What has gone wrong with Lido Finance?
2020 is colloquially referred to as the ‘Defi Summer’ — this was the year when Defi took off. A sudden adoption of popular apps such as UniSwap created a ton of interest among retail investors.
Exactly 2 years later, Defi is under extreme pressure and almost every Defi enthusiast now talks about #DefiWinter. The pains of growing ‘too big, too fast’ are showing clearly among Defi protocols. $UST/Terra created a big crack in the #Defi Armor and the impact of that crack is felt across the crypto space.
In this article, I discuss how Lido Finance is under huge pressure and why the ‘liquidity staking’ business has increased leverage and contributed to the fragility of the #Defi ecosystem.
The Eth 2.0 Merge
One of the biggest anticipated event in Web 3.0 world is called ‘The Merge’. It is when Ethereum moves from a Proof-of-Work to a Proof-of-Stake consensus. Proof-of-Stake (POS), as the name suggests, works when a large number of users stake their Ethereum to validate new blocks and achieve consensus. For POS to work, Ethereum Foundation has asked all interested holders to stake their ethereum to the chain.
Apart from receiving staking rewards, Ethereum stakers also had to agree to lock the ‘Staked Ethereum’ until the merge is complete. While ETH maxis & long term ETH HODLers were OK with the lock-up condition, normal investors were not too gung-ho about it.
With extreme volatility in crypto space, lot of people were uncomfortable holding an illiquid asset for a long period. Rise of Alt L1’s that were constantly nipping at the dominance of Ethereum was also adding to this uncertainty.
This problem was solved by ‘Liquidity Staking’ platforms, Lido Finance being the largest among them.
Liquidity Staking
Ethereum 2.0 required that stakers deposit a minimum of 32 ETH to become a validator. For most crypto investors, this would be a difficult threshold to cross (at the time of writing, this was ~$65k)
LIDO made it easy to stake Ethereum by bringing 2 major innovations:
- There is no minimum ETH requirement for investors. ETH would be pooled and given to a LIDO trusted node providers (validator).
- While the original ETH deposited gets locked up, LIDO will mint a new token stETH that is directly pegged to ETH ( 1 stETH = 1 ETH). The best thing about stETH is that this token can be used by investors as collateral/yield generating asset on other protocols.
These 2 benefits made LIDO a game-changer for staking. Investors would not only enjoy staking rewards but also additional APY by deploying stETH across Defi protocols such as Aave, Curve etc.
stETH wonder
stETH stands for staked ETH token. stETH is minted only when corresponding amount of ETH is locked by LIDO for staking on Ethereum blockchain. This 1-to-1 backing ensures a stETH:ETH peg of 1:1 on paper.
This 1:1 peg allowed other protocols such as Aavesome to accept stETH as collateral against which users can borrow. Additionally, Curve Finance started a stETH-ETH pool — users could now swap their stETH to ETH anytime.
Using stETH as a proxy asset created a lot of leverage in the system that can be dangerous in extended bear markets (I’ll explain this later)
With the incoming ‘Merge’ and additional APY incentive, investors rushed to LIDO to stake their ETH in exchange for stETH. LIDO quickly built up its TVL and became the biggest staker for Eth 2.0, controlling over 4 million ETH deposits & accounting for over 30% of the total staked ETH.
Lido’s Terra Exposure
With the enormous success of stETH, LIDO started opening up its Liquidity Staking services for other chains.
Next big bull run came from ‘Terra’ which was also growing at a phenomenal pace, thanks to its ‘stablecoin savings plan’ on Anchor Protocol offering investors a stable 20% APR.
Just like the ‘Merge’ was a catalyst for stETH’s rapid growth, Anchor was a catalyst for the rise of stLUNA. Users could stake their LUNA (native token of Terra ecosystem) and receive stLUNA.
There were a ton of influencers and Youtubers creating ‘alpha’ for investors by advising them to convert LUNA to stLUNA and then move the stLUNA to the Anchor network to earn 20% APY.
