Exploring Minimum Viable Issuance (MVI)

lang

url

1. Background

The Ethereum Protocol (Gasper) assigns the right to perform protocol work to entities on Earth, relying on these entities, known as Validators, to carry out their duties honestly. Stakers can earn incentives by performing their work faithfully. To become a validator, an entity must lock (stake) a specified amount of ETH. The penalty for dishonest behavior is the confiscation of this ETH (known as slashing).

Stakers have two types of work and corresponding rewards:

Work related to issuance rewards: This includes two tasks. The first is block proposal, where the staker performing this task is called a Proposer. Proposers earn 1/8 of the rewards given to attesters and MEV (including fees) by including attestations in the blocks they construct. The second task is related to attestation (voting), where rewards are earned for voting on correct source and target checkpoints, head blocks, and participating in sync committees. These consensus-related rewards are called issuance rewards because the protocol issues new currency to allocate them. The total weight of rewards for various elements like voting on the source adds up to 64, which is why the value of mentioned later is 64.

Work related to execution rewards: MEV (Fee) is included in the Proposer's execution rewards and is not an issuance reward. It's an additional income that only Proposers among stakers can acquire. Future expected additional income such as Preconfirmation tips will be included in MEV.

The protocol uses this carrot-and-stick approach to regulate validators, ensuring safety and liveness. It relies on an incentive design that guarantees validators will perform their work honestly and correctly.

If the protocol's issuance rewards are high, the yield for stakers increases, incentivizing more to become stakers and potentially increasing the total stake. As the stake increases (assuming the total amount of ETH circulating in the current market remains constant), it requires more resources from potential attackers, thus enhancing security. For example, the cost to cause a delay in Finality scales quadratically with time and is proportional to the total stake.

However, if the protocol is already secure, it may be paying unnecessarily high security costs through issuance rewards. Currently, about 33M ETH is staked out of a total supply of 120M, meaning approximately 1/4 (0.25) is staked. It's worth noting that staking is on an upward trend, with most new stakes reportedly driven by LSTs (Liquid Staking Tokens).

If the current security level is relatively high and the protocol is overpaying, how does this affect user utility? What is the appropriate stake amount to guarantee protocol security? This post aims to organize discussions about Ethereum's issuance reward issues and the concept of Minimum Viable Issuance (MVI), to promote dialogue about Ethereum's staking economy. MVI is the idea that the protocol should not issue more rewards than necessary for security, and it has been a crucial design principle for Ethereum since its inception.

2. Demand and Supply Curves

Let's first discuss the current demand and supply curves for Ethereum staking.

Assumptions

: Staked amount - The total amount of ETH staked, currently about 30M ETH (source: beaconcha.in)

: Circulating supply - The total amount of ETH currently in circulation, about 120M ETH (source: ycharts.com)

: Staking ratio - The ratio of staked amount to circulating supply.

: Reward curve constant - Used in reward curve calculations, typically

: Base reward factor - A parameter directly affecting the reward curve adjustment, currently

: Total REV amount - Annual total Realized Extractable Value (including fees), using last year's value of about 300K ETH

: Constants used to form the supply curve.

: Issuance yield - Yield that varies based on the staked amount under the reward curve (definition to follow)

: REV yield - Yield from REV (definition to follow)

: Total yield - The yield offered by the protocol to stakers, with both demand and supply sides (definition to follow)

Demand Curve

The demand curve represents the demand for staking. It shows how much reward the protocol allocates.

The demand curve is expressed by the total yield , which consists of and . It's the sum of issuance rewards and execution rewards, representing the current incentives.

This results in the following demand curve:

The blue curve represents

The yellow curve represents

The green curve represents

The yield from issuance rewards can be defined as:

Where, as mentioned earlier, and are constants.

Specifically, when are substituted, the value of is approximately 3%.

The yield from REV can be defined as:

This is calculated as the annual amount of extracted MEV divided by the staked amount. Specifically, when are substituted, the value of is approximately 1%.

These yields indicate that the issuance yield based on the current demand curve varies inversely with the staked amount. In other words, as more stakers participate, rewards decrease.

In this demand curve, an increase in the staked amount indicates a higher cost required for attacks. However, excessive security guarantees may lead to a decrease in user utility, so it's important to determine the ideal ratio of staked amount . When considering this, we need to consider the staking ratio , which takes into account the circulating supply .

