Original Author: Ethereum

When liquid collateralized derivatives exceed the consensus threshold, they will no longer be safe

Liquid Staking Derivatives (LSDs) such as Lido and similar protocols are a manifestation of cartelization and pose a significant risk to Ethereum and the pooled capital associated with the protocol when key consensus thresholds are exceeded. Capital allocators should be aware of the risks to their capital and allocate to alternative agreements. LSD protocols should be self-limiting to avoid centralization and protocol risk, ultimately destroying their product.

first level title

Cartelization

In extreme cases, if the LSD protocol exceeds key consensus thresholds, such as 1/3, 1/2, and 2/3, due to coordinated MEV withdrawals, block time manipulation, and/or censorship, compared to non-pooled capital, Pledging derivatives can obtain excess profits, which is the cartelization of block space. In this case, due to large cartel rewards, staked capital is discouraged from being staked elsewhere, increasing the cartel's control over staking.

The LSD protocol can minimize governance, upgradeability, and other risks over time, but the question of "who" becomes part of the set of node operators (NO ) remains. This is also the main reason for cartelization.

Deciding "who" becomes a node operator involves two issues - who is added to the set and who is removed. In the long run, this could be designed in two ways, one through governance (token voting or other similar mechanisms) and two through automation around reputation and profitability.

Option 1: Decide on node operators through governance

If node operators are determined by governance, governance tokens (such as LDO) become a major risk for Ethereum. If tokens can decide who can be a node operator in this theoretical majority LSD, then token holders can enforce censorship, multi-block MEV, etc. cartels activity, otherwise the node operator will be kicked out.

In fact, this economic monopoly only strengthens the token’s grip on node operators. If tokens exercise a monopoly position through destructive mechanisms for excess profits, then in extreme cases, independent operations of node operators will be almost unprofitable. Therefore, node operators determined by the governance token may become a self-reinforcing cartel of the Ethereum protocol, causing abuse of the Ethereum protocol.

Governance-decided node operators have another obvious risk, that of regulatory scrutiny and control. If the collective stake in an LSD protocol exceeds 50%, the collective stake will gain the ability to review blocks (worse, when more than 2/3 can finalize these blocks). In a regulatory censorship attack, there is a unique entity — the governance token holder — to whom regulators can make censorship requests. Depending on token distribution, this could be a much simpler regulatory target than targeting the entire Ethereum network. And, in fact, the distribution of DAO tokens is often so poor that a few entities have the majority of votes.

Therefore, in any form of token governance control over a majority LSD, we rely on the benevolence of the DAO, or the control being well structured. But relying on the benevolence, anonymity, or geographic distribution of entities to prevent attacks is insecure and far from sufficient in the long run.

Option 2: Deciding Node Operators Through Economic Choice

If it was the second way, where node operators were decided based on economics and reputation, we would actually see similar results, except that it would be an automated cartel. First, getting into the set of node operators takes time and money (i.e. stake some ETH, similar to Rocketpool's design, slowly show profitability and get more ETH allocated). While entry into the set of node operators that takes time and money may make it difficult for new entrants, this is not what causes cartelization. Conversely, automatic culling is necessary if a node operator’s performance does not meet certain profitability criteria.

This mechanism is probably the only trustless (non-governance) way to ensure that node operators are beneficial to the mining pool. Defining profitability is not easy - either you define some absolute number (e.g. good baseline issuance rewards) or you need to define some relative number (e.g. no less than 10% of average/normal profitability). Given the unpredictability of MEV/TX rewards within a certain time window, and considering the importance of MEV rewards for long-term profits, we need a dynamic standard and need to be compared with other operators/validators over a period of time. That is, since the economic activity of the system varies greatly over time, the system cannot be designed to have only some absolute metrics, an X variable must be introduced in the transaction fees.

This profitability comparison metric works well when all operators are using "honest" techniques, but if any number of node operators tend to use disruptive techniques like multi-block MEV or adjusting block release times to gain more If there is too much MEV, then they will distort the profit target, so that if honest node operators do not participate in disruptive technology, they will eventually be automatically eliminated.

This means that in either approach (governance determines node operators or economic choice determines node operators), such a pool that exceeds the consensus threshold will be cartelized. It is either a direct cartel under governance, or a destructive, profitable cartel engineered through smart contracts.

Flaws in pledged ETH governance

As an aside, some believe that LSD ETH holders could have a say in the governance of its underlying LSD protocol, preventing an uneven distribution of chaebol tokens.

It is important to note here that ETH holders are not Ethereum users by general definition, and in the long run we expect that there will be far more Ethereum users than ETH holders (holding more ETH than actual exchanges required number of people). This is a key and important fact of Ethereum governance - there is no on-chain governance of ETH holders or stakers. Ethereum is the protocol that users choose to run.

In the long run, ETH holders are only a subset of users, so staked ETH holders are an even smaller subset. In the extreme case where all ETH becomes staked ETH under one LSD, governance vote weights or negative votes of staked ETH do not protect users.

So, even if the LSD protocol and LSD holders are aligned on subtle attacks and captures, this won't include users, who can/will react.

The insidious nature of governance

Even if there is a time delay in LSD governance such that pooled capital can exit the system before changes occur, the LSD protocol is still subject to a boil-frog governance attack. Small, slow changes are unlikely to cause staked capital to exit, but the system can still change dramatically over time.

Also, as mentioned above, LSD holders are not the same as Ethereum users. LSD holders may accept some kind of censored governance vote, but this is still an attack on the Ethereum protocol, and users and developers will mitigate the attack through the means at their disposal - social intervention.

Note: "In the case of malicious governance, pledged ETH can always be withdrawn", from the current point of view, it is not actually technically correct, and it may not be correct in the future. The validator's active key is the only key in the current Ethereum PoS design that allows withdrawal of stake. While there have been many proposals to add exit functionality to the BLS and smart contract withdrawal certificates, there has been no consensus on intent or design.

Capital risk and agreement risk

Much of the above discussion has focused on the risk that LSD pools (such as Lido) pose to the Ethereum protocol, rather than the risk to those holding capital in the pooled system. So, this looks like a tragedy of the commons - everyone makes a rational decision to join the LSD protocol for staking, which is a good decision for the users, but an increasingly bad decision for the protocol. But in fact, when the consensus threshold is exceeded, the risk of the Ethereum protocol and the capital risk allocated to the LSD protocol are one.

Cartelization, abusive MEV extraction, censorship, etc. are all threats to the Ethereum protocol, and users and developers will respond to these threats in the same way as traditional centralized attacks - through social intervention, leak or burn (Annotation: leak refers to inactivity leak lazy punishment). Therefore, funneling capital into LSD for cartelization not only puts the Ethereum protocol at risk, but in turn puts the pooled capital at risk.

These things seem like "tail risks" that are unlikely to happen, or never happen, but if we have learned anything in cryptocurrency - if it can be exploited or there are some unlikely "key edge cases" ", then, the time to be exploited or collapsed will be much earlier than you think. In this open and dynamic environment, rigid systems break down and fragile systems are exploited for fun and profit, over and over again.

The Ethereum protocol and users can recover from LSD centralization and governance attacks, but the process will be awkward. Lido and similar LSD products should be self-limiting for their own benefit, and capital allocators should acknowledge the pooling risks inherent in the LSD protocol design. Due to the inherent and extreme risks associated, capital allocators should not allocate more than 25% of the total Ether staked to the LSD protocol.