old stakers did. In other words, the likely outcome is just harm to the validator set, a drop in SOL prices, a churn of large amounts of SOL to new oners, and eventually the same network configuration as previously, with no significant benefit to defi. Thus the argument that SIMD-228 will drive staked SOL to defi is not convincing.

(EDIT: In order to avoid a no-3-consencutive-posts rule I will put my final post at the end of this post)

I will not vote in favor of SIMD-228. To support my position, I will point out my concerns about the structure of the dynamic emissions schedule. Put simply, it chooses an arbitrary 40% target staking threshold and applies significant stake reward penalties when this threshold is exceeded, and significant stake reward incentives when this threshold is not met. It provides no justification for 40%. This might or might not be a good target, but with no explanation for why this is the right number, nor even any fundamental principles stated from which arguments about what the right number is supplied, there is no way of knowing. I cannot support a proposal that puts forth an arbitrary target staking level without some justification for this level. Some have pointed out that the current inflation schedule was also chosen arbitrarily. This may be true, but even this fact is not evidence that a change should be made. The burden of proof of those proposing a change is on them to show evidence that the change is better than the current system. With no evidence supplied that a 40% target staking level is better than the staking levels that are expected to be achieved with the current inflation schedule, I cannot support a vote in favor of SIMD-228.

[alphagamma]

That’s quite a big jump.Esp considering there is no guarantee that those that stake will be instantly unbonding driven by yield

[laine]

After reviewing this proposal, many discussions in person and online with several of the involved and affected parties, debates on Discord and X as well as reviewing the information here, I believe this proposal is a positive development for Solana and will vote in favour of it.

In addition to Max Resnick’s Leaky Bucket Theory explained here, and the consideration that real yield is the delta between inflation and staking return (i.e. currently 7% - 4.6% = 2.4% real yield, with the proposed curve at current stake ~0.7% but with only about 10% drop in stake approaching the same real yield as currently), below my other thoughts on the topic:

Monetary Theory & the Solana Economy

I see Solana as a nation-state, a sovereign economy. It isn’t entirely, of course, but many properties and analogies work.

Inflation is an incentive to “save” SOL in static stake accounts where they collect “interest”, this interest, or inflation, dilutes all the people who don’t save, just like the Benjamin in your wallet is worth less than it was a year ago, due to inflation in the US economy.

The chain’s value comes from its ability to unlock novel use cases and more efficient implementations of existing use cases, to utilize capital.

Capital efficiency has often been touted on Solana, but why do we need to be capital efficient? Because 63% of our base token is held hostage in stake accounts.

The real circulating supply is unstaked SOL is $38bn worth of SOL, which has created a chain with an FDV of $104bn.

The chain does not need 63% of SOL to be staked for it to maintain its security guarantees. If we can unlock just 20% of that it will add >$20bn to the circulating SOL supply, SOL that will be trading, swapped, providing liquidity, buying tokens and NFTs, investing in RWAs and other Solana protocols.

The way to grow Solana in the long-term is not to focus on 6% v 2% staking return, but how do we massively unlock our warchest of idling capital, sitting on the sidelines ready to go to war for us.

Validator set concerns

What Max does not address are the second-order effects of a drop in stake. While it is true that right now many smaller validators do not significantly rely on inflationary rewards or commission thereon, there are two scenarios that should be reasoned about:

1. If we enter a bear market where REV drops significantly, small validators may need to increase commission on inflationary rewards to remain profitable, while at the same time the loss of APY may drive stakers to demand compensation through larger block rewards sharing, this will compress validator margins
2. If SOL is unstaked and enters the liquid market, as long as this happens uniformly validators are not disadvantaged, as their revenue is determined by their relative share of total stake, however it is reasonable to assume that some validators will suffer disproportionately due to the composition of their delegators. It is unclear which cohort of delegators is most and least likely to unstake.

Shock value

The proposal will likely not be implemented before Q3 of 2025, at which point Solana’s current inflation curve is expected to be at about 4% p.a… The curve as proposed would reduce this value, at current stake levels, to just under 1%, with staking yield being about 1.9%.

The transition between the curves is currently proposed to take 10 epochs.

I have concerns that this remains too rapid, as many stakers have been staking for many years and may not be routinely aware of changes in staking yields, while outside investors may see the rapid change as a shock and risk.

