Why is Ethereum decentralized staking important?

fuse

In the early days of Ethereum’s PoS beacon chain, early adopters thrived. Centralization is one of those devil fruits: if you give up trustlessness, decentralization, or permissionlessness, you’ll be richly rewarded. Web2 giants like Coinbase, Kraken, Binance, and others were more than happy to lock up the funds of willing users. What were the options? Aside from running your own node (a daunting task for many), these CEXs gave users another way to participate in staking. The ring seemed to have been given. The ills of PoW mining pools seemed destined to repeat, and were repeating. With greed and speed came control.

Against this backdrop, another force is growing. A DAO (here, Lido DAO) that seeks to bring benefits to the majority, without the blatant trust assumptions of CEXs. The decentralized staking protocol Lido was launched, and its liquidity staking token stETH appeared, which is the first and currently dominant liquidity staking token. For the first time, anyone on the Ethereum network can stake less than 32 ETH while retaining control of their funds. In Lido's case, users deposit ETH and receive stETH to enjoy the lending carnival driven by the DeFi summer. As capital pours in, an ominous dilemma is slowly brewing in this place that was once a place of awakening pursuit.

The Devil Fruit in the Japanese manga One Piece

Lido Era

CEXs take your assets, and then you have to trust them. Lido gives you a self-custodial receipt token (i.e. stETH), but where does your staked ETH go? The DAO that seeks to provide staked access (i.e. Lido DAO) becomes a black hole and the only access point. The Lido network is willing to happily accept an unlimited amount of ETH and distribute it to a few selected node operators. Trust means speed. In order to onboard more users, these operators who use users' staked ETH to run nodes must act quickly and be trusted as honest operators. A devil fruit is eaten, and power emerges - just like a black hole, this power can be trapped.

Lido’s validator set is full. The centralized giants of Web2 are now fighting Lido, which is the largest staking pool at the time of writing. We are willing to ignore the dangers of centralization issues in pursuit of yield. Capital efficiency is the foundation of all systems in DeFi. Where optimization can be found, it will be found. The DEX war is a clear example. Curve and Uniswap v3 compete in the super efficient token exchange market, and end users are cheering; the concept of liquidity pool tokens allows capital to be active even in your wallet, earning transaction fees for holders. Over time, efficiency is getting better and better, which is good for Lido.

Estimated validator counts in Coinbase, Lido, and Binance. Source: https://pools.invis.cloud/

It’s no surprise that stETH has become a DeFi Lego building block. If you are able to replace ETH in your investment strategy with stETH (or its wrapped version wstETH), you can passively increase your yield by a few percentage points. This increase in yield is not the limit. The essence of yield farming has always been liquidity incentives. For example, the Curve protocol provides CRV Token rewards to users who deposit liquidity into the stETH-ETH pool in the protocol and stake the LP Token crvSTETH they receive.

Introducing the yield aggregator Yearn. The vault system in the Yearn protocol packages these strange yield queues into a simple liquidity token. The protocol automatically recombines all additional liquidity rewards into the base layer. Yearn plays the role of a strategy building center, and the best yield strategies will attract the most capital inflows. Since its inception, Yearn's vault for Curve stETH-ETH liquidity pairing has been highly sought after. Its ycrvstETH Token is one of the most capital-efficient tokens in DeFi (Note: After the user deposits ETH or stETH into the Curve platform and obtains the LP Token crvSTETH, he will obtain the corresponding vault token ycrvstETH after depositing it into the Yearn vault). There is a complete yield ecosystem in your wallet, which is much more complicated than I described.

The ycrvstETH Token is the fourth most popular collateral asset on Abracadabra Finance, one of the most popular lending platforms. This is important because it shows that the Yearn vault token (ycrvstETH) is not the end point, but the base layer for higher yields and more complex strategies. The token represents the maximum yield that a non-custodial passive token can capture. For most of DeFi's history, ycrvstETH was the best asset in this regard.

The constant need to maximize yield has led to the expansion of Lido, the gatekeeper of liquid ETH staking. As Lido approaches 20% of all validators on the Ethereum network, further capital inflows will begin to threaten the security of the Ethereum network. There is a problem here. Currently, stETH will only continue to flow into the market because withdrawals from Ethereum validators will not be unlocked for several months, that is, until a hard fork after the merger. The flood of users chasing yield cannot be stopped - as long as there is yield, capital chasing yield will flow in, unless the consequences are immediately reflected in the price. If the Lido protocol operates 50% of the Ethereum network validators (they may rise in the short term), the returns from stETH will not decline, but in this case, the value and security of the Ethereum network itself may be irreparably damaged. Therefore, the only way forward is to move funds to greener pastures and higher yields.