The ideal staking ratio for optimal security is said to be in the range of 1/4 to 1/8 of all ETH. This is because, thinking in powers of 2, 1/2 is probably too high for security, while 1/16 is likely too low. The reasoning is based on whether there's still a significant risk that a single centralized entity could have enough funds to conduct a 51% attack, even with a certain percentage of ETH staked.

For example, it's said that a single centralized entity in the ecosystem (like an exchange) could collect about 5 to 10 million ETH. With 1/4 of ETH staked, this risk is more certainly mitigated. 1/8 staking is also fine, but going down to 1/16 should be avoided. Based on these considerations, it's mentioned that there should be at least the minimum issuance necessary for at least 1/4 of all ETH to be staked.

While there's room for debate about the ideal staking ratio, this post assumes that at least is likely sufficient for security guarantees.

If the staked amount increases beyond this ideal scenario, ETH holders end up paying for excessive security costs. The current figures are and , resulting in , which can be considered sufficiently secure.

However, the current issuance policy doesn't intentionally target a specific staked amount (staking ratio), so there's no specific mechanism to prevent the staking ratio from exceeding a certain threshold. The MVI approach aims to maintain an appropriate staking ratio by appropriately reducing issuance rewards, thereby decreasing staker incentives.

Supply Curve

The supply curve represents the supply of ETH for staking. It illustrates the willingness of ETH holders to stake at various yield rates and indicates the minimum incentive (or cost) required for participating in staking. This supply curve is considered in terms of the "Reservation yield."

The Reservation yield refers to the minimum rate of return at which stakers are willing to stake their ETH. It's the equilibrium point where stakers derive the same utility from both staking and not staking. If the yield falls below the Reservation yield, stakers will choose not to stake. However, since the Reservation yield cannot be directly measured, the exact shape of the supply curve remains unknown.

To account for this uncertainty, the supply curve is represented by an equation that covers a wide range:

Where are constants with the following values:

First term:

is a constant

represents the stake ratio raised to the power of

Since , this effectively takes the square root of

This term shows the supply elasticity for moderate values of d (e.g., )

It sets the supply elasticity to 2 at the midpoint (0.25)

c₁ is set so that when the staked amount

Second term:

This term becomes dominant when the stake ratio is very high (e.g., )

It causes the Reservation yield to increase rapidly at high stake ratios

This supply curve equation demonstrates that the first term is dominant up to moderate levels of , while the second term becomes dominant as approaches 1 after passing through moderate levels. Additionally, under the current issuance policy, as the stake ratio approaches 1, it indicates an increase in the number of stakers (while the circulating supply remains constant). As a result, the reward per staker decreases. Consequently, the minimum rate of return required to continue staking increases.

3. Staker Incentive Structure

It's beneficial to categorize staking incentives into surplus and costs.

Surplus: The total yield that stakers can acquire.

Costs: These include technical skills, hardware, protection of locked ETH, reduced liquidity due to locking, taxes, opportunity costs, and increased risk premiums. These costs act as barriers to staking.

Surplus

The supply curve, which represents ETH holders' willingness to stake, is determined by the total yield and costs. While the total yield is relatively uniform among stakers, the cost structure varies for different ETH holders.

The total yield can be further classified into endogenous and exogenous yields:

Endogenous yield: Revenue generated solely by participating in consensus through staking .

Exogenous yield: Revenue from DeFi and restaking due to LSTs is not included in the total yield. We'll denote this as .

This classification is useful for modeling what happens when the endogenous yield approaches zero.

For example, let's assume ETH holders are willing to lock their ETH if the total yield (including ) exceeds 0.04. Let be the exogenous yield obtainable by using non-staked ETH as collateral.

If and , ETH holders will choose to stake their ETH.

Therefore, if , the condition for ETH holders to stake is .

However, if , the required condition becomes .

This breakdown of staker surplus shows that staking will only occur if the surplus from locking ETH exceeds the surplus generated without locking.

Costs

Costs can be further categorized into initial costs and long-term costs:

Initial costs are those required to start staking, generally including technical skills, hardware, and capital for collateral.

Long-term costs include protecting locked ETH, reduced liquidity, taxes, and increased risk premiums. These can also be seen as ongoing commitments such as time spent on updates, additional internet connectivity, and accurate execution of protocol work.