At the same time the drop to under 1% feels aggressive, although the free market will ultimately find it’s price, and as stake drops the inflation rate will go up again.

Making the curve too flat/mild and the roll-out too prolonged will destroy any ability to meaningfully attribute resulting behaviours of the economy, base asset and validator set to the change.

I support a flattening of the curve where the initial inflation rate at current stake levels would drop to 1.5% rather than 0.9%, with a transition period of between 30-50 epochs.

Why now?

Why not? There is no exceedingly strong trigger to make a change now, other than the fact that the ecosystem and chain as a whole are at a point of economic and technical strength, poised for continued growth with the impending release of full Firedancer and ongoing performance improvements in Agave.

Together with the upcoming proposals to remove voting costs through an improved consensus algorithm, plug voting loopholes with Intermediate Vote Credits and future innovations such as async execution and multiple concurrent leaders, the ecosystem is graduating from a five-year period of aggressive and volatile growth and maturation into a period of technological domination and continuing mainstream adoption.

We need to act now to make our economy bullet-proof and attractive to bring the next billion users on-chain and become the NASDAQ of the blockchain world. Waiting to act too late will result in missed opportunities or rushed, flawed, implementations.

Conclusion

This change will strengthen Solana’s monetary policy and make it’s economics resilient, but there is non-incidental risk that there will be short- to medium-term upheaval in validator composition.

Ultimately removing friction of the leaky bucket and improving the economic engine of Solana by growing the amount of capital contributing to our Gross Domestic Product is a long-term strength for the ecosystem which will likely ameliorate short-term reductions in yield through future disproportionate growth in ecosystem market cap and base asset price.

Building a sustainable and long-term valuable asset and accompanying economy is difficult and involves balancing of many delicate knobs. A dynamic issuance curve that is linked to market behaviours is a good step forward in unlocking the next stage of exponential growth for SOL the asset.

[Umberto]

The issue I see is that we are confusing what “overpaying for security means”.

The current SOL staked value amounts to ~$53B, which is securing a TVL of ~$15B. Since the cost to control Solana is 66% of SOL staked, we have ~$35B securing ~$15B. That’s for sure an overpayment.

However, this scenario may change in the future and determining / claiming if the current curve is overpaying for security is not so easy.

To see this we can define a “Security Ratio” as TVL / Stake rate and checking for the condition

0.66 < TVL / SR

we can see how different growth assumption may change the overpayment statement.

For example, if we assume a growth of:

• SOL price growth of 20% a year - in 10 years, this means 1 SOL = $866.84
• TVL growth of 62% a year in the first 4 years and 10% growth in the following years - this means Solana DeFi will reach Ethereum TVL in 4 years

the current curve overpays for security for ~2 ys.

Of course, if we assume that SOL price appreciates faster than TVL, the current curve overpays for security even in 10 ys from now.

Now, you can play a bit with parameters to see under what conditions this overpayment holds here. However, this is not the point, since no matter what are the assumptions, no-one knows the future. The point is that this statement depends on how TVL over SOL price will change, and there are instances where the current curve is not overpaying for security,

@laine

While it is true that right now many smaller validators do not significantly rely on inflationary rewards or commission thereon, there are two scenarios that should be reasoned about

This is another misconception. Data shows that small validators do rely on commissions (cohort 3 - 0.05% < SOL Share < 0.5% - and 4 - SOL Share < 0.05%), with ~50% of each cohort having commissions > 0.

And as you said, it is important to understand what will happen if market conditions change. If you look at dynamical evolution of this behaviour (e.g. 50th percentile of commissions by cohort), you can see that medium-small validators (Cohort 3) just changed this around epoch 600 (Apr 9, 2024), and Cohort 4 is just lowering it now.

This is because APR from market dynamic just went up by a lot, allowing small validators to charge less and less commissions to attract stake.

Overall, I agree that Solana should join other ecosystem like Ethereum and Cosmos to have a dynamical emission based on network willingness to move capital. However, we should be aware of what data says, without moving to fast with claims not backed with data. Data rn point to a “too aggressive reduction” and actually we should have more data driven analysis with different models / comparisons.

[laine]

Thank you for this interesting analysis.

Regarding your claims about security in relation to TVL, while I think your approach starts from a sound origin it doesn’t account for the attacker’s ability to actually aquire or purchase the required SOL, or to stake it, with a 9% per epoch stake activation limit.

