EIP-8363 Quantitative Review: What does Ethereum want to regain by cutting the staking "subsidy"?

Author: Mario Chow, IOSG

EIP-8363 Proposal: As the staking rate rises, an increasing portion of validator rewards will be burned, reaching 100% burn when 50% of the supply is staked. This article models its impact on issuance, yield, and staking equilibrium; examines whether ETH's yield truly explains its price; quantifies how much of the on-chain economy genuinely relies on this yield; and presents our conclusions.

All calculations are based on on-chain data and the original text of the EIP. The model is independently constructed, with a match rate of less than 2% with publicly available third-party data. Data updated to August 24, 2026. One-sentence version of the argument

Fee burning is already dead, making issuance the only leverage Ethereum still holds over ETH supply. This proposal halves issuance at the current staking level, rather than reducing it to zero; and it is self-limiting: under any reasonable yield threshold for stakers, the system will ultimately stabilize with 26-34% of the supply staked and an issuance of 0.3-0.5% per year. Meanwhile, the yield it cuts shows no detectable relationship with ETH price.

1. The burning mechanism is no longer effective

EIP-1559 burned 1.48 million ETH in 2022. EIP-1559 burns the base fee, which is essentially congestion pricing; once blobs move rollup data off L1 and the gas limit is raised, congestion disappears: gas usage doubles while the average base fee drops by 96%, and the amount burned has decreased by 98% since 2022. In the past twelve months, it has burned a total of 25,660 ETH, with an even lower running rate in the last 30 days: 39 ETH per day, annualized to about 14,300 ETH.

▲ Daily fee burning amount of EIP-1559 by year: from 8,844 ETH per day in 2021 to 57 ETH per day in 2026------far below the current issuance curve and well below the maximum issuance curve under EIP-8363. Compared to the total issuance of about 1.08 million ETH per year, the burning currently only offsets 2.4% of new supply. As a mechanism, "ultrasound money" has ended. L2 migration and blob expansion have moved the fee base away from L1: during the same period, L1 gas usage actually doubled (from 3.4 billion to 6.7 billion units), while the average base fee dropped from 4.00 gwei to 0.17 gwei: so this is a price effect, not a demand effect. Net issuance: What has happened to the supply? Burning is only half of the ledger. When viewed alongside issuance, the picture is more severe: issuance has never stopped growing, while the offset has directly disappeared beneath it.

▲ Comparison of ETH minted on the consensus layer monthly since the merge and ETH burned by EIP-1559. The issuance bar steadily grows; the burning bar shrinks to nearly zero by 2025. In the 47 months since the merge, only 13 months have been deflationary: the last being March 2024. ETH has been in an inflationary state for 28 consecutive months, and this rate has roughly tripled during this period, from +0.26% per year to +0.87% per year. The reason is not how much issuance has increased (only 4% since 2024), but that the offset has gone to zero.

This reconstructs the entire debate. EIP-8363 is often said to be a choice between staking yield and monetary scarcity. But a more accurate understanding is much narrower and more forced: the issuance policy is now the only remaining leverage Ethereum has over ETH supply, as the demand-driven one is no longer working. Regardless of legislation, all supply issues must now go through the issuance curve.

2. What exactly does EIP-8363 do?

Before dissecting the mechanism, we need to clarify the official core motivation: to defend network security. The proposal's authors believe that once the overall staking rate crosses the 50% red line, Ethereum will lose its ability for "social layer defense" and face systemic parasitic risks from LST oligarchs too big to fail. Therefore, this proposal attempts to lock the staking cap by enforcing a rate cut. However, grand security philosophies often obscure the real flesh and blood on the ledger. Setting aside metaphysical debates about decentralization, what does this mechanism really mean in the real on-chain economy? Here is a purely quantitative deduction.

How is the money deducted? (Core mechanism)

• First earn, then deduct: Validators initially earn the full rewards for their tasks, but then the system will "burn" a portion of the rewards directly at a certain ratio (assumed to be b).

• Deduct based on "theoretical full score," no double punishment: The key here is that the system calculates the burn amount based on the theoretical full rewards you should receive, not the actual rewards you receive. Why do this? Because if you accidentally go offline, you wouldn't earn rewards anyway; if the system deducts based on your actual situation, it would be too unfair to those who are offline. By deducting based on "theoretical value," it ensures that everyone's motivation to work remains unchanged, and those who are offline won't be punished twice.

