With Solana proposals SIMD 550 and SIMD 553 approaching a vote, Anza chief economist Max Resnick published a long-form essay on how the issues behind those proposals tie into Layer 1 valuation.
Resnick said the article was not written as a direct assessment of either proposal. He added that he had already shared his views in the relevant GitHub proposal threads. The focus here, he wrote, is broader: what the proposals are trying to solve, and how those questions connect to the way investors should think about L1 value capture.
Why L1 valuation needs a real pricing framework
Resnick opened with a comparison to the early history of asset pricing. Before a formal theory of asset pricing took shape, investors had no shortage of ways to value companies. Some focused on hard assets and liquidation value. Others emphasized profits, dividends, growth, management quality, or market psychology. What was missing was not data, but a disciplined framework that could explain which variables actually determine value and how they should be weighed against one another.
By the late 1920s, just before the Great Depression, that ambiguity had become dangerous. Investors could point to earnings growth, market expansion, new technologies, and better corporate governance, but those facts were often used to justify market prices rather than to derive intrinsic value.
He cited Graham and Dodd’s 1934 description of that era as one in which analysis gave way to “potentiality and prophecy.” Even when people brought numbers to the table, those figures often turned into what they called pseudo-analysis used to support prevailing fantasies.
Resnick argued that anyone who spends time on Crypto Twitter should recognize the pattern. In his view, current debate around L1 tokens is crowded with the same mix of potential, prophecy, and pseudo-analysis that appeared in equity markets in the late 1920s.
Record numbers of people are building on blockchains. Trading activity has reached all-time highs. Tokens are described as money, collateral, digital oil, or a call option on the future financial system.
Some of those claims may be true, he wrote, and some may point to upside in the underlying networks. Still, if they cannot be connected to residual value for token holders, they do not amount to a coherent valuation framework.
Resnick then turned to John Burr Williams, who helped push asset pricing onto firmer ground. In The Theory of Investment Value, Williams wrote in 1938 that value is “the present worth of future dividends in the case of stocks, or of future coupons and principal in the case of bonds.” Gordon later expressed the same idea more simply in 1959: “Stocks, like other assets, are purchased because they are expected to provide future returns.”
That logic matters for crypto, Resnick said. Equity is not valuable because a company looks impressive, stays busy, matters to the economy, or cannot easily be replaced in technical terms. It is valuable because it gives shareholders a claim on future income.
For L1 tokens, he argued, value can accrue in two main ways. One is through fee burns, which are economically similar to buybacks. The other is through distributing fees to stakers, which are economically similar to dividends.
Staking rewards funded by token issuance belong in a different bucket. Resnick said those rewards are neither revenue created by the network nor a cost borne by the network. The protocol mints new tokens and gives them to stakers, diluting holders who do not stake.
That mechanism may be necessary for security, and it may affect who gradually comes to own the network over time. But at the level of all token holders combined, he wrote, it does not create value and it does not destroy value.
A blockchain can process millions of transactions while creating very little value for token holders if the economic surplus ends up with users, applications, validators, or other intermediaries. A lower-activity chain, by contrast, could be worth more if it converts a larger share of economic activity into value for token holders.
Fee quality matters as much as fee volume
Not every dollar of fees deserves the same valuation, Resnick wrote. In ARR, the “R” stands for recurring, meaning income that is durable and repeatable. He cited Dichev, Graham, Harvey, and Rajgopal’s 2013 discussion of earnings quality, which said high-quality earnings should be “sustainable and repeatable.” He argued that the same standard should apply to L1 fees.
A dollar of fees generated by long-term financial activity is not the same as a dollar generated by an airdrop, meme coin mania, cascading liquidations, or temporary network congestion.
Some fees come from durable demand for scarce block space. Others are just the exhaust of speculative cycles. Once incentives fade, volatility drops, or users run out of money, that activity can disappear with them.
In his framing, fee quality depends on durability and defensibility. Are users paying because the chain provides long-term economic utility, or because a short-lived event happens to be taking place there? Can the protocol keep charging those fees without driving users, applications, or order flow elsewhere? Can the token continue to capture that revenue, or will the value be competed away to validators, applications, searchers, block builders, users, or rival chains?
Resnick said crypto investors have often made two opposite mistakes on revenue quality at the same time.
They overestimate it because so much crypto activity is speculative, reflexive, and cyclical. They underestimate it because they fail to appreciate how powerful L1 network effects can be.