Assuming an attacker has no pre-existing staked SOL, as long as >33% of total supply is staked, they can never gain the supermajority. However even below that, there will be insufficient liquidity and increasing costs to aquire the required SOL.

Even if they already control some stake, with sybils perhaps even 10-15% of stake, they will struggle to achieve a 67% control. In this context I think the value of staked SOL securing the TVL could be 0.5 (7.5bn staked SOL to secure 15bn TVL) without major concerns for safety, the attackers cost to aquire the SOL will equal or exceed the TVL they seek to attack.

[Umberto]

I understand your point, however I still want to relay on the actual definition of security:

Cost to attack > Profit → the chain is secure.

Tomorrow majority of stake can coordinate an attack, and in a decentralized system you can’t assume this wont happen. In this scenario, 66% of SOL staked can be used to attack the network!

If you trust so desperately the current superminority will not hurt the network, decentralization is not needed. We can concentrate everything there and Solana will be faster.

So, when dealing with “security” we should always be sure the worst case scenario is covered!

[Umberto]

But in any case, my point was exactly on this.

When defining “overpayment” everyone has its own mental model. However, no one made a study to show how the mental model can change and under which condition it holds.

We did it with a very simple toy model because we wanted to address exactly this variability. But we focussed on several aspect related to SIMD228, and diving more in this topic was too reductive to provide a complete overview.

We should have validators running these kind of analyses, so there could be coordination and peer review of models. At the moment, the industry has a complete lack in this with few entities (like Chorus One) really involved in these kind of analyses.

Another important aspect when dealing with security is MEV. Is it Solana ready? I understand Max is working to improve it, but as far as I know there is no SIMD on this, so we should trust (?).

Indeed, despite the amount of MEV and fees depends on block proposals, and then from share of staked SOL, the relative gain per SOL depends from SOL staked.

In other words, a share of staked SOL of 1%, $S_1$, produce on average $M$ from MEV and fees. If fraction of SOL staked goes down, the same share of 1%, $S_2$, still makes $M$ from MEV, this time with $S_2 < S_1$. Since, $M/S_1 < M/S_2$, revenue for staked SOL increase. This behaviour is depicted in the image below - fixed share of staked SOL of 1%.

Above you can see the evolution of APR from MEV and fees for 1% of the stake. This is risky because you can now make more profits in relative terms from MEV. Put it simply, bad actors can accumulate SOL with discounts coming from unstaking.

Is that still making the current curve overpaying for security? I don’t know, but rn it is keeping SOL staked, making SOL accrual via different sources more costly.

[zantetsu]

As an addendum to my previous posts explaining my no vote on SIMD-228, I would like state that after further consideration, I have decided to ask my stakers to direct their stake-weighted fraction of my vote to their preferred option, and for those stakers who do not express a vote preference, I will vote NO as per my previous arguments.

The reason for this is that I believe that SIMD-228 is really best decided by stakers since it is the inflation schedule that they believed would be honored when they bought into Solana that is to be altered, and only they can decide if they accept this. For those who do not express a vote preference, I feel justified in voting NO because without knowledge of how they feel about SIMD-228, a NO vote at least retains an inflation mechanism that they previously indicated they were happy with.

[kankanivishal]

thank you for looking out for your stakers

[kankanivishal]

@Umberto

ying for security? I don’t know, but rn it is keeping SOL staked, making SOL accrual via different sources more costly.

thanks for sharing this analysis. yes, sol needs to be used more in different places.

[kankanivishal]

@alphagamma

considering there is no guarantee that those that stake will be instantly unbonding driven by yield

1. We have extended the rollout to 50 epochs instead of 10 epochs to smoothen the impact.

2. The drop should be considered from the day it is fully implemented and rolled out (which is likely a year out). Rationale below:

3. Even if SIMD-0228 passes, it is not gonna get implemented for the next 6 months.

4. It will have 50 epoch rollout

5. The right comparison for inflation should be inflation pre SIMD 96

6. Then decay it by 12 months.

7. That emissions number is roughly 3%, assuming it is even okay to benchmark with an arbitrary curve.

8. Stake % is bleeding. So, implied inflation from this curve would be ~1%, and implied staking rate would be ~1.7%

Above is the nominal yield impact. In fact, real yield impact would be even lower:

Today real yield: ~7% vs 4.7% = ~2.3%
Next year real yield post SIMD-0228: ~3% vs ~1.3%