• Extreme case protection: If the Ethereum network encounters serious issues (entering inactivity leak state), the burning of rewards for proof will be paused.

• Snapshot taken externally: This proposal only affects the rewards on the consensus layer. The "extra income" you earn by running a node: that is, MEV and Priority Fees, not a penny less, is completely unaffected.

Two common misunderstandings in the community Misunderstanding 1: "Ethereum's issuance will be directly cut to zero"

• Truth: Not at all. The current staking amount is about 42.2 million ETH, and at this level, the burn ratio b is 58.6%.

• To bring issuance to zero, the staking amount would need to soar to 60.25 million (43% higher than now). So the accurate statement should be: this proposal at this stage only cuts issuance by about half, and it is still far from complete zeroing.

Misunderstanding 2: "Yields will plummet instantly, triggering a DeFi collapse on the first day"

• Truth: The official design includes an 18-month "soft landing" period, making the first day almost imperceptible.

• To prevent an instant shock, the proposal will double the base parameters (base reward factor) for calculating rewards when it first goes live to 128. This doubling operation just fills the 58.6% burn amount mentioned earlier.

• In other words, on the first day of the upgrade, the net issuance across the network can still be maintained at about 83% of the current level. Then, over the next 18 months, the parameters will gradually return to the normal 64, and the issuance will slowly slide down to the current 41%.

• Summary: The decline in yield will be gradually diluted over a year and a half, not a sudden drop overnight. Those worried about "instantaneously bursting the DeFi bubble" are actually ignoring this buffer mechanism.

3. Baseline: Where is Ethereum now?

Supply dynamics

Where does issuance come from? All comes from staking rewards. After the merge, the new ETH has only one source: the consensus layer pays validators and distributes it according to fixed weights (denominator 64) across three categories of responsibilities: proof 54/64 (84.4%, 911,672 ETH/year), block production 8/64 (12.5%, 135,063), sync committee 2/64 (3.1%, 33,766).

Issuance is modeled as I(S) = 940.9 · √(S/32) ETH/year, which is the protocol's own reward curve. At S = 42.2 million, the corresponding consensus layer APR is 2.560%. The actual measured priority fee was 2,623 ETH in the 23 days before August, annualized to 41,500 ETH: equivalent to 0.098% on the staking base. Adding both gives 2.658%, which is almost identical to the published 2.66%.

Based on the measured priority fee, at least 96% of validator income comes from issuance, and at most 4% comes from fees. Payments from proposers exceeding direct priority fees in MEV-boost are not captured, so the fee proportion is a lower limit. In any case, issuance dominates absolutely, and this proportion is the key to the entire debate.

4. Modeling the proposal

The earliest attempts were to design an entire network around "hiding," rather than patching an existing network. There are two coins leading this path, but the bets are completely opposite. The third case is designed for banks, not individuals, but belongs to the same family. Applying at current staking levels, without considering behavioral responses

Issuance reduction: −58.6%. Staking APR reduction: −56.4%. Removed dilution: 633,000 ETH/year = $1.55B/year = 0.53% of ETH's market value per year.

▲ The relationship between annual ETH issuance as a proportion of supply and staking rate. Today's curve rises steadily; the EIP-8363 launch curve peaks at about 1.0% near a 20% staking rate; the permanent curve peaks at about 0.5%. Both curves drop to zero at a 50% staking rate. Complete curve (after full transition)

Issuance peaks at around 25 million staked, about 0.505% of the supply, then declines------consistent with the EIP's own statement. Equilibrium------the number that truly ends the debate Stakers are not passive. If the yield falls below their required return, they will exit, which will both push up the gross APR and lower b. Solving for the fixed point:

▲ The relationship between total staking yield and the amount of staked ETH under current rules and EIP-8363. The EIP-8363 curve intersects the 2% threshold at 31.2 million staked, and intersects the 1.25% threshold at 40.9 million.

Reading this table against the two loudest claims in the debate:

• "Issuance will go to zero." This only holds if marginal stakers are willing to work for about 0.5% returns. Under any reasonable required return, ETH remains at 0.3-0.5% annual inflation. Supporters have exaggerated.

• "Staking will collapse." At the 2% threshold, the staking rate will stabilize at 26%: lower than today's 35%, but roughly the level for the entire year of 2024. Critics have also exaggerated.

