Welcome to the third edition of Fluff and Air, a recurring publication by Carnation and ct_zpy. Today we’re looking at how the moon people successfully tested liquidity across all our chains.
Just a heads up: This post is just for education, and this post is not financial nor astrological advice. We may hold tokens of some of the projects mentioned here.
After a certain illiquid algorithmic stablecoin that we covered in the previous piece exploded, we have decided to do a follow up and assess the damage to the ecosystems of our favourite digital currencies.
A little thought experiment about the Illiquid Factor
We gave a little bit more thought as to what the illiquid factor represented, and came up with a thought experiment to demonstrate it.
Scenario 1)
Suppose on a chain we have only one Uniswap V2 style AMM with a single ETH/USDC pool, and there is 1 ETH and 2000 USDC in the pool. The TVL of this pool is 4000 dollars (2000 USDC and 2000 dollars worth of ETH).
If you tried to spend 2000 USDC to buy ETH, x*y=k says you can get 0.5 ETH out, which is a 50% slippage.
So the illiquid factor is TVL/(amount to cause 50% slippage) = 4000/2000 = 2
Scenario 2)
Suppose we examine our pool which now has 0.5 ETH and 4000 USDC, the TVL is now 6000 dollars, and it would take a 4000 USDC buy order to cause 50% slippage.
The illiquid factor now is 6000/4000 = 1.5
So the illiquid factor goes down when there are more stables relative to the volatile token inside the AMM.
However, if one treated the TVL as not just the tokens locked into the AMM, but as all the assets on chain, then the scenarios would look as such:
Scenario 3)
Same setup as scenario 1, but we include the USDC that went into the swap as part of the TVL
TVL = (2000 USDC in AMM)+(2000 worth of ETH in AMM)+(2000 worth of ETH in wallet) = 6000
Slippage for 50% = 2000
Illiquid Factor = 6000/2000 = 3
Scenario 4)
Same setup as scenario 2, but we include the ETH that came out of the swap as part of the TVL
TVL = (4000 USDC in AMM)+(2000 worth of ETH in AMM)+(2000 worth of ETH in wallet) = 8000
Slippage for 50% = 4000
Illiquid Factor = 8000/4000 = 2
The point we wish to highlight here is that regardless of where one draws the boundaries for the definition of TVL, the illiquid factor correlates with the ratio of volatile token to stables inside the AMMs.
The TVL data we gathered from Defillama represents the funds that are actively participating in the DeFi ecosystem of all these chains, and probably is closer to scenarios 3 and 4 rather than scenarios 1 and 2.
It is our opinion that the illiquid factor represents some level of risk in participating in DeFi. For example, when a chain that has a high illiquid ratio, it means that a smaller proportion of funds would be able to get out with less than a 50% slippage.
We leave as an exercise to the reader of how adding or removing liquidity to the AMM affects the illiquid factor.
Methodology
We followed the same methods as our last article. Market caps were pulled from CoinGecko. TVL was pulled from DefiLlama. The DEXes used to determine slippage were DEX aggregators where possible (like 1inch, Kyberswap, and Jup.ag), or the largest DEX on a chain if not possible.
All numbers were pulled between 5/15 and 5/16. We imagine the general ratios between MC and TVL will remain relatively stable, but the underlying data will quickly become stale as prices change.
We checked the slippage when selling tokens that fit into the categories of: primary gas token (gas token of the chain), secondary gas token (gas token of the underlying L1), and governance token of the chain. We examined a total of 17 chains and 23 chain-token combinations.
We attempt to sell the token in question for the most liquid stablecoin on the same chain.
Taking a similar approach to (link to thread), we asked the following Qs:
- What happens when you try to dump 1% of the Market Cap on chain
- What happens when you try to dump 1% of the TVL on chain
- How much dumping will bring slippage up to 50%
Chains were then assigned an “illiquid factor”, which was calculated by dividing TVL by the amount of money required to reach 50% slippage. A really high illiquid factor means that even if a tiny fraction of TVL decided to exit, they would cause terrible slippage and also take away a large fraction of stablecoin liquidity. The naïve interpretation is that high illiquidity factors are bad.