The rise of Rocket Pool

Lido was once the only one in the space, but not anymore. Other Ethereum staking service providers have emerged with their own liquidity staking tokens. However, none is more important to the long-term health of the Ethereum ecosystem than Rocket Pool. While Lido’s permissioned nodes are willing to take on all validation duties, it may pose long-term risks to the network, and Rocket Pool refuses to cut corners. With Rocket Pool, there are almost no trust assumptions, only code and crypto-economic incentives. There is no speed on this path. Rocket Pool cannot onboard an unlimited amount of ETH in a day like Lido can. Instead, Rocket Pool has a framework by which a decentralized set of nodes can securely support a liquidity staking token network. The Rocket Pool network has over 800 node operators, some of whom are known major Ethereum players and others are anonymous.

At the time of writing, 106,432 ETH is staked in the Rocket Pool protocol, with 853 node operators. Image source: https://rocketpool.net/

Creating this framework was not easy. Lido Lido had been accepting user deposits for about a year before Rocket Pool launched. However, despite stETH’s huge head start and almost complete dominance of liquidity staking, in the three months since Rocket Pool launched, massive capital inflows have pushed the protocol to 1% of all validators on the Ethereum network, with over 100,000 ETH staked in the Rocket Pool protocol. The reasons for the protocol’s adoption are complex, but at its current rate of adoption, Rocket Pool is likely to attract increasing inflows from its competitors. If markets are any good, though, it’s that the prospect of future growth is reflected in the price at this moment. We can accelerate Rocket Pool’s development to directly benefit the health of the ecosystem.

stETH's 01 moment was the liquidity pairing with ETH on the Curve platform, thus compounding rewards on top of this LP Token (i.e. crvSTETH). Most stETH use cases use the stETH-ETH pool as the base layer. The pool is currently one of the largest single holders of ETH and one of the deepest liquidity pools in existence. What if we could create another 01 moment? The first such moment was the trading pair from ETH to stETH; the second such moment has already arrived, with the potential to double returns by doubling ETH's liquidity staking exposure in this largest building block of DeFi (i.e. Rocket Pool).

To understand why this is a turning point for DeFi, let’s look at where the yield generated by ycrvstETH comes from. Remember, these rewards generated by Yearn are automatically compounded. In order of complexity, starting with 0.5 stETH and 0.5 ETH:

The annual return rate of 0.5 stETH is 4.5%, and the overall return rate is 2.25% (because the underlying 0.5 ETH does not receive rewards);

stETH-ETH’s Curve LP (liquidity provider) earns a small portion of revenue from transaction fees;

Curve LP receives CRV+LDO Token rewards;

Convex LP receives CVX Token rewards.

The user's ycrvstETH yield is 4.51% APY (annual yield). However, this is after the Yearn protocol deducts a 2% management fee and a 20% performance fee. Therefore, the real APY is about (4.51% / 0.8) + 2% = 7.63% APY. Therefore, the total reward for steps 2-4 above is a 5.38% boost. This income is passive, but it needs to be subsidized for the benefit of certain protocols. In a sense, this income is not sticky to the underlying Token pairing. The income comes from rewards and can be paid to any Token pairing. This 5.38% is variable.

Now let me introduce a successor to Lido's powerful stETH-ETH Curve pool - rETH-wstETH pool. rETH is Rocket Pool's liquidity staking token (users will get rETH after staking ETH through Rocket Pool), which will passively increase in value relative to ETH; wstETH is Lido's wrapped version of stETH, and wrapping stETH also allows it to increase in value relative to ETH over time, just like rETH.

Above: rETH-wstETH pool on Curve, https://curve.fi/factory/89

Part of the beauty of the original stETH-ETH pool is that the stETH token is pegged to ETH, so holders do not experience impermanent loss due to asset price deviations. This is critical because an ideal passive collateral token should carry minimal passive value risk. Since wstETH and rETH grow in value at roughly the same rate, the Curve pool between these two tokens (i.e., the rETH-wstETH pool) will inherit the benefit of this negligible impermanent loss. Additionally, the novel enhancement of this pool is that there is no "cold" ETH exposure, and both assets (i.e., rETH and wstETH) will gain staking exposure. Let's see how this 100% exposure to liquid collateral tokens changes the APY calculation.