This mechanism is designed to be self-limiting. This is the most interesting property of the design and the least discussed point.

5. Can staking yield explain ETH's price?

First, address the "problem behind the problem" Is there a correlation between staking rate and yield? Yes: completely correlated, and it is determined by definition, not observed. This must be clarified first, as it determines what the data can and cannot explain.

The reward pool paid by the protocol scales with the square root of the staking balance, so the yield per ETH has a closed-form solution:

issuance(S) = 940.9 · √(S/32) ETH/year APR(S) = issuance(S)/S = 166.28 / √S The more ETH staked, the more the same pool is shared among more coins. The correlation between staking rate and issuance yield is structurally −1. Plotting both together is like plotting an identity.

The only free variable is the difference between the published yield and the formula value: fee income. It was about 1.34 percentage points in 2022, and today it is 0.10 percentage points. Correlation itself Answer: There is no correlation. For 43 months, from January 2023 to July 2026. (The data refresh on August 24 did not rerun this item; the window ends in July 2026, and subsequent price fluctuations do not affect this result.)

Regression results

▲ The relationship between end-of-month ETH price and staking APR, along with OLS fitting. The fit looks strong, but the residuals show serious autocorrelation.

This level regression is "significant" at p = 0.006------but it is worthless. Durbin-Watson is 0.40, indicating serious serial correlation in the residuals, which is a textbook characteristic of spurious regression between two trending series. Both variables have trends, so they are correlated; the standard error is underestimated, and the p-value is unusable. Retaining this chart serves as a warning, not as evidence.

▲ Scatter plot of monthly ETH returns versus changes in staking APR for the same month, with the OLS fit line nearly horizontal and a wide residual band.

After differencing to eliminate trends, the relationship disappears: p = 0.73, R² = 0.003. Durbin-Watson is 1.75, indicating that this setup is clean. The 95% confidence interval comfortably crosses zero in both directions------the data cannot even determine the sign of the effect, let alone its magnitude.

▲ Rolling correlation of changes in staking yield with ETH returns over 12 months, oscillating around zero and mostly falling within a range indistinguishable from zero. Moreover, this is not a stable relationship hidden in noisy averages------the rolling correlation repeatedly crosses the zero axis, spending most of the time in a range indistinguishable from zero. Between January 2023 and July 2026, ETH's staking yield dropped from 3.98% to 2.50%, while ETH/BTC fell by 57%. In the same window, the monthly correlation between changes in staking yield and ETH returns was −0.05. The yield has always been there.

It did not protect the price, nor did its compression cause a decline. If the existing reward curve's natural 37% yield reduction had no detectable price effect, then the burden of proof falls on anyone claiming "another cut will have an effect." Note: The staking series is reconstructed from on-chain traffic, approximately 5% higher than published data. The direction and shape are reliable, but the absolute level is not precise. Supply growth rate also does not explain If yield does not affect price, then what about the supply numbers that this proposal truly changes? The same test, same window, replacing yield with net supply growth rate.

▲ The relationship between monthly ETH returns and annualized net supply growth rate. The fitted line slopes downward, but the scatter points are very dispersed, and the relationship is not significant.

Slope −6.9 (for each percentage point increase in annual supply growth, monthly returns drop by 6.9 percentage points), p = 0.18, R² = 0.044, Durbin-Watson 1.82. The 95% interval for the slope is from −17.1 to +3.4.

Please read this result honestly, as it cuts both ways. This relationship is statistically insignificant, and the interval crosses zero, so it cannot serve as evidence that "reducing supply growth will raise prices." However, it is about fifteen times stronger than the yield relationship (R² 4.4% vs 0.3%), and the sign is consistent with theoretical predictions. If either of the two variables truly has a marginal effect, the data indicates supply, not yield------and this is precisely the exchange that EIP-8363 makes.

6. How deeply does the on-chain economy rely on ETH yield?

Liquid staking

Lido alone accounts for 48% of Ethereum's total $48.5B DeFi TVL. Any claim that "DeFi will be fine" must first withstand this number. What does the yield cut mean for them? Liquid staking: income is impaired. Lido handles about $602M in staking rewards annually, taking a 10% fee (about $60M/year). This cut of 58.6% in issuance equates to 633k fewer ETH rewards per year; based on Lido's 22.8% share, it will earn about $35M less in fees per year: nearly half of its income from this segment. This is not insignificant for Lido, but on the entire Ethereum level, it is inconsequential. Moreover, regardless of how yields change, wstETH remains far superior to WETH for any borrower wanting ETH exposure, maintaining its role as collateral.

