Interpreting "Minimum Viable Issuance" (MVI): Seeking a Balance Between Security and Inflation in ETH

Original Title: Minimum Viable Issuance

Original Author: Anders Elowsson

Original Compiler: 深潮 TechFlow

Introduction

I believe it is crucial to achieve "Minimum Viable Issuance" (MVI), which is an important commitment to the average Ethereum user. Staking should ensure the security of Ethereum, rather than becoming an inflation tax that reduces utility and liquidity, leading to oligopoly risks.

Ethereum is continuously evolving and may drive the global financial system in the future. We must assume that the "average user" will have a similar understanding of how Ethereum operates as the average person has of the current financial system.

Of course, we cannot assume that the average user will be driven by some ideology, as was the motivation for Ethereum's initial creation. Our job is to ensure that the right incentive mechanisms are in place so that Ethereum can develop unhindered.

An important design principle that has existed since the inception of Ethereum is "Minimum Viable Issuance" (MVI), which states that the amount of ETH issued by the protocol should not exceed the amount strictly necessary for security. This principle is reasonable under both Proof of Work (PoW) and Proof of Stake (PoS).

Under PoW, the role of MVI is to prevent miners from charging ordinary users excessive inflation taxes. Therefore, the block reward was reduced from 5 ETH to 3 ETH, and finally to 2 ETH.

Under PoS, the MVI principle should also be adhered to, ensuring that ordinary users are not charged excessive inflation taxes. Ordinary users should not have to worry about the details of staking to avoid having their savings eroded or supporting potentially censorious validator sets, etc.

Thus, MVI is essentially about maintaining the staking rate (the proportion of all ETH that is staked) at a sufficiently high level, but not higher. In this article, I will attempt to clarify why issuing more than the "minimum viable amount" would reduce the utility of Ethereum.

Benefits of MVI for User Empowerment

For individuals, participating in staking comes with various opportunity costs. It requires resources, focus, and technical knowledge, or it requires trust in third parties, while also reducing liquidity. Liquid staking tokens (LSTs) are not as reliable as native tokens and are less suitable as currency or collateral.

Therefore, individuals wish to earn returns through staking. They define their minimum expected return as the lowest yield they are willing to stake (using their best staking method). Thus, Ethereum's (inverse) supply curve derives from the minimum expected return of future ETH holders.

The reservation yield of holders can be described as the "indifference point," at which the utility they gain from staking is equivalent to not staking. This means that reducing the issuance actually increases everyone's utility, including stakers, as long as Ethereum remains reliable and secure.

Consider a hypothetical supply curve with an elasticity of 2 (in blue). In this example, I set it so that when the amount staked D reaches 25 million ETH, the yield reaches 2%, meaning that when staking 25 million ETH, the minimum expected yield for the marginal staker is 2%.

The real supply curve is a rather complex phenomenon, and we have not yet reached a balance point where we can anchor its position, but we will start from this simple and fairly realistic scenario. We will also ignore the complexities of compounding.

The burn rate b is set at 0.008. This represents the proportion of ETH burned as a percentage of total supply, expressed on an annualized basis since the merge. But this is not the key point, as we focus on the transformation of supply and demand between the mid-term balance point (circle), rather than the drift of total ETH supply.

Realized Extracted Value (REV) (slightly over 300,000 ETH per year) has been added to the protocol issuance to form the black demand curve (current policy) and the green demand curve (halving the issuance by reducing the base reward factor F from 64 to 32).

Halving the issuance reduces the yield y (red arrow). This reduces the issuance yield yi = y - yv (where yv is the yield from REV), thereby lowering the issuance i = yid and the circulating supply inflation rate s = i - b (orange arrow).

In one year, the change in the proportion of circulating ETH holdings P that someone can obtain depends on s and each holder's yield y, according to the formula: P = 1 + y / (1 + s) - 1.

The current issuance policy gives P1, and the halved issuance gives P2, so their ratio is: P′ = 1 + P2 / (1 + P1) - 1.

Define the relevant change in utility as u′ = P′, but when calculating P2, for those who stop staking, use their respective minimum expected yields. Below that yield, they would not have staked in the first place, so as yields further decline, they do not suffer additional utility loss.

According to this definition, everyone will gain higher utility under the new balance point. Stakers may see their yields decline, but the decline in the supply inflation rate is greater, allowing them to gain a larger proportion of ETH.

