Original author: Ronan, Collab+Currency

Original compilation: Deep Chao TechFlow

This article focuses on an overlooked but critical factor in decentralizing the Ethereum staking layer. Including the impact and returns of Ethereum ETFs, challenges facing decentralization, the scale of institutional capital flows and Lido, etc.

I. Ethereum ETF will follow Bitcoin ETF approval

People are starting to turn their attention to spot Ethereum ETFs, and confidence in the approval of spot Bitcoin ETFs stems from the SEC’s apparent inconsistency in approving futures-based products, but refusing to approve spot-based products. Blackrock’s application for an Ethereum spot ETF on November 9 further fueled this concern.

Considering the existence of the CME Ether futures market and multiple futures-based Ethereum ETFs, the logic of approval appears to be fairly transferable. Even the regulatory approach to Ethereum in the United States is non-security based. The possibility of Gensler or future regulators changing previous treatments is unlikely for a number of reasons.

Indeed, recent SEC legislation targeting Coinbase’s listing of securities excludes ether.

II. Ethereum ETFs with Rewards are a natural extension of Uncollateralized Ethereum ETFs

Until a spot Ethereum ETF is approved, issuers will be scrambling to find an implementation that would allow them to earn Ethereum staking voting rewards. ETH with rewards is better than ETH without rewards and may attract new investors who have been on the sidelines so far.

Issuers will compete to be the first market participant to launch staking voting rewards. Initially, issuers seemed irrational in running their own validators, given knowledge barriers, business model challenges for node operations, and increased regulatory risks.

To be first to market, issuers must come up with a solution that complies with existing regulatory frameworks and can be approved as quickly as possible. Therefore, the least laborious route is for the ETF issuer to enter into a contractual agreement, including a lending agreement, with a third-party centralized staking voting provider, which charges a small fee.

This is already the solution adopted by 21 Share’s staking Ethereum ETF AETH. 21 shares will host its ETH at Coinbase Custody and may lend underlying ETH to Coinbase Cloud, Blockdameon, and Figment.

AETH has attracted $240.77 million in NAV, equivalent to 121,400 ETH, all of which has been pledged to centralized providers. These inflows are completely independent of the yield, as some fixed percentage is programmatically collateralized regardless of the yield.

It is reasonable to assume that US ETF issuers will likely enter into lending agreements similar to 21 Shares AETH. This programmatic funding flow has the potential to drive centralized providers to gain greater market share relative to decentralized providers, as many custodians offer vertically integrated staking voting products (e.g. Coinbase Prime → Coinbase Cloud) Or have an existing service level agreement (SLA) with a centralized staking voting provider such as Figment. As a specific example, considering that ARK is partnered with 21 shares, it is reasonable to assume that they will use the same provider as the AETH product.

III. Institutional staking voting will shift to liquid alternatives, bringing new challenges to Ethereum decentralization

Savvy institutions may seek to stake their votes with providers outside of ETF wrappers that have more favorable cost structures and greater utility. Decentralized protocols such as Lido are already available as assets to existing institutional clients in various custody, QC and regulated environments. As decentralized protocols, they offer a consistent experience and institutional-level security, but are open to all market participants wishing to stake any amount of ETH.

On the other hand, some new programs are positioning themselves specifically to address institutional needs. Especially like Liquid Collective One such company is building a “liquidity staking solution designed specifically for institutional compliance needs.” Institutions can mint lsETH, which is staked by one of 3 centralized providers (Coinbase, Figment, and Staked) that also manage the project. The idea here is twofold:

• Liquidity staking is a better product than regular staking voting; you retain the monetary properties of your ETH and most of the rewards.

• Institutions must stake with a KYC/AML compliant provider or risk civil and criminal liability.

The first point is fairly obvious.

Liquidity Staked Tokens (LSTs) can be used as collateral throughout DeFi, are the underlying asset in liquidity pools, and avoid withdrawal queue times.

Additionally, institutional products such as ETFs benefit from liquidity to manage fund redemptions in less than a day. For illiquid funds, this is typically managed by keeping a portion of uncollateralized ETH in escrow. This carries the risk that a massive influx of withdrawals could cause a virtual bank run while the remaining ETH is uncollateralized, as well as drag down the reward rate during normal operations.

Having liquid staking tokens will make it possible to manage redemptions more smoothly and also increase the proportion of funds that can be staked at any given time. In order for this to be a real possibility, obviously the liquid staking tokens must be liquid. Simply offering a token is not enough if there is not enough liquidity available for that token. Currently, the only liquid staking token with any substantial on-chain and off-chain liquidity for institutional use is Lido’s stETH.