LST as lending collateral------the real dependence

▲ The proportion of liquid staking tokens in TVL: SparkLend 66.9%, Aave V3 38.7%, Morpho Blue 10.4%, totaling 34.2%.

In Ethereum's three major lending markets, $31.1 billion of collateral includes $10.63 billion (34.2%) in staking yield derivatives. SparkLend is a typical single point of failure: two-thirds of it is wstETH. ETF channel, quantifying it The most frequently cited objection is that cutting yields will siphon off institutional buying, as staking ETH ETFs market yields to those who cannot directly access returns. This channel is real. But it is also very small today.

Products explicitly targeting yield account for only 5.4% of ETF assets, 0.53% of all staked ETH, and 0.19% of total ETH supply. BlackRock's non-staking ETH product is ten times its size. Regardless of what is drawing institutional funds into ETH, staking yield is not the main selling point------allocative funds overwhelmingly buy non-staked exposure.

There are two points that prevent this conclusion from becoming a nailed-down conclusion. First, the category of staking ETFs is still very young and growing: Bitwise and Grayscale are now working on staking ETFs for Solana, and Grayscale has also created one for Hyperliquid, so future risks are greater than the current AUM. Second, declining yields may slow the conversion speed of existing non-staking ETF assets to staking share categories, but this is an impact on growth rates, not capital outflows. Both points do not change the order of magnitude: ultimately, this is a $0.5B group fighting over a $1.55B/year benefit transfer.

7. Conclusion and Judgment: When "security anxiety" collides with "interest redistribution"

First, cut through the grand security narrative.

We must acknowledge that the core authors of EIP-8363 (such as Justin Drake and Jerome) have extremely serious intentions regarding network security. From a game theory perspective, once the overall staking rate crosses the 50% life-and-death red line, Ethereum will lose its ability for "social layer defense" against extreme attacks and face systemic parasitic risks from LST oligarchs too big to fail. Therefore, this proposal attempts to forcibly lock the staking rate in a safe zone through economic means of enforced rate cuts.

But on the flip side of security philosophy, the reality of on-chain data is much colder.

Since 2022, Ethereum's "burn mechanism" has become virtually non-existent: the burn amount has plummeted by 98%, now only offsetting a mere 2.4% of new issuance. Regardless of your stance on the security intentions of EIP-8363, an unavoidable fact is that the previous mechanism of "allowing ETH supply to dynamically adjust with market demand" has ceased to operate. Under the current L2 economics dominated by blobs, expecting L1 fees to surge to revive the burn mechanism is akin to dreaming. Ethereum's monetary policy has entered a state of "autopilot without a steering wheel," and adjusting issuance is the only trigger we can still pull.

Setting aside emotions, the real policy impact lies between the two extreme narratives.

Supporters shout "end ETH inflation," while opponents warn of "staking system collapse," both rhetorical stances deviate from mathematical facts. At the current staking scale of 42.2 million ETH, this proposal will only reduce issuance by about 58.6%, with staking APR dropping by about 56%. Want to bring issuance to zero completely? This requires staking to soar to 60.25 million (43% higher than now). More importantly, this mechanism has built-in brakes: as yields decline, some stakers will exit, and the system will ultimately stabilize at "a staking rate of 26% and an annual inflation rate of 0.48%." What it actually delivers is merely halving the dilution, not destroying or overturning anything.

Is the "half percentage point" saved really important? The numbers are more honest than words.

At current prices, reducing issuance by 633,000 ETH per year equates to preserving $1.55 billion, about 0.53% of the total market value. Don’t underestimate this proportion; it is roughly five times the entire L1 fee economy of Ethereum (about 0.10% per year). For an asset whose fee income has already dried up, plugging a structural bleed of 0.5% per year is by no means a "rounding error," but the largest economic leverage we can currently utilize.

So what is the cost? Will DeFi really collapse? The risk does exist.

