My (e)thesis: settlement, data availability, execution

# My (e)thesis: settlement, data availability, execution — in that order. > "*What our generation has forgotten is that the system of private property is the most important guarantee of freedom.*" > **~Hayek, The Road to Serfdom** <img src=https://storage.googleapis.com/ethereum-hackmd/upload_32be190efd3362f3ee6258848379e47d.png width=92%> $\cdot$ *by [mike](https://twitter.com/mikeneuder) – monday, september 9, 2024.*                                                        $\uparrow$ [9/9](https://youtu.be/Fpu5a0Bl8eY?si=k8LrzjoBD6a_s0Ba) $\cdot$ *This is an opinion piece; reviews are not necessarily endorsements. Many thanks to [Nate](https://x.com/nan0gloss), [Jon](https://x.com/jon_charb), [Ryan](https://x.com/RyanSAdams), [David](https://x.com/TrustlessState), [James](https://x.com/_prestwich), [Vitalik](https://x.com/VitalikButerin), [Justin](https://x.com/drakefjustin), [Dankrad](https://x.com/dankrad), [Sreeram](https://x.com/sreeramkannan), [Kydo](https://x.com/0xkydo), [Barnabé](https://twitter.com/barnabemonnot), [Terence](https://x.com/terencechain), [Preston](https://x.com/preston_vanloon), & [Benji](https://x.com/0xFunk) for comments on dracts of this article. Additional thanks to [Sassal](https://x.com/sassal0x), [DCInvestor](https://x.com/iamDCinvestor), [gphummer](https://x.com/gphummer), [Francesco](https://x.com/fradamt), [Macrosam](https://x.com/sjerniganiv) for discussions that helped shape my thinking.* $\cdot$ **Contents** [(1)](#1-The-essence-of-Ethereum-property-rights) The essence of Ethereum: property rights [(2)](#2-Ethereum-and-rollups) Ethereum and rollups   [(2.1)](#21-Ethereum-DA-as-a-shared-security-substrate) Ethereum DA as a shared security substrate   [(2.2)](#22-Fees-and-value-accrual-to-ETH) Fees and value accrual to `ETH` [(3)](#3-On-Bitcoin) On Bitcoin --- # (1) The essence of Ethereum: property rights Ethereum is fundamentally about property rights. The Ethereum protocol creates a digital, self-custodied, permissionless asset whose value can be transmitted globally and cannot be seized or censored. Ethereum's relentless pursuit of decentralization is the means to that end. *Any compromise on decentralization opens an attack surface for regulatory or corporate capture, fundamentally limiting the market opportunity of a digital property rights system.* The three fundamental pillars of this argument are, 1. Blockchains are differentiated from traditional finance through their strong property rights – the inalienable right to store and transmit value. 2. Centralization provides a means by which powerful entities can influence the outcomes of blockchains. 3. The value stored in a property rights system is directly correlated to the credibility of said system's property rights. Taken together, it follows that a centralized system that can (and will) be coerced by nation-state actors *cannot provide the same property rights* of a decentralized system and thus is less valuable. It's a common misconception that Ethereum's decentralization is only necessary "in a [WWIII scenario](https://x.com/deww400/status/1826336394220015875)" or "[post-dollar world](https://www.youtube.com/watch?v=7CMMLdVrtFE&t=2842s)", but this is a strawman – decentralization matters today. The attack model for blockchains shouldn't only consider the adversary who wants to revert finality but also must account for a more subtle actor aiming to control the economic outcomes of the system without outright destroying it. This "death by a thousand cuts" shows up in protocol through coercion of validator behavior (see this recent [Staff Report](https://www.newyorkfed.org/medialibrary/media/research/staff_reports/sr1112.pdf?sc_lang=en) from the NY Fed) and enforcement of strict KYC/AML requirements for onchain activity (see the [details](https://www.steakhouse.financial/projects/blackrock-buidl) of Blackrock's BUIDL fund). Solana's stated goal of becoming "[best, most permissionless, accessible financial market in the world](https://www.youtube.com/watch?v=7CMMLdVrtFE&t=3731s)" and "[globally shared state that anyone can access permissionlessly](https://x.com/0xMert_/status/1828934969672204761)" is not possible without a clear story about maintaining credible neutrality of their block production. Without this, the outcome is a regulated but transparent financial transport layer, likely under the jurisdiction of the USG, which seems far less novel, impactful, and valuable than a digital property ecosystem centered on the original cypher-punk goals of censorship resistance and self-custody. Beyond the validator set, Ethereum has meaningfully decentralized many other parts of the ecosystem. Some examples are (i) the `ETH` token distribution arising from the crowd sale and Proof-of-Work mining, (ii) the [stake