Overall, these long-term costs are likely to have a greater impact on stakers' ongoing decision-making than initial costs. To stake, one can either stake independently or participate through delegation to Staking Service Providers (SSPs).

Solo stakers bear all costs themselves. For SSPs, we need to consider different cost structures for node operators and participants (ETH holders):

SSP node operators have fixed staking costs and receive a portion of participants' earnings as fees. In low-profit situations, they can increase efficiency by reducing the number of nodes and increasing the number of validators per node.

Participants only delegate their stake amount to node operators, so they incur no costs but pay a portion of their earnings as fees. Due to LST issuance, they may not feel the impact of reduced liquidity and might not perceive the fees as a cost.

Thus, while costs equally burden solo stakers and node operators, SSP users face lower costs. As SSP operators substitute the barriers to solo staking, users utilizing SSPs don't see a significant increase in costs for additional staking. Moreover, if they receive LSTs, liquidity increases, allowing access to additional yields without requiring long-term commitment.

As SSPs simplify the staking experience (cost structure) and improve staking liquidity, the inequality holds even as the endogenous yield decreases.

Given that the current issuance policy doesn't target a specific stake amount , it's expected that as the stake amount increases, the supply curve will shift downward and flatten due to improvements in exogenous yields and cost structure. This indicates that even if issuance rewards decrease, changes in the supply curve (i.e., Reservation yield) will be small, minimizing the impact on stakers. This trend encourages participation as stakers regardless of fluctuations in issuance rewards, potentially leading to the protocol issuing more rewards than necessary for security and forcing users to pay an infrastructure tax.

Image showing a flattening effect

The motivation behind MVI stems from concerns about the supply curve flattening out, potentially imposing excessive inflation on users from the perspective of staking surplus and costs. Therefore, MVI is an approach aimed at reducing the staking ratio to an issuance level that is sufficiently high for security purposes, but not any higher than necessary. Since the current staking ratio is likely secure, discussions about MVI often revolve around reducing issuance rewards.

MEV-burn

When considering issuance reduction, it's crucial to differentiate between scenarios with and without MEV-burn. This is because MEV rewards can fluctuate significantly. MEV-burn refers to the concept of burning a portion of MEV to stabilize relative rewards among stakers.

Generally, issuance rewards have an upper limit, while MEV doesn't. If we only reduce issuance rewards without addressing MEV, stakers rewards will become relatively more MEV-dependent. This increases pressure to pursue MEV, potentially disadvantaging solo stakers who may lack extraction capabilities.

A desirable issuance reduction proposal should provide appropriate yields for all stakers, even without MEV-burn. This involves a tempered issuance curve that slightly reduces issuance to limit stake growth. If stake amount remains high after implementing MEV-burn, a phased approach of further cutting issuance is considered ideal.

4. Advantages of MVI

Let's consider the advantages of MVI (Minimum Viable Issuance) from both micro and macro perspectives.

Assumptions

We'll use the following variables:

: Total annual rewards

: Implicit total cost for stakers

: Staker surplus

: Reduction in total cost for stakers

: Reduction in staker surplus

: Proportional return for stakers and non-stakers

: Annual issuance amount

Circulating supply

: Annual burn amount (using the average since The Merge)

: Inflation rate of circulating supply

': Change in utility

This framework allows us to analyze the economic impacts of MVI on both individual stakeholders (micro level) and the overall system (macro level).

Micro-level Benefits of MVI

All stakers are ETH holders and can benefit if the reduction in issuance increases the value of ETH.

The total annual reward for stakers is , and the total reward is .

The costs under the supply curve related to include the broader total costs of stakers mentioned earlier. The area above the supply curve represents the surplus .

Image showing a hypothetical supply curve in blue, indicating the amount of stake deposited at various yields (S=120M). Point 1 shows the expected for the next few years, and point 2 shows the curve with reduced .

The supply curve is drawn based on the costs for ETH holders to stake for maintaining security. This reflects the yield they think is worth staking at given their costs.

The green demand curve shows reduced issuance rewards. As long as Ethereum's security is guaranteed in the shift from 1 to 2, issuance reduction might reduce costs for users.

The changes in costs and surplus can be quantified as follows:

The change in cost is calculated by the definite integral of the inverse supply curve :

The change in surplus is quantified as:

These results suggest that moving the demand curve from 1 to 2 reduces costs by 446k ETH and shifts 252k ETH of surplus. While the surplus represents a loss in stakers' yield , the reduction in issuance is received by ETH holders in the form of a decreased issuance rate , distributing and .

