Liquid Collective pitches efficiency standards for Ethereum validators

An Ethereum liquid staking protocol for institutions says it’s now holding those who run its network to new standards and hopes others will follow suit — perhaps boosting yield and attracting more users along the way.

Liquid Collective, developed by Delaware-registered blockchain startup Alluvial Finance and launched earlier this year, introduced the series of service level agreements (SLAs) last week.

These new standards will require Liquid Collective node operators, which include Coinbase Cloud, Figment and Staked, to perform in the top 50th percentile of ETH staking providers operating 100 nodes or more, as ranked by Rated Network’s validator dashboard.

If they fall out of the top half, those who own the nodes must reimburse Liquid Collective for the revenue they missed due to their inefficiencies — effectively penalizing them until they maximize their yield.

Staking to earn yield, in Ethereum’s case, is the process of locking crypto in a proof of stake network to participate in consensus. This involves managing software to validate transactions and generally ensuring the blockchain functions well.

Staking rewards are designed to incentivize validators to act honestly. Operate in bad faith, say by pushing erroneous transactions, and face slashing. This removes validators from consensus and results in lost capital, usually 1 ETH ($2,200) per event, although penalties can be higher if misbehavior continues.

On Ethereum, a solo staker must deposit 32 ETH ($69,600) to join the network as a validator. Ethereum validators who stake directly to the chain can currently earn an annualized yield of 4%, which the network pays out in ETH as block rewards (some additional yield can be gleaned by adopting MEV boost relays).

Likewise, if the machine running the Ethereum node is slow and unreliable, or the internet connection is poor, that node may miss out on chances to participate in consensus and ultimately lose rewards, which can convert to lower overall yield.

2/ Performance data related to ETH validators and their operators is critical for staking’s development.

Establishing a common language for evaluating performance ensures clarity, consistency, and unlocks comparative assessment of a service provider’s long-term viability.

— Liquid Collective (@liquid_col) December 5, 2023

Liquid Collective has published an open source methodology showing how it grades validator performance

Users who don’t want to maintain their own nodes can opt for a staking service, like the ones Coinbase and Kraken offer, which manage Ethereum validators on their behalf and forward rewards minus a cut.

Protocols like Lido, Rocket Pool and Liquid Collective go one step further: users stake ETH in return for a different cryptocurrency, a liquid staking derivative token (LST), as a receipt.

Many liquid staking services allow smaller ETH stakes in return for scaled rewards, with Rocket Pool’s mini-pools accepting stakes of 8 ETH ($17,800), for instance.

Participants in the liquid staking protocols — often third-party partners — then similarly handle the Ethereum nodes instead of the user, and charge fees for the trouble. Coinbase and Kraken also accept partial stakes.

LSTs such as Lido’s, Rocket Pool’s and Liquid Collective’s can then be traded on crypto exchanges, used as collateral for loans and even staked again in liquidity pools across DeFi to earn additional yield. All this turns what would have been an illiquid ETH stake, subject to days-long withdrawal periods, into a more liquid digital asset.

There’s about $80 million staked in Liquid Collective, per DeFiLlama, much less than industry leader Lido’s $22 billion. All Liquid Collective stakers and operators must pass know-your-customer and anti-money laundering (KYC/AML) checks, and the platform has 69 unique depositors compared to Lido’s 200,000 and Rocket Pool’s 7,900.

There are now one million Ethereum validators and about 40% are managed by either Lido or Coinbase. The ETH staked with Lido represents one-third of all ether staked and nearly 8% of the entire circulating supply.

The growing ETH supply within Lido has become a point of contention over centralization of risk and security concerns.

Make validators better and yield will follow

Due to their fees, liquid staking protocols offer slightly lower yield than the Ethereum blockchain.

The two biggest liquid staking protocols, Lido and Rocket Pool, at the moment pay annual percentage rates of 3.93% and 3.57%, respectively. Liquid Collective doesn’t yet publish its yield data.

As for fees, Lido charges 10%, Rocket Pool takes 14% and Liquid Collective 15%.

With margins so thin, protocols compete to offer the highest yield possible (and the most attractive tokenomics). Part of that means ensuring operators keep their nodes, which typically run on dedicated servers, operating as efficiently as possible — hence Liquid Collective’s push to contractually obligate its own maintainers to meet certain service level standards.