distribution](https://explorer.rated.network/?network=mainnet&view=pool&timeWindow=1d&page=1&pageSize=15&poolType=all) remaining diffuse, (iii) meaningful activity and volume across [L2s](https://l2beat.com/scaling/tvl), (iv) client diversity continually [improving](https://x.com/VitalikButerin/status/1827301405029507534). Ethereum's "[human capital](https://www.youtube.com/watch?v=7CMMLdVrtFE&t=4451s)" is also remarkable – an invaluable side effect of the network being built in the open by individuals and teams across the globe is that it empowers many people to contribute and invest in the future of the protocol. This true decentralization of value, power, and intellectual capacity seems extremely difficult to replicate. Additionally, since most of the technology is being researched and developed in an open-source and public-domain setting, Ethereum can inherit some benefits of execution-scaling-focused ecosystems. The tech can be commoditized, while Ethereum's decentralization cannot be. It is worth noting, however, that the market, rather than values, decides the outcomes of these ecosystems. The most decentralized chain may be less valuable if the marginal cost of said decentralization in terms of worse L1 execution, UX, and value accrual is too high. The bull case for Solana, Monad, BSC, Tron, etc., is that a much lesser degree of decentralization provides sufficiently strong property rights for most users and applications. I tend to believe that in the medium-term, [censorship](https://x.com/RyanSAdams/status/1828446420644294703), asset seizure, KYC/AML regulations, and validator coercion will call into question the robustness of centralized systems and will make their addressable market constrained to a single jurisdiction. In a multi-polar world where nations **do not trust** each other and attempt to regulate and monitor their citizenry through capital controls and financial surveillance, it seems likely that global economic activity **will not naturally funnel through a single system**; Ethereum has a unique claim on credible neutrality at a time when international cooperation and coordination is scarce. `ETH` is the asset that derives value from this credible neutrality by serving as the go-to choice for a genuinely permissionless store of value in the system. In contrast, centrally-issued `USD` stablecoins offer no property rights to their holders. As Sreeram put it, any holder of `USDX` runs the risk of getting their "[wings clipped](https://x.com/sreeramkannan/status/1830348370470416811)" arbitrarily by Circle or Tether; you can't have [programmable money](https://x.com/chainlinkgod/status/1832166952254751147) with counterparty risk. I hope `ETH` and `ETH`-collateralized stables & derivatives become the default option to retain sovereignty over digital property. # (2) Ethereum and rollups Ethereum's neutrality and censorship resistance make it the best place to settle, store, and express value. However, L1 settlement alone is an incomplete picture of what the rollup-centric roadmap aims to accomplish. Ethereum also serves as an affordable settlement and DA layer for rollups. I see rollups (and their corresponding rollup platform à la Optimism [Superchain](https://docs.optimism.io/stack/explainer) and Arbitrum [Orbit](https://arbitrum.io/orbit)) as independent fiefdoms. Each platform will compete to give users what they want: fast transactions, cheap fees, simple onboarding flows, etc., at the cost of decentralization. I call the fiefdoms because it seems likely that the teams responsible for creating and growing the ecosystems will continue to have outsized influence over their respective spheres – *this seems OK!* Rollups, by definition, make tradeoffs that the Ethereum L1 is unwilling to. If rollups needed to be as decentralized as Ethereum, what would be the point of having this symbiotic relationship in the first place? Rollups *depend on Ethereum for security and decentralization*, while Ethereum *depends on rollups to expand the economic activity of the ecosystem*. One big caveat here is that the rollups must achieve "[Stage 2](https://ethereum-magicians.org/t/proposed-milestones-for-rollups-taking-off-training-wheels/11571#stage-2-no-training-wheels-4)" status, where the rules of upgrading the bridge contract are maximally robust and provide a clear offboarding path for assets in the bridge. Notice, however, that Stage 2 makes no distinction over (i) the decentralization of the rollup sequencing, (ii) the destination of the fees and MEV generated from rollup activity, or (iii) the interoperability between rollup ecosystems. Stage 2 instead sets a norm for how rollups take advantage of Ethereum's security and decentralization without being overly prescriptive on the other dimensions of rollup design. I won't get into the arena of debating how or when rollups will decentralize their sequencers (though I generally align with [Max](https://x.com/MaxResnick1/status/1829127389437128843) on this – I don't see the incentives for them to do so). Still, I agree with Vitalik that [this shouldn't be the top priority](https://www.reddit.com/r/ethereum/comments/1f81ntr/comment/llmnpjz). I think the most important tasks at hand for rollups are (1) inheriting Ethereum security by becoming Stage 2, (2) inheriting Ethereum censorship resistance by having a transparent and efficient (not time delayed as is currently implemented) force-inclusion mechanism. In my mind, these are the key ingredients, and they both circle back to the theme of Ethereum providing the most robust digital property rights system for both L1 and L2 assets. ## (2.1) Ethereum DA as a shared security substrate One critical element of rollup design is where the rollups post transaction data, and indeed, we see [new](https://www.eclipse.xyz/), [projects](https://megaeth.systems/) already shifting the Overton window by being alt-DA from the get-go. I don't believe in trying to use social pressure or coercion to force projects to use Ethereum DA (and this path would be unsustainable anyway). Instead, we should examine what properties Ethereum DA provides that others cannot and consider the possible network effects of blob consumption. The primary advantage of using Ethereum DA is inheriting Ethereum's property rights and censorship resistance (do I sound like a [broken record](https://www.youtube.com/watch?v=rgesBcZe5Y0) yet?!). I like to think of this as embuing the "[Freedom of Movement](https://en.wikipedia.org/wiki/Freedom_of_movement)" for assets on a rollup. As a user, I'd be happy to engage in most of my daily financial activity on a rollup with less-than-Ethereum decentralization, so long as I know my assets cannot be seized and I have the same self custody assurances. With this in mind, let's examine the following scenario: > **Scenario:** Consider an L2 user who bridged `ETH` into the L2 via the canonical smart-contract bridge. They want to know under what circumstances they can withdraw their funds from the bridge to a different Ethereum account on the L1. *The ability to use the escape hatch of an L2 depends on where the L2 posts data.* - If the L2 is an Ethereum rollup and posts transaction data to Ethereum blobs, **the escape hatch is available for use unconditionally.** Because the data behind each state update on the bridge is committed to on an Ethereum blob, rollup users are guaranteed the ability to prove the validity of their withdrawal and get it included utilizing an L1 transaction (they retain sovereignty over their L2 assets). - If the L2 posts transaction data to an alt-DA layer, **the escape hatch is only available if the rollup is live.** By posting the L2 transaction data to a different chain, the state update for the bridge on Ethereum is coupled with the availability of the transaction data on the alt-DA chain. Namely, suppose an invalid state root is posted to the bridge without transaction data on the alt-DA chain (commonly called a "data withholding attack"). In that case, the L2 users cannot prove their exit is valid and thus cannot withdraw their `ETH` to the L1 (they lose sovereignty over their L2 assets). It's worth noting that the second outcome would require the L2 to permanently stop producing blocks in order to seize all the assets in the canonical bridge. This level of intervention would be pretty extreme. But the simple fact remains: only a Stage 2 Ethereum rollup that posts transaction data to an Ethereum blob can bestow the same level of property rights on assets bridged into the L2 (leaving aside the discussion of whether the ledger of record for a token is (i) a different chain (e.g., the `OP` token and Optimism mainnet) or (ii) a company (e.g., `USDC` and Circles internal database) – see [Jon's canon](https://dba.xyz/rollups-are-l1s-l2s-a-k-a-how-rollups-actually-actually-actually-work/) for more on this debate). The scenario above highlights the first (and best candidate in my mind) network effect of Ethereum DA: **a rollup posting their data to Ethereum DA benefits from other rollups doing the same because all assets across the chains share the same trust assumptions.