Today we present a short note on what we see as the worse-than-expected adoption of stETH and Ethereum Liquid Staking in general.
stETH and LSD are losing market share as a % of total staked ETH.
stETH has not really achieved a Defi network effect to meaningfully differentiate itself from smaller liquid or non-liquid competitors who tend to offer better rates.
The battle for stETH to overcome WETH in Defi is harder than many have thought. Behavioral inertia, mindshare of native tokens, declining community momentum, and Lido's slow-moving governance model all pose drags on Lido's adoption prospects.
Due to stETH's worse-than-expected network effect, we believe investors should apply a discount, not a premium, on top of Lido's consensus standing as a squared exposure to ETH (ETH price x EL Activity Fee). This discount accounts for the risk of declining "stETH/Total staked ETH" market share due to market fragmentation.
Observation: more people chose to stake non-liquid rather than liquid
From April 2023 to August 2023, Ethereum's staking ratio has grown from 14.13% to 21.32%, a healthy 50% increase over just four months.
However, of the additional 8.6 million ETH staked, the vast majority (67%) went to non-liquid staking. Lido stETH only captured less than 25% of all staked ETH growth. This means Lido's market share of all ETH staked has declined from 35% in April to only 31.5% in August.
Even if we consider the additional Ethereum emission from a higher staking ratio (square root of staking ratio), and exclude volatile factors like execution layer rewards, Lido's total revenue actually declined between April and August despite growing from 5.8M to 8M stETH in circulation.
Why? We hypothesize a few reasons:
Price: Lido's 10% take rate hurt staker incentives. Elsewhere on the market, Validator-as-a-Service providers offer 4.4% APY while Lido stETH only offers 3.8% today.
Lack of Network Effect: Lido's stETH/wstETH has only one real use case today: borrowing USDC/DAI against wstETH on AAVE and Maker. But at a net borrowing rate of 4% - 5% and a maximum LTV of 72%, borrowing/levering up alone has not proven a strong driver for stETH demand. Otherwise, there is no real Defi composability or network effect for stETH (details later).
Where is stETH used today?
Let's take a bird's eye view of stETH in the Defi ecosystem.
55% of all stETH tokens are held passively outside any defi contracts (Storage of Value)
40% of all stETH tokens are held in lending protocols like AAVE and Maker (Lending/Collateralization)
3% are used at liquidity positions on Curve and Uniswap, facilitating ETH-stETH swaps. (Medium of Exchange)
<1% is used in new "LSDFi" protocols like Pendle. (~4% if you double-count Lybra here.)
Lending/Collateral is stETH's most successful use case so far. stETH has surpassed ETH and becomes the No.1 collateral asset on all major lending platforms on ETH L1.
See below, stETH/wstETH is the most used collateral asset on both Maker and AAVE. On the other hand, virtually no wstETH/stETH has been borrowed from AAVE, leading to a rock-bottom supply APY of 0%. This indicates a lack of use cases outside collateralization.
Another caveat here is that collateralization as a use case does not require nor generate much network effect for the token: AAVE can happily take 10 different LSD flavors. The presence of stETH would not hurt rETH's LTV ratio or interest rate. Even as stETH becomes the largest deposited asset, this fact alone does not marginally give anyone more reason to buy, hold, or borrow against stETH over rETH.
Medium of Exchange: no traction on Uniswap
A big part of the liquid staking story is Defi compatibility: one day, stETH will become what WETH is today. A big step in that process is unlocking stETH's utility as a medium of exchange: people will casually hold stETH if they believe this is the most convenient token to swap into any other token, and whales will buy it to provide liquidity on trading venues like Uniswap.
However, on both Uniswap V2 and V3, even though hundreds of new pools are launched every month since April, they are almost exclusively based on WETH. Of the limited stETH presence on Uniswap, it is treated as a commodity to swap from WETH/USDC (the consensus MOEs), not as a medium of exchange itself.
On a side note, CEX has an even lower interest in stETH than DEX LPs. This tells us there is indeed a much higher cognitive barrier for someone to adopt stETH compared to ETH. Newcomers who heard about the magic of Ethereum, or who check CoinMarketCap will likely continue to happily use and hold ETH in their transactions. (Native ETH will soon be supported in Uniswap V4. Goodbye WETH.)
Inter-EVM Currency: no traction on L2s and bridges
Another potential use case of stETH is cross-chain. As more and more of Ethereum's defi innovation happen on L2s, it is strategically important for Lido to have stETH available at the forefront of Defi innovation. However, if we look at bridging data between EVM chains, WETH and USDC are safely the dominant inter-EVM currencies, and stETH is nowhere to be found.
New Frontiers: Pendle and Lybra
A final piece of the puzzle is what markets call "LSDFi" innovations. One may say "Forget about old Defi toys, stETH finds its runaway success in new places?"
Sure. Two top projects on this front today are Pendle and Lybra. Let's talk about them.
Pendle is an interest-rate swap market based on APY-generating assets. Pendle's liquidity is dependent on stETH's market cap. Looking at real-world interest rate swaps, it is unlikely a derivative rate market would drive the adoption of the underlying asset in reverse.