After all, more rewards for validators converts to more rewards for end users. There have been about 400 slashing events in total since Ethereum’s beacon chain went live three years ago. At 1 ETH each, that works out to be nearly $900,000 in crypto lost due to slashing, at current prices.

More specific to liquid staking, the largest protocol Lido Finance faced a 20 ETH ($43,500) penalty in October after 20 of its nodes, operated by LaunchNodes, were slashed due to “non-optimal fallback procedures during datacenter connectivity issues.” In other words, its infrastructure wasn’t properly prepared to handle sudden downtime.

“It’s essential that institutional investors look for protocols who fulfill KYC/AML requirements. The remaining 20% could be maintained in spot ETH, offering additional immediate liquidity.” – @mleinweber2 @Josh_Deems https://t.co/kcEZUid246

— Liquid Collective (@liquid_col) October 17, 2023

Liquid Collective wants to be the liquid staking protocol of choice for entities with strict KYC/AML requirements

It is important to note that all three node operators on Liquid Collective are either investors in its creator Alluvial, in the case of Figment, or are owned by a company that has invested in Alluvial, as with Staked and Kraken. Coinbase Ventures similarly invested in Alluvial while its sister organization, Coinbase Cloud, runs nodes on Alluvial’s network.

Other backers include Brevan Howard, Variant and enterprise-grade staking provider Kiln, the latter of which led the build of Liquid Collective’s Ethereum platform. Alluvial itself was co-founded by the former CTO of Kiln alongside two former executives from Figment and Index Cooperative’s DAO.

The scope of the agreement also only applies to the validators those companies run for Liquid Collective. So, Ethereum nodes maintained by Coinbase Cloud, Figment and Staked in other contexts won’t have to stick to the new standards.

Liquid Collective measures node performance with a system developed by Rated, RAVER, which determines and compares validator health by analyzing whether they maximize consensus layer rewards over time.

(At time of writing, the RAVED system rates Lido’s node effectiveness at 97.9% — which is in the top 45% percentile and above Liquid Collective’s new benchmarks. Rocket Pool, meanwhile, is in the bottom 10th percentile with 94.7% effectiveness. Liquid Collective doesn’t yet have an effectiveness rating.)

“To perform in the top 50th percentile, staking providers that operate validators must ‘beat the competition’ by having better-performing infrastructure,” Mara Schmiedt, CEO of Alluvial, told Blockworks.

“This isn’t just about being better than average; it’s about guaranteeing node operators meet the high standards that enterprise clients are looking for when joining the space,” she said. The benchmark will be reviewed on a three-month rolling basis by Liquid Collective to make necessary adjustments based on performance analysis and risk metrics.

Other staking protocols can also adopt Liquid Collective’s standards if they so wish. The firm is betting that its open source methodology for ranking nodes might inspire the wider Ethereum ecosystem to set its own standards, which could help bring more of traditional finance over to crypto.

“In the competitive landscape of securing Ethereum, it’s imperative for stakers to be confident in the proficiency of their chosen service providers,” she said. “Establishing SLAs is more than just a formality; it presents a guarantee of excellence and reliability for the entire ecosystem.”

SLAs may not minimize risk — which institutions avoid

The pseudonymous adcv, a Lido DAO contributor, expressed it was great that research and new ideas are being shared in the liquid staking space. But they remain cautious about the roles SLAs like Liquid Collective’s will play moving forward.

“The clear main concern for any regulated institution is to shield their customers from risks, and the two major risks in liquid staking are counterparty exposure and liquidity risk,” adcv told Blockworks.

“I believe the fundamental reason institutions have overwhelmingly chosen [Lido’s] stETH — open source staking middleware — to be the leading enterprise-grade staking solution is because it deals with both of these issues.”

Adcv warned that centralization in liquid staking protocols could concentrate counterparty exposure, a big no-no in terms of risk and something that SLAs alone may not sufficiently mitigate.

“I am not convinced that SLAs will overcome these basic needs,” they said.

Alluvial’s Schmiedt, on the other hand, said institutions evaluating liquid staking protocols must also consider whether liquid staking providers meet their quality and compliance needs. That’s part of identifying all risk vectors, which includes centralization of stake in a single protocol.

“While existing solutions like stETH are great for crypto-native retail investors, they do not meet the needs of all market participants,” she said, adding that Liquid Collective’s products are tailored to meet the needs of certain customer segments.

David Canellis contributed reporting.

Source