** Sreeram calls this the ["trustless composibility network effect"](https://x.com/sreeramkannan/status/1770686691441062104) – a term I like. It's not clear how valuable this is from a user perspective (as pointed out by [VA](https://x.com/0xBalloonLover/status/1830444573304229968)). However, we are still very early in L2 adoption, and it doesn't seem helpful to speculate on. What seems much more important is ensuring that rollups don't immediately have an incentive to use external DA sources (see Dankrad's [excellent post](https://www.reddit.com/r/ethereum/comments/1f81ntr/comment/llmfpwd/) on this during the recent AMA). The goals of scaling Ethereum DA throughput with PeerDAS and Danksharding seem very well aligned with this vision of providing abundant blobs for rollups – making the decision to use them a no-brainer. In the future, it is conceivable that other network effects from Ethereum DA arise (see Wei's [tweet](https://x.com/_weidai/status/1830488020556083533) on this). For example, in a world with real-time proving of transaction validity and preconfs (sometimes called "[Universal Synchronous Composibility](https://www.youtube.com/watch?v=MnsjUZo7RRI)" by Justin), rollups consuming Ethereum DA may have access to much better cross-chain UX, more liquidity, and more users. I find this argument a bit too futuristic regarding technology and rollup participation to have a firm conviction about its inevitability. Yet the network effects of DA are critical only if we fundamentally view DA fees as a core component of the value of `ETH` the asset. Let's explore this a bit deeper. ## (2.2) Fees and value accrual to `ETH` Until now, we have yet to discuss fees and how they drive value to `ETH` the asset, despite that being a dominant narrative over the past few weeks. As evidenced by the organization of this article, I see it as tertiary in importance to (1) Ethereum's property rights and censorship resistance as a settlement layer and (2) Ethereum's ability to extend security and decentralization to rollups as a DA layer. With that being said, it is worthwhile to consider more "direct" forms of value accrual to `ETH` the asset. Personally, I align most closely with Dankrad's view on the topic of DA fees: > "*I do not believe that fees from blobs will be the best value capture mechanism for Ethereum. The data availability market is too fickle -- while Ethereum provides the best security, it is too easy to get something 'close enough' that it will never be a good way to extract value.*" > **~Dankrad on the [AMA](https://www.reddit.com/r/ethereum/comments/1f81ntr/comment/llbwkc1/).** Fundamentally, I don't see Ethereum DA as being that sticky. The network effects mentioned above do not feel strong enough to command a consistently high blob fee, but I don't see that as a problem. By providing cheap DA to rollups, Ethereum encourages them to build and grow the amount of economic activity in the Ethereum ecosystem. As such, proposals to rent-seek on blob pricing to drive short-term burn rates higher seem like entirely the wrong direction (again agreeing with Dankrad [here](https://www.reddit.com/r/ethereum/comments/1f81ntr/comment/llbwkc1/)). Francesco also has a [great response](https://www.reddit.com/r/ethereum/comments/1f81ntr/comment/llmoac5/) on the AMA sketching out the math for how many L2 transactions are possible under the proposed DA scaling. The other source of direct value accrual to `ETH` is the burn associated with fees from L1 execution. Max and co. have started an [information campaign](https://x.com/RyanSAdams/status/1831350573037178977) to bring back all DeFi execution to the L1; Justin holds that there is "[no future](https://www.youtube.com/watch?v=Rd_04vVLE-4&t=1733s)" for L1 execution; I land somewhere in between. Quoting Dankrad again (last one, he just had some bangers in the AMA), > "Ethereum L1 will be the crossroads between all those subdomains, and a lot of very valuable activity will continue to take place on it and that will generate valuable fees. (A decent amount of L1 scaling will be necessary for this to happen.)" > **~Dankrad on the [AMA](https://www.reddit.com/r/ethereum/comments/1f81ntr/comment/llmfeww/).** It seems like valuable activity will always occur on Ethereum, and creating a platform to facilitate large amounts of L2 economic activity will also drive base-chain usage. Thus, scaling the L1 execution layer to facilitate this activity is essential, though I see it as less urgent than maintaining and improving the properties of Ethereum as a settlement and DA layer. Again this underscores my core premise, which is that Ethereum should maximize economic activity within its platform (including rollups) and `ETH` should position as a true permissionless, digital store of value (see related tweets from [Doug](https://x.com/0xdoug/status/1830090816473190684) & [Sassal](https://x.com/sassal0x/status/1688372941577482240)) more so than an interest-bearing capital asset. The stated focus on the store-of-value properties of `ETH` naturally suggests the question, "Why not `BTC` instead?" We conclude with a short response to that question. # (3) On Bitcoin There is much to say about Bitcoin and `BTC`, especially as it [re-activates](https://docs.google.com/presentation/d/1Mxyp4rDPz0FJuY0Ai3KKmFJgz0Ilq4Idi78_HuK5wNY/edit#slide=id.g1bfa905fe34_0_633) its research and development ecosystem around ordinals, runes, rollups, the BitVM, etc., but this piece is not the venue for the more nuanced discussion (nor am I the person best equipped to write it). I will, however, highlight a few critical points as they relate to the vision of Ethereum described above. 