The second point is less clear. Regulated institutions are generally required to fulfill a range of obligations to reduce the risk or likelihood of money laundering or the promotion of crime. To this end, KYC/AML obligations exist to maintain an auditable trail of funds flowing between institutions and their clients. Additionally, there may be higher requirements for a “qualified custodian.” That said, qualified custodians and institutions in general should be able to meet their KYC/AML obligations without compromising asset selection, whether it is the LST token or the staking provider.

Even if the staking voting provider explicitly enters into a contractual relationship with the owner or custodian of the funds, I do not expect regulators to create entirely new KYC/AML compliance obligations for Ethereum staking voting. This is because I believe regulators will learn over time that Ethereum staking voting is a special computational activity with characteristics that do not match “money flows” in the traditional or financial sense. Holders and custodians of LST should be able to perform KYC/AML on any assets within their sights and meet their compliance obligations to reduce the risk of money laundering or criminal financing.

The centralization of staking within a centralized entity presents a variety of emergent challenges in developing the Ethereum blockchain. Overall, various possibilities for disrupting the blockchain become increasingly likely at different levels of stake share:

• block reorganization attack

• Final confirmation delayed

• Fork choice

• coercion

The first two attack types interfere with the normal functioning of the blockchain, and Ethereum has some incentives built into it to discourage attackers from trying to do this, such as gradually diluting the attackers rewards and staking balance. However, when the market share among a single node operator reaches 33% or higher, that actor may start delaying final confirmations even if disrupting network operations becomes expensive. At higher levels of market share, such as 50%, attackers can effectively fork the blockchain and choose a fork they approved of. At 67% or higher, the blockchain effectively becomes a delegated database controlled entirely by a single party.

These attacks are not just theoretical, they are at the heart of Ethereum’s value as a settlement layer.

Ethereums proof-of-stake mechanism makes it possible for centralized players to individually accumulate a large and potentially controlling share of the total Ether staked through market forces. For example, centralized entities are already well-positioned to quickly convert multiple lines of business (including custody relationships) into staking relationships and solidify their position in the staking space. It is already the largest validator on the network with a market share of 16% and operates multiple acquisition channels such as cbETH and Coinbase Earn for retail customers, and Coinbase Cloud and Coinbase Prime for institutional customers.

Behind this influx of new capital lies a potentially dangerous challenge to Ethereum’s neutrality, or as the Ethereum Foundation describes it, a “Layer 0” attack on the “social layer.” Even with all the good intentions, adding a staking voting provider that just assumes only KYC/AML compliance is allowed, even if that compliance is illusory and not backed by legal or technical facts, will only hasten the passage of regulations firmly and seize up Centralized players of market share use staking voting.

Centralized entities increasing their market share poses a risk to Ethereum’s staking layer. They fundamentally have a different set of obligations than decentralized protocols. Decentralized protocols exist as a smart contract incentive layer for the purpose of coordinating activities among many participants.

For example, Rocket Pools rETH has nearly 20,000 holders, 9,000 RPL holders and more than 2.2,000 pledged deposit addresses, representing unique node operators. Or a smart contract layer like Lido, which coordinates 39 node operators distributed around the world, nearly 300,000 stETH holders and approximately 41,000 LDO holders.

However, a company has a fiduciary duty first and foremost to its shareholders and is accountable to local legal authorities and regulators. While Ethereum’s decentralization may be somewhat relevant to a company’s business (such as an exchange), it does not transcend these two obligations.

There is also a risk of other aspects of over-regulation, namely a soft “Layer 0” attack, even with good intentions. If Ethereum is to become the world’s settlement layer, it must have nuclear-level censorship resistance and trusted neutrality. If large enough, these centralized entities will hinder Ethereums core goals.

IV. The scale of institutional flows can be considerable

Some may overlook the arrival or impact of spot Ethereum ETFs due to a lack of institutional interest. However, there are some useful precedents in the commodities space that can tell us about the level of interest exchange-traded products (ETPs) generate in new asset classes. As financial instruments, ETFs and ETPs are highly effective in providing standardized and democratized access to institutional and retail investors. When these instruments enter distribution channels for institutional allocations, pension funds, or social security contributions, there are significant inflows into the underlying asset class.

The same benefits of the spot BTC ETF apply to the spot Ethereum ETF. For example, approximately 80% of U.S. wealth is controlled by financial advisors and institutions, who have access to participation and are generally recognized by regulators and governments. We can expect this increased legitimacy and recognition to drive demand for Ethereum beyond its ETF wrapper. We are already seeing signs of institutional interest. Intermediaries such as Bitwise heard potential allocations increased from 1% to 5%. Brian Armstrong stated that in their Q3 financial report, Coinbase had doubled the number of institutional users to more than 100.