The proportion of benefits transferred to ETH holders is:

This means that as stakers' costs are reduced, their surplus is reduced by 56% in proportion to costs. This reduction in surplus affects the overall surplus improvement for all ETH holders. As long as Ethereum's security is established when the demand curve moves from 1 to 2, the cost and surplus redistribution resulting from the reward curve shift benefits all ETH holders.

Classification of ETH Holders

Next, we consider the utility for ETH holders classified into 1) stakers, 2) de-stakers, and 3) non-stakers.

To calculate utility, we first need to determine the proportional yield considering inflation.

Reducing issuance decreases the issuance yield, lowering both the issuance rate and inflation rate . The inflation rate is , derived from the issuance rate and burn rate . A higher issuance rate means an increase in . Conversely, if more ETH is burned due to EIP1559, decreases. We consider , the approximate burn rate since the Merge.

As stakers' yield increases, the inflation rate increases. When rises, additional ETH is issued, reducing everyone's proportion. The annual change in someone's ETH holding proportion depends on each ETH holder's yield and the inflation rate , expressed as:

A simple approximation is , where is called the proportional yield considering inflation (real yield).

We compare two different equilibria where an ETH holder's real yield changes from (before issuance reduction, ) to (after issuance reduction, ). (We use to denote before and after for other variables as well.)

This comparison represents the following real utility change:

The utility here is cardinal utility, not ordinal utility. In other words, it considers not only the order but also the level of the utility function.

Image illustrating the utility changes

In conclusion, observing the issuance reduction by MVI in terms of utility change shows that all ETH holders can benefit, meaning MVI is not a zero-sum game. In other words, stakers, de-stakers, and non-stakers all have . Paradoxically, having a higher yield than necessary to guarantee security means ETH holders bear a high cost. Maintaining MVI allows all users to gain non-negative benefits.

1. For Stakers

Stakers lose some yield , but the inflation rate decreases similarly, so and becomes 0. This value is the same for all stakers.

2. For De-stakers

In calculating , de-stakers use . They unstake when the yield falls below their reservation yield, so there's no additional utility loss if the yield decreases below the reservation yield.

They lose some yield in the transition from to , but then the yield drops to , and they benefit from the decrease in inflation rate like non-stakers. In other words, they gain some utility after unstaking.

We will omit the detailed calculations here.

This function behaves linearly for (the de-stakers part in the above figure).

3. For Non-stakers Non-stakers have , so they're not affected by yield and receive all the utility from the decrease in inflation rate . Thus, the for non-stakers can be calculated as:

This value is the same for all non-stakers.

Macro-level benefits of MVI

Let's consider the macro-level impact of MVI.

In a scenario where most ETH is staked through SSPs, the improvement in user surplus and cost structure due to SSP usage suggests that if one LST becomes popular due to its currency functionality, it's expected to be utilized by more projects.

Moreover, considering additional revenue pressures to promote SSP adoption, such as airdrops (points) in RSTs (Restaked Tokens) and adoption by platforms like Swell and Puffer, the likelihood of continuously increasing stake amounts rises. In this situation, under the current issuance policy, the issuance reward per staker would decrease, making it harder to continue solo staking from a surplus/cost perspective. Therefore, users might prefer joining LSTs as they are more resistant to protocol reward changes compared to solo staking.

Image showing launch dates and stake amount changes for LSTs and RSTs

From a macro perspective, SSPs tend to follow a winner-takes-all trend. If one or a few LSTs become dominant as currency, they will penetrate all layers and applications due to their currency functionality. An economy driven by an untrustworthy currency, even if it doesn't threaten the consensus itself, goes against Ethereum's trustworthy neutrality.

Currency functionality refers to the medium of exchange for goods and services, store of value, and payment standard across globalized DeFi. Therefore, in a situation with a sufficiently high staking ratio and relatively low circulating supply, LSTs could surpass transaction volumes and maintain a dominant position as currency. If this winner-takes-all trend exceeds important consensus thresholds like 1/3, 1/2, or 2/3, SSPs can access greater benefits through MEV extraction, timing games, and censorship.

When utilizing such cartelization layers, a tragedy of the commons occurs where participation in SSPs is beneficial for users but not for the protocol. Since LSTs depend on their issuing entity, users and projects are at risk of being similarly affected in cases of smart contract issues, governance problems, bugs, fraud, or if regulations or sanctions are imposed.