Of course, non-stakers will clearly fare better, as the only change for them is that the ETH issued to stakers has decreased. Those who stop staking are the only participants who see a reduced proportion of circulating ETH under the new balance point.

Despite the friction, their situation still implicitly improves due to the increase in utility. For example, marginal stakers at the old balance point would have been indifferent to staking, so they could stop staking and gain the utility improvement from the full supply inflation reduction.

When those who stop staking find themselves in between, still benefiting from the decline in inflation but suffering some losses in yield until they become indifferent to staking and unstaking. We have illustrated that, from a utility perspective, the issuance policy is not a zero-sum game.

Moreover, any group that gains utility increases will also generally benefit all token holders.

As long as one holds the underlying ETH, everyone can benefit from MVI. This does not include CEXs and other staking service providers (SSPs) that profit from staking fees. They will not benefit from the reduction in supply inflation and will seek to maintain high yields to sustain high reductions.

However, issuance above MVI forces unwilling stakers to endure utility declines when staking or suffer economic consequences when not staking. Under the real supply curve, even willing stakers will find the situation worsened. Note that this example does not even consider tax implications.

For a PoS cryptocurrency with a yield of 5%, where everyone stakes, and the average tax on staking yields is 20%, one would spend 1% of their market value on taxes each year. This is higher than the amount that will dissipate to miners after the next halving of Bitcoin.

The argument does not necessarily depend on users' views on tax levels or how they interpret staking yields. We can still conclude that by implementing MVI, Ethereum remains more neutral regarding the differences in tax policies between nation-states.

It can be argued that PoS requires lower rewards to achieve the same level of security as PoW, and it is crucial to leverage this to maximize user utility. For example, with a yield of 2% and staking 25 million, the total reward is Y = 0.02 × 25,000,000 = 500,000 ETH.

Maintaining this solid secure "return rate" is approximately r = Y/S = 0.4%, a surprisingly low figure. We fully leverage this to maximize utility for users. The potential balance of the current issuance policy is represented by the black circle.

With a yield of about 3%, staking 50 million ETH, i.e., Y = 1.5 million ETH/year. The difference in rewards of 1 million ETH (over 1 billion dollars at current token prices) can be rewarded to Ethereum users without diluting token holders.

For MVI, an average extraction of 15% in staking fees would provide CEXs and SSPs with about 250 million dollars in excess profits annually. Some will be passed on to corporate shareholders, and some may be used for lobbying to keep yields permanently above MVI.

Macroeconomic Benefits of MVI

I often think that Ethereum's penetration into the ecosystem is desirable. In terms of L2, bridging ETH binds L1 and L2 together and provides external funding for users on L2, enhancing their financial security.

If in the system you create, users must rely on some obstinate ETH derivatives as funding to avoid inflation taxes, then the entire ecosystem becomes more vulnerable to disruption.

For example, consider the following scenario: users who cannot stake provide their ETH to organizations (SSPs) that run validators for them. These organizations can issue LSTs as collateral and use them on Ethereum.

If the protocol does not operate under MVI but runs at a higher deposit ratio, one or several LSTs may replace currency in the Ethereum ecosystem, embedding it into every layer and application. What impact would this have?

First, the positive network externalities brought by the currency function may allow LSTs to maintain dominance, even if the services provided by their SSPs are worse than competitors (e.g., charging higher fees or only offering poorer risk-adjusted rewards).

Second, and most importantly, LST holders and any applications or users needing LSTs to maintain value will share a common fate with LSTs and ultimately the LST issuing organizations (SSPs).

This requires Ethereum to destroy a significant portion of itself. Affected users may be more willing to reinterpret errors or misconduct as entirely different things. Once you become the currency of Ethereum, you become, to some extent, part of the social layer. We no longer only care about the proportion of staked ETH under LSTs, but rather the proportion of the total ETH under LSTs. Corrupt institutions are correspondingly also positioned one layer above the consensus mechanism.

It is evident from The DAO that if the proportion of the total circulating supply affected by the results becomes large enough, the "social layer" may waver in its commitment to the potential expected consensus process.

If the community can no longer effectively intervene in events such as a 51% active attack, then risk mitigation in the form of a warning system discussed by Buterin may not be effective.

In this case, the consensus mechanism becomes so large and interconnected through derivatives that its ultimate arbiter—the social consensus mechanism—becomes overloaded.