Guessing the exact amount of net inflows to institutions is difficult. However, golds GLD ETF attracted $3.1 billion in net inflows in its first year alone. Buying any amount of gold before ETFs was much more difficult than buying BTC/ETH. You need to transport and store it yourself, verify and authenticate its purity, and pay dealers high transaction costs. Gold ETFs refine the underlying characteristics of the asset and democratize its value proposition through mega-fund allocations to millions of individual investors.

Bitcoin and Ethereum are the digital assets of choice, allowing billions of dollars to be transferred in minutes. The barriers to entry for physical gold do not exist here. While ETFs don’t necessarily improve the fundamental value propositions of Bitcoin or Ethereum—they could threaten them—ETFs offer a similar degree of democratization and access to both asset classes.

In fact, the vast majority of people around the world will likely never own any actual cryptocurrency, but will likely have some financial exposure to some extent through pension allocations, private savings or investments. Regulated financial channels in the developed world are already highly penetrated and capillary. ETFs can help demystify the asset class, making it easy for individual investors who may have been on the sidelines to get involved within an already familiar framework, such as a bank or brokerage.

V. What does this all mean for Ethereum’s staking layer?

Large-scale, yield-insensitive inflows from institutions could push the overall amount of staked ETH higher than either economic principles suggest or probabilistic targets based on crypto-endogenous variables. The vast majority of these inflows will likely accrue disproportionately to centralized providers within and outside of ETF wrappers. Greater centralization of staking within centralized entities will reduce Ethereum’s censorship resistance and credible neutrality without credible successful counterforces.

VI. Lido as an effective balance

Much of the focus of the public conversation today revolves around Lido – whether they control too much of the stakes, and the attack vectors this could introduce in a worst-case scenario. Here are some quick overviews of i) how much control, if any, Lido governance actually has over node operators, and ii) how the DAO handles governance risks, and iii) how the DAO considers expanding the Node Operator (NO) portfolio and decentralizing collection of validators.

• Governance risk: Hasu’s GOOSE submission + dual governance proposal

• Expanded NO combination: pledge router + DVT module

• Decentralized validator collection:Jon Charbs PoG + Grandjean paper.

Also see this excellent article on the real risks of Lido dominance by Mike Neuder of the Ethereum Foundationarticle。

Criticisms of Lido are largely based on a static view of the staking market. It does not take into account how the staking market develops and its practical realities. In order to fully assess the staking market, future growth and market forces must be considered:

• Future growth: Institutional adoption may drive concentration of centralized entities

• Market Power: (Liquidity) Staking has a winner-take-all nature.

Much of this article is devoted to outlining future growth, as this is an infrequently discussed but important aspect. The dynamics of winner-take-most have been discussed at length in the digital public sphere, but often lack the context for future growth. Rational market incentives, including incentives to maintain a decentralized Ethereum network, may not prevent institutions from taking the most expedient path and bringing new capital into Ethereum.

The only effective check and balance is to increase the market share of decentralized liquidity staking protocols at the expense of centralized market share. While multiple decentralized protocols will likely gain enough share to form an effective backstop, Lido is currently the only viable option to keep the Ethereum staking layer robust and decentralized:

• It has been objectively successful in attracting new ether from holders, as the Lido smart contract has bootstrapped over 30% of all staked ETH, and stETH has almost 300,000 holders

• It is objectively successful in limiting the growth of individual node operators, as they all use Lido as a successful channel to attract new ether staking, but cannot grow their individual market share within Lido beyond other node operators

• Its governance is currently objectively minimal, and with the advancement of initiatives such as dual governance, this degree will be further reduced.

Although the Ethereum base layer is designed to be “governance-less” or with very limited fork-choice governance, having an intermediary layer with one or more decentralized protocols can fill necessary gaps that Ethereum cannot fill. Jon Charbonneau describes it this way:

“In particular, LST governance can manage the additional subjective incentives required by decentralized operators (e.g. different modules may receive different rates). A free market economy will not lead to independent stakers or uniform staking in the long term Distribution. The Ethereum core protocol is largely built on the idea that it should be objective and non-opinional where possible. However, in order to achieve a decentralized mix of operators, subjective management and incentives will be required.

Although governance minimization is often desirable, it is always possible that LSTs require some minimal form of governance. Some procedures are needed to match the need for staking with the need to run validators. LST governance will always need to manage objective functions of the operator portfolio (e.g. targets for staking distribution, weighting of different modules, geographical targets, etc.). This kind of fine-tuning may not happen often, but this high-level goal setting is critical to monitoring and maintaining decentralization of the operator portfolio.