Liquidity, applications, wallets, infrastructure, users, developers, assets, and order flow reinforce one another, he wrote. Those network effects can make some fees harder to dislodge than they first appear. They also suggest that major blockchains such as Solana and Ethereum may have stronger pricing power than the market commonly assumes, which means they may benefit from charging higher fees.
Separating revenue, inflation, and supply
In the second section of the essay, Resnick tried to define a minimal L1 fundamentals model that could support valuation multiples.
He said it may be too early to call it a “standard model,” because there is no widely accepted standard model for valuing L1s today. Even so, he argued that the categories he laid out are what such a model should look like, and he intentionally designed them to resemble the frameworks equity analysts use for companies.
He said spelling this out is necessary because the market has not even agreed on the most basic accounting objects. Resnick wrote that he has discussed this framework with some of the smartest people he knows, yet those conversations often break down over basic classification questions.
Among them: whether validator rewards paid via inflation should count as a cost; whether foundation spending should be treated as an operating expense; whether unused foundation token allocations should be counted in supply; and whether MEV paid to validators should be treated as protocol revenue, validator revenue, or neither.
Part of the confusion comes from the fact that there can be more than one correct valuation model. Accounting classification is not unique. As long as offsetting entries are adjusted consistently, an item can move from one side of the ledger to the other without breaking the model.
But Resnick warned that this flexibility cuts both ways. Many models are internally consistent. Many are not. The fact that more than one correct approach exists does not reduce the number of wrong ones.
Inflation rewards are the clearest example. Staking rewards funded by inflation are, in substance, token holders paying stakers through dilution. From the perspective of all token holders together, those effects cancel out. When the protocol mints new tokens, it has not earned revenue. When it distributes those tokens to stakers, it has not incurred a real external cost.
He said a model can correctly treat inflation rewards as a cost, but only if newly issued tokens are also counted as a source of value to offset that cost. Without that symmetric treatment, the model can produce absurd conclusions, such as claiming Solana is unprofitable because it pays large staking rewards.
His proposed baseline separates three things: revenue, costs, and total supply.
Resnick said those definitions are the closest match to the models equity analysts already know, which makes them easier to interpret and reason through. Other classifications may also be valid, he added, but anyone departing from this structure should have a clear reason for doing so. Bespoke models come with two costs: they are harder for others to understand, and they make it easier to miss dependencies across line items.
He gave an example. If foundation spending is classified as a cost, then unused foundation token allocations cannot also be included in total supply, or the model double-counts them. If inflation rewards are classified as a cost, newly issued tokens must receive symmetrical treatment.
Higher fees do not automatically mean higher revenue
The final section turned to supply, demand, and price. Resnick noted that several blockchains have recently raised fees explicitly to increase revenue. But revenue is price multiplied by quantity.
Charging more raises revenue per transaction for the activity that remains, but it also causes some transactions to disappear. The net effect is uncertain. It depends on the price elasticity of transaction demand.
To understand the reasoning behind those fee changes, Resnick said he spoke with some of the decision-makers involved at those chains. Their view was that fees had been too low, so demand in that price range was relatively inelastic.
That may be true in some cases, he wrote, but it is not a general rule.
Resnick said he had previously used randomness in EIP-1559 pricing to study the issue. His analysis suggested that gas demand has a price elasticity of roughly 0.6 to 0.8. In practical terms, that means a 10% increase in price is associated with a 6% to 8% decline in quantity demanded.
He added that those figures only capture short-term price movements. They do not include broader effects such as applications moving activity off-chain or optimizing their software in response.
For that reason, he called uniform pricing a fairly blunt revenue tool. Different types of on-chain transactions have different willingness to pay.
A small wallet transfer, a large stablecoin transfer, and a liquidation may consume the same amount of block space, but they do not create the same total surplus, and they do not justify the same fee level.
He also included a note on a chart in the piece: blue dots represent fee payers that initiated more than 250 transactions during the period, making them more likely to be bots and therefore more price-elastic. Bots often run on thin margins, so when prices rise, they tend to cut resource usage sharply.
Resnick said protocols should charge more to transactions with a higher willingness to pay. Compute-based fees are a step in that direction, but not the end point.
For financial activity, he argued, nominal transaction size often reflects willingness to pay better than compute usage does. That is why exchanges usually charge in basis points.
He suggested that token programs could support a pricing model based on transaction value by modifying the token program to collect a small proportional fee on token transfers. Under that approach, a high-value transfer would still pay more than a low-value transfer even if both consume roughly the same amount of compute.