0.75 rETH earns 4.3%, 0.25 wstETH earns 4.5%, and the total earning is about 4.35%;

Curve LPs (liquidity providers) of the rETH-wstETH pool receive a small portion of the revenue from transaction fees;

Curve LP will receive CRV Token rewards;

Convex LP will receive CVX Token rewards.

Comparing the two Curve pools, the rETH-wstETH pairing is yielding 2.1% more than ETH-stETH. This alone covers 39% of the additional rewards added to the Lido stETH-ETH pool. When dealing with billions of dollars in liquidity TVL (total value locked), it is difficult to overstate the magnitude of growth. In the March budget, Lido had to allocate 3.25 million LDO tokens to the Curve stETH-ETH pool for direct incentives to LPs (liquidity providers). At the current rate, this equates to a monthly expense of $6.9 million. A 39% capital efficiency upgrade saves millions of dollars per month. If these liquidity options were added to rETH-wstETH, their efficiency would be greatly improved.

Another way to look at this is that the problem faced by the stETH-ETH pool is that it must subsidize 50% of stETH returns, otherwise users are better off just holding stETH. The rETH-wstETH pool does not need subsidies because it provides essentially the same (if not more) returns as just holding rETH or stETH. Over time, the funds flowing into ETH staking will be huge, and it will not ignore this obvious improvement in capital efficiency.

Looking ahead

More philosophically, we should learn lessons from DeFi veterans like MakerDAO, who abandoned their original single-collateral DAI model (which initially only allowed ETH to be used as collateral to mint and borrow DAI) and instead moved to a multi-collateral model. This essentially eliminates the risk of a single point of failure. The same risk management approach should have appeared in the liquid staking ecosystem long ago. We are currently in a danger zone, with Lido having 86% of the liquid staking market share and Rocket Pool in second place with only 4.5% market share. As shown in the figure below:

Image source: https://dune.xyz/eliasimos/Eth2-Liquid-Staking

I don’t want to portray the wstETH-rETH pool as a vampire attack on Lido’s market share. Instead, I see this pool as the first step in a larger restructuring of the space. Reducing the ecosystem’s reliance on stETH will reduce the chances of catastrophic events like zero-day vulnerabilities or hostile protocol takeovers. Competition is inevitable, and collaboration will lead to more vibrant gains.

Collaboration is another benefit of the wstETH-rETH pool, which is specifically for those who participate in the governance chat through Discord like me. DeFi runs on liquidity. Lido is willing to spend millions per month to fund the stETH-ETH pool. However, why must there be only one protocol to increase liquidity? One day in the future, a basket of staked tokens in a single liquidity pool may mean a basket of different liquidity incentives from different staking protocols. An LP token that maximally supports the health of the Ethereum ecosystem. This basket of staked tokens represents the long-term security of public goods.

We have an opportunity to get ahead of the curve in situations like the stablecoin market. Instead of competing with each other for liquidity, we must expand the basket to allow different liquid collateralized ETH derivatives to benefit from shared liquidity, thereby reducing risk across the network and continuously dividing the cost of long-term incentivized liquidity.

The ultimate purpose of staking on the Ethereum network is to secure the network. Incentivize the security of the network and encode rules that minimize attack vectors. However, the Ethereum community’s responsibility to continuously secure the Ethereum network must never be relaxed. The network of people running the validator nodes themselves is not just made up of Ethereum enthusiasts. As more value accumulates to the network, more and more individuals will try to attack it wherever they can. We must always remain vigilant. The unrestricted expansion of the Lido network is precisely a tail risk that can be avoided. In fact, the safest and most capital-efficient outcome in the future is for many staking providers to compete with each other and drive down commission rates.

It’s easy to get a false sense of security when you see that there is currently $27 billion worth of ETH locked in the Eth2 deposit contract. But don’t be fooled. The merger has not happened yet. The vast majority of ETH is not staked. We are still early and therefore have a responsibility to safely guide this space into the future. The staking dynamics that exist today represent the early stages of this space. Decisions made now will have a greater impact on the history of Ethereum that has yet to be written. We call on the larger DeFi ecosystem to integrate ycrvwstrETH Token. We call on all protocols that use ycrvstETH Token to also allow the use of ycrvwstrETH Token (if it is safe). This will be a capital efficient future, a future of fair staking providers, and a secure future.