1. **The 21 million promise.** The core promise of Bitcoin is the fixed supply of tokens. The revolutionary concept of creating digital scarcity was incredibly powerful and has placed it in the [Top 10](https://companiesmarketcap.com/assets-by-market-cap/) most valuable assets in the world (tenth with a one trillion market cap. as of Sept. 2024) – an astonishing accomplishment for the 15-year-old asset. But I see the 21 million promise as the fatal flaw in the system because I believe that Bitcoin's fork-choice rule is fundamentally "[unstable under diminishing block rewards](https://www.cs.princeton.edu/~arvindn/publications/mining_CCS.pdf)." The standard response to this claim is that fee revenue will be high enough to incentivize honest mining behavior, but I don't buy that argument either. The plot below shows the fees generated by the network over the past six years. ![Screen Shot 2024-09-06 at 12.56.07 PM](https://hackmd.io/_uploads/rk0EN2u3C.png) I am simply not convinced a mining company could remain profitable under this revenue stream. Consider the two-year window of mid-2021 to mid-2023, where the fees consistently remained below 1 `BTC` ($10^8$ `sats`) per block. A more optimistic potential equilibrium is that a majority of the `BTC` is accumulated by ETF issuers who decide to subsidize the mining industry to keep earning fees through their AUM business model. However, that doesn't seem like a very cypher-punk outcome. Additionally, the projection that fee revenue will incentivize mining seems directly at odds with the "buy & hodl" norm. Where are the fees coming from if everyone holds? 2. **Bitcoin disrupting itself to become a settlement and DA layer.** The most viable answer I've heard to the fee source question is that Bitcoin becomes a settlement and DA layer for L2s who pay fees. This is plausible and similar to the path that Ethereum is taking, but with two distinct differences. - (1) The core security model of the Ethereum network does not depend on the fees generated by settlement and DA because of Ethereum's issuance model. I even argued above that I don't think DA fees are a critical component of the value of `ETH`. For Bitcoin, there would be an *existential need* to *constantly generate* fees, which appears more circular: "The security of the L1 depends on the fees paid by L2s, who depend on the L1 for security." - (2) Bitcoin has neither a scaling roadmap nor a norm for upgrading the network. Again, this is a blessing and a curse. While stability and predictability are core features of the Bitcoin system, they also might impede its ability to make a full pivot into becoming a settlement and DA layer. This seems to be a classic innovator's dilemma as the system may be too big and successful to make extensive changes like adding `OP_CAT` and increasing the block size, which may be necessary to provide L2s the resources they need to scale usage meaningfully. I am happy to be proven wrong here, and I have much less visibility into the Bitcoin ecosystem, but this is the vibe I get. More on Bitcoin soon, but I will leave it at that for now. `BTC` has a strong case for being digital gold – a "pet rock" that is highly valuable but static. I believe `ETH` has a more dynamic future as a censorship-resistant, programmable store of value underpinning a much larger digital economy by providing permissionless settlement, DA, and execution. ## Final remarks Ethereum's unwavering focus on decentralization aims to create the most secure and censorship-resistant substrate for the on-chain economy. The rollup-centric roadmap seeks to expand the economic activity of the platform without compromising on the features of the settlement layer. As a DA layer, Ethereum provides a cheap and maximally trustless way for rollups to sacrifice decentralization to attract more users while ensuring the sovereignty of their user's assets. I agree with [Myles](https://www.youtube.com/watch?v=it3n5LFs46Y&t=2084s) when he says that, regardless of the exact mechanics of the value capture, `ETH` will be more valuable as more economic activity takes place in the ecosystem; optimizing for value capture today feels premature. Finally, while I believe maintaining settlement properties and scaling DA are the most critical roadmap features, scaling L1 execution can and should be done in parallel by building on the technology and innovation across the space. Fundamentally, I consider `ETH` to derive most of its value from being a global, permissionless store of value, and while it is engaging to discuss the value accrual story as it relates to scaling of the ecosystem, long-term user and developer growth should take precedent over short-term focus on token mechanics. The rollup-centric roadmap makes a lot of sense; settlement, DA, & execution – in that order.