Interpretation

The interpretation of the issuance policy is described.

Demand curve:

Supply curve:

Given the current forms of demand and supply curves, where are constants and are variables not directly controllable by the Ethereum protocol, we can only modify the demand curve by adjusting from a protocol perspective.

Therefore, what we can mainly control is the demand curve, especially the demand curve in cases where MEV burn is not introduced, and more specifically, when no other adjustments are made to the demand curve.

Overall, MVI can be viewed as "an optimization problem of through the manipulation of ". If is not optimized, SSP may simplify surpluses and costs, flattening the supply curve and potentially forcing unnecessary costs onto users. Additionally, discussions about end-game issuance levels such as tempering issuance or cut issuance are attempts to optimize by changing the form of the equation. Therefore, the current direction is focused on making more substantial changes to the demand curve, such as adding other terms to the demand curve equation like dividing by .

Consequently, through a set of hypothetical supply curves, we define a demand curve that optimizes the known trade-offs related to staking: long-term and short-term economic security, its costs, the composition of staking pools, reward volatility, trustless money, low currency dilution, and the resulting utility. This implies a reduction in issuance.

By reducing the issuance, the following results are expected:

At the micro level, there are gains for all roles: stakers, unstakers, and non-stakers

At the macro level, when comparing SSPs (Staking Service Providers) and solo stakers, SSPs are structurally advantaged as they are more resistant to changes in protocol rewards

In this way, reducing issuance may increase user utility and decrease the equilibrium amount of stake for both SSPs and solo stakers. However, since SSPs are structurally advantaged over solo stakers, the stake amount will not decrease in equal proportions. Solo stakers might unstake before SSPs and participate as delegated stakers instead.

5. Discussion

We will discuss the clarification and solutions to the MVI problem based on previous interpretations.

Clarifying Assumptions

MVI is an optimization problem of through the operation of , generally indicating a reduction in issuance rewards. It's beneficial to categorize the reasons for issuance reduction into two when discussing:

The supply curve may be flattened by delegated stakers, leading to an increasing trend in stake amounts, potentially forcing users to pay unnecessary costs.

As the supply curve flattens, delegated stakers may be led towards a winner-takes-all tendency, making it potentially more rational for solo stakers to transition to delegated staking.

These situations arise due to the following reasons:

Benefits: Improved liquidity through LSTs

Costs: Simplified cost structures such as reduced capital efficiency for staking (ability to stake with less than 32 ETH) and long-term commitments

One predictable result of MVI is that reducing issuance rewards may lower the equilibrium stake amounts for both delegated and solo stakers. This might be a positive trend as it could increase the utility for ETH holders.

However, delegated and solo stakers are unlikely to see their equilibrium stake amounts decrease in equal proportions. In reality, due to factors like reward volatility and access to additional revenue, people might shift from solo staking to delegated staking, attempting to compensate for the costs lost due to issuance reduction through restaking. It's also important to note that issuance reduction may relatively increase the attractiveness of REV rewards while decreasing the appeal of attestations.

Problem Clarification and Solutions

Overall, the main issues with issuance reduction appear to be:

Risk of validators not attesting accurately

Reward volatility due to MEV

Centralization vector towards SSPs and elimination pressure on solo stakers

1. Risk of validators not attesting accurately

Theoretically, when deriving yield, more than 2/3 comes from issuance rewards, with the remainder being REV. The setting of the basic reward coefficient F determines the proportion of yield from issuance, and as decreases, rewards from issuance decrease, relatively increasing the yield from REV.

For example, if issuance reduction, almost half of the rewards would come from REV generated by block proposals. aelowsson defines these proportions as: as a good state, as not a good state, and as a state to be avoided.

At , the incentive to attest works equally with REV, so attesters will likely attest honestly. However, in states like or where the proportion of issuance rewards is too small, the incentive to attest becomes smaller compared to REV, increasing the risk of validators not attesting accurately. As a result, as long as inactivity leak is not triggered, there's no cost for inaccurate attestations, making it rational to play timing games to maximize REV while avoiding slashing. This situation is not good for consensus. To address this problem, increasing penalties for non-participation seems preferable. For more details, refer to the discussions between aelowsson and vbuterin.