We did add a few more columns that allow us to compare how the illiquid factor changed from pre-crash to post-crash.
Observations
We will try to pick out a few interesting things that caught our intention.
We welcome the readers to dig into the full data here: On-chain Liquidity and Slippage Comparison - May 15 2022
We’d also like to refer the reader to the work and discussion surrounding similar analyses on Twitter:
Ethereum
Ethereum drew down roughly 28%, but is only slightly less illiquid than before, with a small 3.5% illiquid factor increase.
Tron
Tron held up very well, likely because Justin Sun is essentially the chain personified. If His Excellency does not capitulate, then the price will not go down. His Excellency indeed did not capitulate, so the price did not go down. TRX drew down a mere 2%, and its illiquid factor actually improved by 3% post lunar explosion.
We take this as a sign that contrary to several YC startups, His Excellency’s farming funds are safu.
BSC/Sol/Matic/FTM
BSC, Solana, Matic, and FTM drew down 23%, 38%, 37%, and 56% respectively. However these four chains saw an increase in liquidity, their illiquid factors improved by 12%, 21%, 22%, and 11% respectively. This could be due to a wide variety of reasons. One possible reason is that the TVL of BSC, Solana, and Matic proved to be stickier than the rest.
BSC performed especially well, seeing less TVL drawdown than the rest while also holding its price against BTC quite well. A pessimistic explanation to this is that BSC shitcoins are so far removed from everything else in crypto and form such a large percentage of BSC TVL that a $40B asset can collapse and have extremely little effect on BSC shitcoin prices and the BSC TVL.
FTM on the other hand was absolutely obliterated. A possible explanation is that although FTM TVL got destroyed, the stablecoin liquidity of FTM was less affected. People removed less than the proportion of stablecoin liquidity in comparison to how much FTM went down in price.
Terra
Our sincere condolences. Goodbye Moonmen
On a serious note, suppose one wanted to manage risk through on-chain perps and hedging protocols. How effective would it be to purchase insurance when the payoffs are denominated in UST (not to mention the chain was halted and funds were not accessible)?
This highlights an argument that hedges against a chain failing cannot be placed within said chain.
Osmosis
In the previous piece, we looked at OSMO/UST. Obviously, this pool exploded. We also mentioned the fact that ATOM/OSMO was the most liquid pool on Osmosis - many users exited LUNA/UST through ATOM and OSMO. This caused prices of ATOM and OSMO to dump hard and ct_zpy’s familia will never recover from this loss.
A side note, a lot of liquidity providers in the LUNA and UST pools really got the short end of the stick because of the bonding unlock periods. We suppose the lack of hedging methods in IBC didn’t make things any better.
Near and Aurora
On Aurora, liquidity got better for the NEAR token and worse for the ETH token relative to the beginning of May. While on Near, the liquidity got worse for NEAR. We speculate that this indicated that the general opinion is leaning towards Aurora being a NEAR-centric ecosystem rather than ETH-centric.
Layer 2s
The ETH liquidity has actually improved on Arbitrum and Boba, while getting worse on Optimism and Metis. It was surprising that the liquidity of ETH was not sticky on Optimism and Metis, but at the same time went up for Boba. Maybe someone who knows someone who might know something.
On Arbitrum, we’d like to point out this: GMX was integrated into 1inch, and the max size you can push through is roughly 20K ETH or 40M USDC. The slippage on that is ~0.5%, For comparison, a 20K ETH swap to USDC would be roughly 2.5% slippage on Ethereum mainnet. This actually makes Arbitrum the singular best place right now for low slippage large ETH trades.
Concluding thoughts
The markets presented us with a unique opportunity to study the composition of liquidity on-chain from a very broad level. While something like the illiquid factor might be too broad to offer any specific actionable insight, we were able to examine some outliers of extreme improvements and deteriorations of liquidity across the various chains. Potentially, the illiquidity factor can serve as something like a “health indicator” for DeFi ecosystems.
We may do one more round of data collection on this series to study in a few weeks to see how the liquidity flows after the markets had more time in the post-crash recovery phase.