Opponents often cite collateral, such as two-thirds of wstETH in SparkLend. But we need to clarify the difference between "exposure" and "dependence": as long as wstETH still has positive yield, it will always be superior to ordinary WETH as collateral, and this foundation remains solid. What EIP-8363 will truly shatter is the "leveraged staking loop." When the base staking yield drops below 1.16%, unable to cover the interest on borrowed ETH, the funds relying on leveraged arbitrage will disintegrate. In other words, it is the leveraged bubble that will shrink, not the collateral system itself. As for direct losses on the protocol side, Lido will lose about $35 million annually: nearly half of their fee income.

Concerns that "cutting yields will crash the market" have actually been settled by the market.

In the past 43 months of data, no significant correlation can be found between the fluctuations in staking yield and ETH's price performance (p = 0.73, R² = 0.003). The staking yield dropped from 3.98% to 2.50%, yet it did not prevent the ETH/BTC exchange rate from plummeting by 57% in July 2026. Yield is neither a moat for price, nor has its compression become a trigger for a crash. If the previous 37% yield decline had no price impact, then those claiming "another cut will cause Ethereum to plummet" need to provide stronger evidence. Why is this debate so intense? Because this is a zero-sum game where "losses are highly concentrated, and gains are extremely dispersed."

Peeling away the obscure technical language and grand security rhetoric, the essence of EIP-8363 is a brutal redistribution of wealth: currently, stakers take 100% of the newly issued ETH, but they only hold 35% of the tokens across the network. This means they are shifting the cost of inflation onto the remaining 65% of token holders. Cutting this $1.55 billion in issuance is equivalent to forcibly returning $1 billion of hidden wealth each year from stakers (intermediaries) to all non-staking ETH holders.

This is the real reason why all parties are so heated:

• The harmed parties are highly concentrated: Lido, LST issuers, re-staking protocols, leveraged players. This is a small, well-funded, and highly organized interest group. They are very aware of how much real money this proposal will take from their pockets (Lido directly loses half its profits, and the leveraged loop dies outright).

• The beneficiaries are extremely dispersed: Ordinary token holders who account for 65% of the supply. They can bear 0.5% less dilution each year, but this money is spread across nearly $300 billion in market value, making it virtually silent; no one will rally in the streets for it.

This explains why the current debate is always filled with "hollow rhetoric." When an interest group cannot openly state, "this will take away a billion dollars from our profits each year," they will raise the shield of "this will destroy DeFi"; and when researchers want to forcibly retract the faucet of currency issuance, the most politically correct weapon is "defend network security." Please interpret the volume of opposition and support as the concentration of interest distribution, rather than the mathematical correctness of the proposal itself. Our final judgment Strategy: Mildly bullish on ETH, clearly bearish on staking intermediaries/infrastructure. The proposal is likely to be rejected.

• On the asset level, we are bullish not because of the "scarcity myth," but based on common sense: when the only lever to adjust supply fails, removing a structural selling pressure of up to $1.55 billion per year (given to those who do not genuinely bear the cost of yield) is a highly cost-effective transaction. It may not be a dramatic reversal, but the compounding effect should not be underestimated.

• On the intermediary system level, the logic is flawless. The core valuation logic of Lido, LSTs, and LRTs is entirely built on the "staking yield" that is about to be cut in half. This is not emotional panic; it is a tangible profit statement shrinking by 58.6%.

• As for the fate of the proposal? The probability is extremely low. In decentralized governance, "concentrated harm vs. dispersed benefit" is the standard script that kills a good proposal. Economically correct, but politically difficult to birth, this is our baseline expectation.

Conditions that would trigger a change in our viewpoint (falsification indicators):

1. On-chain data proves "issuance is reinvested, not sold": If the flow of funds shows that newly minted ETH remains globally in auto-compounding LSTs and does not flow into exchanges to crash the market, then our selling pressure assumption would not hold.

2. Staking ETFs bring in massive buying: The current scale of $500 million is negligible. But if it expands tenfold, the marginal demand power will exceed the significance of the inflation reduction.

3. L1 fees miraculously recover: If the burn mechanism re-dominates the fundamentals, the urgency of artificially intervening in issuance will vanish.

4. The negative correlation between supply and price is thoroughly evidenced: Currently, the relationship between the two is very weak; if data in the coming year can confirm that "reducing supply must raise prices," this will become the most unassailable quantitative pillar for being bullish on ETH.