Lybra's heavily subsidized ponzinomics will come under increasing pressure as stETH's yield secularly go down. While we recognize there is a theoretical chance of eUSD achieving breakout success, the chances of launching a new successful stablecoin in Defi today are just so low. (Look at Maker and Luna) But the systematic downside risk of mass redemption/liquidation of eUSD dragging stETH into hot waters is significant.
To sum it up, we have not seen encouraging signs that "LSDFi innovations" can drive the adoption of stETH in the near future.
Let's try to guess why: behavioral inertia and the deja vu of early Maker
History: a difficult game
Liquid staking is a hard game. Look at Solana. It has a high staking APY (started at 8%+), a high staking ratio (70%+), and very low friction (<$0.001 gas). Yet Solana's dominant LSD provider, Marinade Finance's mSOL, has only <$200M TVL against >$7B total staked SOL tokens. Much worse than Lido's numbers. Similar to stETH's difficulties today, even during Solana's Defi heydays, the mSOL token was only found as loan collaterals and like-to-like swap pools.
The same story is broadly true elsewhere with PoS frontrunners including Polygon, BNB, Polkadot, Avalanche, and Luna.
Behavioral: Network inertia
Lending is relatively easy. stETH is a transparent synthetic asset on the back of ETH's credibility. AAVE and Maker voters can easily approve a new asset. There is no obvious network effect required here.
But DEX and L2 are hard. To facilitate a transition from WETH to stETH on Uniswap or an entire L2, Lido needs to convince everyone to move at about the same time. Otherwise, inertia and liquidity fragmentation will discourage everyone from moving early.
Mindshare and learning curve: everyone with ETH knows ETH is a legit native token representing the chain. The learning curve for stETH is higher, especially for people new to the on-chain world. Especially as Shanghai passes and attention shifts, LSD adoption will further lose momentum -- fewer Defi protocols are actively talking about LSD integration today.
Paying gas fees and interacting with CEX still require native tokens. Like how most people use only one credit card for everything, stETH can seem optional in contrast.
Logically, speeding away from the network inertia requires aggressive efforts in adoption incentives, partnership BD, brand awareness, L2 expansion, and more. But such initiatives have historically proven quite incompatible with a slow, steady decentralized governance model Lido has chosen. Lido's situation today reminds us of early MakerDAO.
Both are building a public good for Ethereum and Web3.
Both were up against a formidable competitor (USDT and ETH) with a network effect.
Both didn't choose to go beyond ETH L1.
Both are slow in Defi integration beyond lending.
Will DAI's decline be stETH's tomorrow? We hope not. But we are worried.
Conclusion & Compare Notes with Bull Thesis
Finally, we want to compare our notes with some of the most prominent LDO bull thesis. For the latest refresher on bull thesis, we direct you to the August 8 piece by Bryan Tan and Arthur from DeFiance.
We want to point out 2 mistakes in the bull thesis and add one strategic risk.
Staking rewards don't grow linearly with ETH staking ratio. The numerical relationship is actually the square root of ETH staking ratio. For example, if ETH staking ratio increase from today's 22% to a whopping 88%, the total reward will only double, not 4x. This caps a major upside for Lido's revenue.
LSD market share is shrinking, not growing. We believe here Bryan used the wrong comparative. Although stETH fared better than Coinbase and Binance's LSDs, LSD is losing market share against non-liquid Validator-as-a-Service solutions. We have presented this in the first section.
Strategic risk: as we investigate the reasons behind LSD's market share decline, we find that stETH has not successfully built a moat where network effect matters (DEX, CEX, or cross-chain lockup).
The lack of network effect will likely hinder stETH's growth going forward.
Without network effect, stETH's current use cases of SOV and Lending Collateral are pretty commoditized. Lido will continue to face competition and likely fee compression from both non-liquid staking and LSD competitors.
Of course, there are reasons to be bullish on Lido that we didn't refute. We agree it feels like a squared exposure to ETH: benefiting from both ETH appreciation and ETH network activity fees. For the purpose of investing, we believe our piece offers two contrarian takeaways:
Until stETH's network effect happens, Tier 2 LSDs will continue to have a chance to fight price wars against Lido and fragment the LSD landscape.
Until stETH's network effect happens, Lido's squared ETH exposure should be discounted with a risk of its chronically losing market share as a % of staked ETH (to both non-liquid and LSD fragmentation). This risk may manifest directly with loss of market share and thus revenue, or a cut in take rate to match competition (which seems to be discussed in governance already?)
The bear market is not over yet, and Lido still has time and money to build its network effect and set it up for positive feedback loops when markets recover. We will keep watching. If you're interested in our ideas, please get in touch.
Correction: a previous version of this post miscalculated stETH's April market share to be 41%. Our best-attempt figure should be 35%. The corrected figure shows a slower rate of decrease in Lido's market share. Other parts of the analytical process and our conclusion remain unchanged.
Thanks to Lawrence from Mint Ventures for pointing that out. Obtaining correct data about Lido has been quite difficult among a number of conflicting sources (Etherscan, Llama, Various Dune Dashboards, CryptoQuant, etc.) We tried our best. Your discussion and correction are greatly appreciated to make crypto analytics better for all.
Research Lead, Maverick Crypto
DMs are open.