Now consider the different situation under MVI. First, each LST will face tougher competition from non-staked ETH. Therefore, the ability to monopolize the currency function and then charge exorbitant fees or offer higher-risk products will be weakened.

Second, the social layer will continue to be inherently bound to Ethereum and native ETH, rather than to external organizations and their issued ETH derivatives. By keeping the staking rate at a sufficiently low level through MVI, it changes the risk calculations of participants.

Under MVI, when the staking rate is low enough to prevent moral hazard from developing, delegating staking to the dominant LST's delegators—the principal-agent problem (PAP)—can be priced more accurately. No LST will grow to the point of being "too big to fail" in the eyes of Ethereum's social layer.

This pricing will reflect the fact that the larger the share of staking controlled by agents acting on behalf of delegators (or any party able to intervene in that relationship), the better their chances of profiting by degrading consensus.

Delegated stakers must always consider what security guarantees they have (e.g., the value risk of the staking agent or intervening party), knowing that if the worst-case scenario occurs, they could lose everything.

Removing Ethereum's direct dominance as currency and assuming that under MVI the deposit ratio has grown to a scale that maximizes utility, larger SSPs are likely to find non-monopolistic strategies more profitable (i.e., increasing fees).

This is merely a comment relevant to the present. But importantly, it reflects the fact that for every "cartel layer" we can eliminate, the value proposition of secure and value-consistent SSPs increases relatively.

An important step towards MVI is MEV burning, which may also have the potential to eliminate "cartel layers" that are more significant than currency functions. MEV burning helps reduce the reward variance for independent stakers, which would increase if the issuance yield decreases.

It also brings greater precision to targeting MVI, as it eliminates a source of income that may change in unpredictable ways over time.

It is worth noting that various methods may be adopted in the future to address certain aspects of the principal-agent problem in delegated staking (i.e., one-time signatures). However, the fundamental issues of building trust, monopolistic incentives, and censorship capabilities may be difficult to escape.

Another benefit of MVI is that it improves the conditions for (independent) staking, which relates directly to the scale of collateral, the number of validators, and the size of validators. If the scale of collateral changes, the size or number of validators (network load) will also change.

This effect will ripple through the entire protocol design space and will affect any targets that may be substituted for higher or lower network loads, such as parameters related to variable validator balances.

This is a fundamental property of the current consensus mechanism. If the issuance policy leads to a mid-term balance point of d = 0.6 instead of d = 0.2, independent staking will require three times the ETH to maintain the same network load, all else being equal.

At its core, I believe the most important benefit of MVI is its ability to provide utility to ordinary users. Ethereum is in a unique position to make native cryptocurrency a global currency, and I believe this is an opportunity worth pursuing.

When nations implement price inflation by increasing the monetary base, they control the time preferences of ordinary people and believe that this control remains viable in a digital and globalized world.

Ethereum should not control ordinary people, nor should it force them to save liquidity. We should allow them to maximize the convenience of using Ethereum currency and derive utility from it. The "risk-free rate" in Ethereum is simply holding (and trading) ETH.

Addressing Potential Issues with MVI

After outlining the potential benefits of MVI, the second part will address some suggested drawbacks. These include reduced economic security and the notion that if we lower yields, delegated staking will replace all independent staking.

When it comes to the first point, this is indeed correct, as higher deposit ratios do force attackers to spend more resources, such as restoring finality. This is not something to be taken lightly.

Our goal is not "minimum issuance." We must always ensure it is "viable." Buterin provided some intuitive explanations of how expensive a 51% attack on Ethereum should be.

We can also consider that the nearly 14 million ETH that ensured Ethereum's security at the time of the merge is the "preference" for the collateral scale that the ecosystem believes is sufficiently secure under the current consensus mechanism (in terms of resisting bad actor attacks, not just super committee accountability).

At the same time, having a considerable margin is indeed nice; the current staking rate (d ≈ 0.2) relative to the staking rate at the time of the merge (d ≈ 0.1) may also provide a meaningful enhancement in resisting bad actor attacks.

The slope of the reward curve cannot be too steep, which is why we may want to operate at a certain distance from the preference point and ultimately determine d from the probabilistic analysis of staking supply and demand.

Some may argue that delegated staking somehow makes the allocation of attack resources easier, which is merely "apparent" security. However, by penalizing all staking and eliminating moral hazard (through MVI), delegators must be very careful when delegating staking, as mentioned earlier.