2. Reward volatility due to MEV

MEV creates volatility in staking rewards, increasing uncertainty. MVI considers MEV-burn an important step, addressing negative externalities by reducing REV reward volatility through burning a portion of MEV.

As mentioned in the interpretation, currently, what we can mainly control in issuance policy is the demand curve, especially the reward curve when MEV-burn is not introduced, and more specifically, F when no other adjustments are made to the demand curve. Therefore, discussing MEV-burn is slightly off-focus. However, we'll touch on it briefly to facilitate discussion.

MEV-burn models are discussed as follows:

Simple MEV-burn: MEV-burn in Block Auction

Execution Auctions (EAs, f.k.a. APS-Burn): Execution proposing rights are deterministically allocated to each slot in advance. The slot's execution proposer can purchase this right by bidding in the Slot Auction held 32 slots prior.

Execution Tickets (ETs): Execution proposing rights are not deterministically allocated. Proposers buy lottery tickets in advance, and winners are randomly selected from the ticket pool before each slot to gain proposing rights.

The simple MEV-burn model is introduced as ePBS, where block builders specify base fee and tips in PBS auction bids. About 2 seconds before the slot starts, Attesters observe the highest base fee among the bids, setting it as the minimum base fee for the Proposer's block. Only bids with base fees at or above this minimum are accepted for the Proposer, and the base fee is burned.

EAs and ETs, derived from this simple model, are introduced as Attester-Proposer separation. Both designs incorporate MEV-burn into the allocation process for execution proposer rights. EAs sell rights for 32 slots in advance through the Slot Auction model, allowing for reward smoothing and Preconfirmation benefits. ETs are redeemed at random future times, making it difficult for a single entity to acquire execution proposer rights for consecutive slots. However, ETs make Preconfirmation more challenging as the execution proposing right holder is unknown in advance.

From a reward volatility perspective, MEV-burn in Slot Auctions (EAs or ETs) results in more equalized revenue for stakers (reducing execution reward variance) compared to Block Auction MEV-burn. This is because bidders in Slot Auctions have less accurate knowledge of block value and only know the distribution their self-made blocks' values follow, leading to slightly lower bids accounting for risk.

3. Centralization Vector Towards SSPs and Elimination Pressure on Solo Staker

From a reward volatility perspective, is low-volatility while is high-volatility. As issuance reduces, the proportion of high-volatility validator rewards increases. Reward volatility affects validator pool size; larger pools can smooth out fluctuations by aggregating rewards from many validators, thus decreasing volatility.

High reward volatility makes income prediction difficult. Furthermore, as issuance reduction further increases volatility for solo stakers, it becomes more rational for them to quit solo staking and delegate instead. This might be somewhat mitigated by MEV-Burn.

A specific scenario of centralization vector towards SSPs and elimination pressure on solo stakers assumes that solo stakers will leave (before delegated stakers) due to issuance reduction. Some might argue that current solo stakers have already borne fixed costs, making their supply yield elasticity low. However, long-term costs, which require ongoing commitment, are likely more burdensome than initial costs. Therefore, solutions like MEV-burn need to be seriously considered. While diverging from direct MVI discussion, proposals aiming to create solo staker-friendly structures like Rainbow Staking might reduce the elimination pressure on solo stakers.

6. Summary of Opinion

1) Each hypothetical supply curve has one optimal equilibrium point, and the demand curve should intersect each hypothetical supply curve near this point. My question, orthogonal to MVI, is what the actual range of staking participation should be as a condition, and how to determine the demand curve equation for this optimal equilibrium point.

2) While MVI can increase user utility by adjusting the staking ratio, some issues arise. The first issue is strengthening penalties for fraudulent proofs, which seems solvable.

3) The next issue is how to reduce reward volatility, with MEV-Burn considered a crucial step in MVI. Block Auction MEV-burn has the drawback that builders have no incentive to bid 2 seconds before the burn amount is determined.

An alternative solution is Sealed Execution Auction, which changes the Block Auction execution method to a second-price sealed bid and burns the second-highest price. Being a second-price auction, builders no longer have an incentive to delay bidding.

Pre-sale Slot Auction-type mechanisms (EAs and ETs) are superior in reducing staker reward volatility. EAs have an advantage in implementation simplicity, while ETs are advantageous in decentralization but sacrifice preconfirmation. I currently think Sealed Execution Auction is particularly excellent.