Investors started converting LUNA to stLUNA and then converting stLUNA to bLUNA (bonded LUNA). bLUNA was then used as collateral on Anchor Protocol to borrow UST and put that UST back into Anchor to earn a 20% return. This round-tripping of capital created more and more leverage and put the Anchor Protocol on steroids.
Users also started to put stETH as collateral on Anchor protocol to borrow UST and re-invest back into Anchor. This positive feedback loop was rapidly increasing TVL for LIDO as well.
Entire Defi is a bunch of legos stacked on top of each other. Eth 2.0 expectations led to LIDO’s first wave of growth, Anchor led to the first wave of LUNA growth. Anchor’s growth fueled the second wave of LIDO’s growth which inturn boosted Anchor’s growth. So everything in a ‘Lego’ structure builds on each other.
As on the date of writing, Lido Finance has taken out all information regarding stLUNA from their website. I have this screenshot from one of my presentations (taken in March) that shows a significant TVL exposure.
The Terra Collapse
Terra’s spectacular collapse meant all those stLUNA’s that were issued against LUNA turned worthless. Thanks to LUNA’s death spiral, over 6 trillion LUNA are currently in circulation.
Any weakness in a lego structure quickly cascades and exponentially increases fragility —we experienced this fragility with LIDO in the week following Terra’s collapse. Entire stLUNA of ~9 billion value as on 30'th April got reduced to 0, thanks to LUNA’s hyperinflation. LIDO lost >50% of TVL in a matter of 5 days.
Thankfully, companies like Celsius that had deposited over 216,000 ETH (read stETH) in Anchor could recover almost all of their stETH intime. Had Terra chain collapsed or went into a forced shut down, stETH holders who posted their collateral on Anchor would have been stuck and helpless.
Loss of Confidence
In the aftermath of $UST collapse, investors seem to have lost confidence in LIDO. I say this because of two reasons:
- LIDO’s dominance in the Ethereum Liquidity Staking business seems to have taken a massive hit. LIDO completely dominated the ETH staking
2. It’s easy to think that this is because of a high ‘fear index’ in the current market. But LIDO’s competitors, notably Rocketpool has seen a significant rise in market share right after $UST collapse. ETH 2.0 stakers seem to prefer other platforms to stake their ETH. Whether this loss of trust in LIDO is temporary or permanent, only time will tell.
While this was happening on the ETH staking front, another interesting phenomena was unfolding on the Curve $stETH:$ETH pool. Let’s look at that…
The de-peg
Starting the 11'th of May, right after $UST collapsed completely, $USDT, a centralized stablecoin started to de-peg. While the world’s attention was on the largest stablecoin, there was a silent de-peg happening elsewhere. The $stETH:$ETH 1:1 peg started to go off-limits. It fell for the next 2 days and went alarmingly close to 0.95 (point where time-bomb activates, more on this in the next section) before it recovered.
As mentioned above, Curve Finance has an active trading pool based on $stETH:$ETH pair. Since conventional wisdom is that both assets move together, the impermanent loss for investors in this LP pool is low (ignore this if you don’t understand what impermanent loss is). And since $ETH is one of the biggest coins out there, this pool has deep liquidity ($2 billion at the time of this writing).
Right after $UST collapse, investors got worried that Terra might take down LIDO with it (remember the intertwined lego structure). This introduced dear and uncertainty for investors— investors swarmed to the $stETH-$ETH pool and started dumping $stETH in-exchange for $ETH. Since the AMM pricing model of Curve LP re-adjusts the price based on pool inventory, $stETH experienced a continuous price drop viz-a-viz $ETH.
What was supposed to be a 50:50 pool now started getting more and more skewed. Even at the time of writing this article, the pool continues to remain heavily skewed with a 30:70 ratio of $ETH:$stETH.