In this setup, the market determines the appropriate capitalization ratio for staking operators and prices the risks of staking. In contrast, Ethereum is responsible for penalizing misconduct and maintaining the value of ETH relative to the value it secures.

By ensuring that ETH tokens penetrate the real economy and that all consensus participants have real stakes, we set a price for attacks that is harder to evade through financial engineering.

I mention this because there are indeed some interesting alternatives being discussed, where Ethereum intervenes in the delegation process, and delegators bear no risk. Then the risk for delegators contributing to consensus degradation would be much lower.

Or at least it seems so. When Ethereum forks and/or must be saved through social intervention, if the worst-case scenario occurs, risk-free delegators may be surprised at how the social layer assesses their delegation and the damage they are perceived to cause.

Here I return again to Buterin's request not to overload consensus. My point and a theme of this article is that when the proportion of ETH involved in the consensus process is very high, everyone becomes involved and may not achieve a "neutral" outcome.

The conclusion to the first issue is that d under MVI must remain sufficiently large to ensure security; delegation does indeed reduce security to some extent, but as long as their staking is at risk, parties will strive to assess risks and delegate wisely.

Retaining independent stakers is indeed a complex dilemma. Economies of scale are difficult to design out, and we have not paid enough attention to liquidity in staking. However, there are some nuances in the current argument that favor MVI, which I hope to bring up.

Independent household stakers incur certain costs when staking. They prepay a significant portion of the costs, including knowledge acquisition. They also incur variable costs such as bandwidth, troubleshooting time, and interruption risks.

Many of Ethereum's SSPs incur significant costs in designing their services and bear other types of operational costs that independent stakers do not have to worry about. However, they rely on economies of scale to reduce the average cost of operating validators.

We must assume that SSPs seek to maximize profits and can consider what their fees might be under different balances. What is the difference in economies of scale between d = 0.2 and d = 0.6? It seems reasonable to assume that the average cost for SSPs is much lower at d = 0.6.

Remember that at d = 0.2, individual stakers may be able to run validators that are three times smaller than those at d = 0.6. In terms of the proportion of individual stakers we can attract, the minimum validator amounts of 32 ETH and 96 ETH (or 11 ETH - 32 ETH) may differ.

Thus, not only will a higher d force independent stakers to have more ETH under the same network load, but they will also have to compete with SSPs that can charge lower fees. While fees will be set according to market strategies, average costs should ultimately be significant.

If we lower yields, SSPs will want to increase fees to adequately compensate and amortize costs. The costs for delegated stakers are variable, including PAP and fees. They can easily shed increased costs.

The argument that lowering yields will drive independent stakers away (earlier than delegated stakers) should be taken seriously. However, since current household stakers have incurred fixed costs, their current personal yield supply elasticity may not be high.

However, if we lower yields to the point where independent household staking becomes unfeasible (including for newcomers), their lower elasticity in the short term will be of no help. If we want to maintain independent staking, there exists a lower bound on total staking yields that we cannot fall below.

Assume the total cost of independent household staking (in ETH) is C, and consider the annual risk R of funds when staking, among other factors. Then, yields must be above y > C/32 + R; even if re-staking brings liquidity, reasonable margins are still required.

Here, I also want to discuss the impact of DeFi yields. All stakers will receive yields intrinsic to staking y. This "intrinsic yield" comes from issuance, MEV, and priority fees. Some may also receive "extrinsic yields" yc outside the consensus mechanism.

It is not simply a matter of summing y + yc for LST holders and concluding that regardless of how y declines, LST holders always profit relative to independent stakers. It is expected that ETH tokens will provide higher utility relative to LSTs (not considering their intrinsic yields).

Delegated stakers must weigh y (1-f), where f is the percentage fee, against the risks/costs including PAP and the inherent disadvantages of LSTs relative to native ETH, and only decide to stake when y (1-f) exceeds these costs.

When y = 0, agents will not delegate staking. They can obtain better liquidity or higher yc through native ETH and face severe adverse choices by delegating staking to SSPs operating at a loss. Independent stakers may also choose not to stake.

For those who want to hold ETH regardless, the decision may not depend on whether yc is 1% or 5%. At 5%, it can be expected that ETH will provide +5%. Of course, that 5% comes with risks and is not free currency (our yields should not be, hence MVI).

As y rises, potential independent stakers and delegated stakers will gradually find staking claims worthwhile, starting with the most ambitious/risk-taking individuals. Here we are forming a supply timeline where each agent makes decisions based on their specific circumstances.