4) Regarding the centralization vector towards SSPs and the disappearance of solo stakers, I'm bullish on rainbow staking proposed by Barnabé. While Rainbow Staking may not be the focus of current issuance policy discussions, I'd like to express my personal opinion.

Rainbow staking targets MVI and counters the emergence of dominant LSTs alternative to ETH. It introduces heavy and light services. SSPs and solo stakers are expected to leverage their respective strengths, enhancing overall competitiveness and diversity, and strengthening solo stakers' economic value and agency.

Professional operators are suited for providing heavy services due to their sophistication in many aspects (economies of scale, capital requirements, knowledge, reputation). Solo stakers, with their high preference entropy, are suited for providing light services as they surface protocol-specific signals.

A simple scenario predictable from rainbow staking is that competitiveness works separately in each tier, creating opportunities for solo stakers to gain more stake. While SSPs may become active in light services, the structure might be more solo staker-friendly than the current situation.

Currently, SSPs can perform all of Gasper's work including IL, but separating tiers could create a structure where preference entropy, low capital requirements, and ETH holders can support chosen providers without risk, potentially giving solo stakers means to compete against professional operators like SSPs.

Rainbow Staking is referred to as 2-D MVI because solo stakers can effectively participate under their own conditions. Previous MVI discussions were 1-D MVI. In 2-D MVI, instead of all participants considering a single staking reward and its conditions, the discussion becomes about how much to allocate to heavy and light services. Therefore, the current MVI thinking is expected to transition to heavy services, determine the demand curve corresponding to that staking ratio, and then decide the issuance amount for light services based on heavy service rewards.

Image showing a 2-D MVI with Rainbow staking

However, while Rainbow Staking is a method to explicitly characterize the economic attributes of various classes of service providers, it is not a method to clearly identify economic attributes. Even if we determine the optimal allocation to heavy and light services, if SSPs become dominant in each tier (especially in light services) through some means, the optimal allocation would no longer be meaningful.

Considering such outcomes, I'd like to consider the following questions:

Is it reasonable to promote the identification of service providers' economic attributes through Rainbow Staking? If so, does this distinction not contradict credible neutrality, and how would it be implemented? As discussed here, it might be achievable by assigning metadata to validators exceeding a certain value.

As a result of promoting the identification of economic attributes, can we characterize a structure that is more advantageous to solo stakers? For example, in the heavy service layer, we could lower (or completely eliminate) the Burn amount for solo stakers only, or distribute the Burned amount to solo stakers. In the light service layer, we could increase the weighting of solo stakers' preferences.

However, Rainbow Staking is a method to explicitly identify the economic attributes of service providers and solo stakers. If we could more clearly identify economic attributes through another method, we might be able to create a solo staker-friendly structure without implementing Rainbow Staking.

7. Conclution: Minimum Viable Issuance

MVI is a simple principle of not over-securing, and can be seen as a problem of optimizing by manipulating . If we do not optimize d, the supply curve may be flattened by the SSP simplifying surpluses and costs, forcing users to pay more than they need to.

Through a set of virtual supply curves, we define a demand curve that optimizes the known trade-offs regarding staking: long-term and short-term economic security, its cost, the composition of the staking pool, reward volatility, trustless money, low currency dilution, and the utility that comes with it. And that means issuance reduction.

Issuance reduction can have negative effects such as impact on consensus, reward volatility, and selection pressure for solo stakers, so increasing penalties, MEV-burn, Rainbow Staking, etc. are being discussed as methods of mitigation.

8. Reference

Google Presentation

Electra Issuance Curve Adjustment Proposal - Ethereum Magicians

Minimum Viable Issuance - Ethereum Notes

Properties of Issuance Level Consensus Incentives and Variability Across Potential Reward Curves - EthResearch

FAQ: Ethereum Issuance Reduction - EthResearch

Reward Curve with Tempered Issuance - EIP Research Post

Endgame Staking Economics: A Case for Targeting - EthResearch

Risks of LSD - Ethereum Notes

Unbundling Staking: Towards Rainbow Staking - EthResearch

MEV burn—a simple design

Burn Incentives in MEV Pricing Auctions - EthResearch

MEV-Resistant Dynamic Pricing Auction of Execution Proposal Rights - EthResearch

How Optional Non-KYC Validator Metadata Can Improve Staking Decentralization - EthResearch

Execution Auctions as an Alternative to Execution Tickets - EthResearch