Normally Curve AMM ensures a smooth price movement — it takes a huge sell-off to move the price to 0.95. The fact that this level was reached and that the peg is still at 0.97 levels, a week after $UST collapse, tells us that investors are unsure about LIDO business model. This also explains a loss in market share of LIDO in their core ‘ETH staking’ business. In the current environment, investors don’t seem to mind the 3–5% loss when swapping $stETH to $ETH.
Investors seem to think its better to hold ETH or convert it into cash/stablecoin than stay exposed to $stETH.
LIDO has taken some countermeasures by offering more incentives to liquidity providers. But incentives only work in normal markets and don’t usually have a desired effect in extreme ‘bear’ markets.
The stETH-ETH time-bomb
This takes me to the last section — what happens if $stETH peg goes down further. To answer this, we need to look at all the places where $stETH is deployed in Defi.
One of the major $stETH usage is in Aave. A total of 1.37 million stETH worth $2.6 billion is locked up as AAVE collateral.
The staking yield on $ETH was ~4% whereas variable borrowing rate on Aave was ~1.7–2%. Investors used a ‘leveraged staking’ strategy where they deposited $stETH and borrowed $ETH, staked it again. The new $stETH again found its way into Aave as collateral and fresh $ETH was borrowed. This process was repeated in a loop to maximize asset yields. Unfortunately, this also meant an extremely high leverage and a low margin for error.
As per recent risk report published by Lido (before the de-peg), liquidation threshold is at 75%. This means that whenever loan to collateral value ratio becomes equal to 75%, loan is considered under-collateralized and liquidated.
Since the general assumption has always been $ETH:$stETH = 1:1, investors were not worried about the the prospect of liquidation. Because $ETH and $stETH are always supposed to go up and down together, it was assumed that this leveraged borrowing is safe & will never get liquidated.
The stETH-ETH parity was taken for granted & investors leveraged their $stETH to maximum possible limit with very low margin for error.
Second order effects
Let’s run a scenario here:
Say, $stETH:$ETH is at 0.95. Say 10% of $stETH borrowings ($260 million) had a LTV of 70% when $stETH:$ETH = 1:1. These are investors who have deployed maximum leverage and at the highest risk. Here is a possible sequence of events
- A liquidation event is triggered and the liquidator gets the assets ($stETH in this case) at an average 6% discount.
- Liquidator will most likely go to CURVE to swap $stETH:$ETH and realize instant profits. Since there is a constant selling pressure on $stETH, Liquidator is happy exiting his position as quickly as possible. This will add an additional $260 million to the Curve pool that is already short on Ethereum
3. This adds further selling pressure on $stETH -> price goes down further -> more liquidations -> price goes down even further
4. This will continue until enough demand gets generated where users buy $stETH at a discount and hold it to exchange for $ETH after merge. This will only happen when liquidity is available and buyers are willing to lock their liquid assets for an illiquid asset.
In extreme liquidity squeeze scenarios, such death spirals can significantly increase risk and fragility in the system. Everything hinges on $BTC — if $BTC price is on a down spiral, all leveraged assets can collapse because of such hidden, second-order risks. If $BTC is anyway falling, investors will prefer to use cash to buy a falling $ETH instead of a de-pegged $stETH
I hope we reach peg quickly and LIDO’s incentives boost liquidity and investor confidence.
Conclusion
To conclude, here are my key observations about LIDO
- A big beneficiary of Eth 2.0 merge
- Overstretched itself by dipping into the explosive growth of Terra
- Lost 50% of TVL and market confidence post Terra debacle
- Losing market share for its core ETH liquid staking business
- Risky environment is making people dump $stETH for $ETH at a discount
- LIDO incentives are not enough to reduce ‘fear’ among investors
- ‘stETH’ collateral has built a ‘leveraged Ethereum Staking’ Model on Aave
- Potential for mass liquidations and $stETH trading at deep discount to $ETH if $BTC falls further
- All the above is making Defi even more fragile in the current scenario