How the distribution of minimum expected yields between potential independent stakers and delegated stakers looks is still unclear. At the mid-term balance point of d = 0.2, the proportion of independent stakers may be lower than at d = 0.6, but another option is also likely.

A higher d may allow SSPs to diversify more, but the cartel layer of the currency function exerts pressure on this. The proportion of individuals with enough ETH for independent staking is also limited, setting a soft upper limit on the total number of independent stakers.

This is indeed a topic worth further research. The key is that the opportunity cost of staking must always be fully considered, and economies of scale and monopolies will affect the underlying equilibrium analysis in quite complex ways.

Finally, re-staking has the potential to make independent stakers more competitive. It allows them to "re-stake" their staking when they wish (however, if they want to provide economic security, they may also encounter principal-agent problems themselves).

One benefit of re-staking is that if active validation services (AVS) can quantify decentralization, it can also endow decentralization with economic surplus value. This is something Ethereum as an open protocol cannot achieve.

The previous arguments also apply to re-staking on the protocol-external EigenLayer functionality. At very low yields, users would be better off using non-staked ETH directly (free staking). For many use cases, it seems reasonable for AVS to prefer a token that will not easily evaporate.

It is also worth noting that if PEPC expands its scope beyond "block production use cases," the yields generated may become more intrinsic, depending on the remaining utility provided.

Looking Ahead

The discussion on the merits and drawbacks of MVI concludes here. While there are some concerning issues with independent staking, MVI is fundamentally a reasonable design policy that gives Ethereum a real opportunity to provide users with the best digital currency ever.

Each argument has its nuances, and some discussions cannot be succinctly expressed in tweets. However, I believe that considering all factors, it should be possible to accept that MVI is also a favorable design principle under PoS.

We must always focus first on the "average user," which requires research based on microfoundations, and assess how we can maximize utility for ordinary people as Ethereum (hopefully) becomes their new financial system.

So the question is how we achieve MVI, which is a question I have been delving into. Dietrichs mentioned the importance of conveying current issuance policy research in a recent developer call, and my process began with this tweet.

Changing the issuance policy is a sensitive issue. What we desire is an issuance policy that maximizes utility without requiring further developer intervention, so that it can forever proportionally allocate MVI that maximizes utility.

However, the current reward curve does not allow the protocol to influence the staking rate (security), but rather the scale of collateral. In the mid-term, the results are closely related to both, but in the long-term balance, there may be significant divergence as circulating supply drifts.

This is the theme of my article in Ethresearch in 2021 and my talk at Devconnect: defining how circulating supply S drifts towards balance (i = b) so that we can improve the reward curve and achieve minimum viable issuance under PoS.

According to the current reward curve, the issuance i can be expressed as i = cF√d/√S, which changes with the circulating supply (the staking rate d provides some room for adjustment). The illustration shows Ethereum's issuance rate diagonal and the average b since the merge.

The burn rate b will not depend on circulating supply—the demand for block space does not change due to changes in monetary units. If i > b, S will rise and pull i down until it equals b. If i < b, S will fall and pull i up until it equals b.

In 2021, stakers did not have REV, so I directly used the minimum expected yield y- to derive Ethereum's security as d = b/y.

Today, we simply add the "REV rate" v to the equation, yielding d = (b + v) / y. This point indicates that we cannot control the staking rate and security in the long term unless we are prepared to change F from time to time.

We can cut F as a temporary solution to avoid paying excessive security costs (which will be discussed in the next tweet). However, Ethereum will ultimately return to the same long-term balance staking rate under lower circulating supply (all else being equal).

This is why we ultimately want to change the reward curve to be related to d rather than D. Then simply replacing D with S0d (where S0 is the current circulating supply) seems tempting. This moves us a step closer to autonomous issuance policy, but it still does not guarantee its realization.

Assuming there is MEV burning, then the protocol can fully adapt to income changes, but it still cannot adapt to permanent shifts in expected yields, i.e., the supply curve. This can be addressed by allowing the entire reward curve (demand curve) to drift slowly.

The ultimate goal is a dynamic balance where circulating supply can change at a constant rate without external influences. Whether it is inflationary or deflationary depends on the supply curve and how the value of block space is reflected in the market value of ETH.

Thus, we achieve what Polynya calls "constant" security, which I believe aptly describes our ultimate goal: to ultimately take control of issuance away from developers and enable Ethereum to achieve autonomy under MVI.