As aforementioned, stETH is the token issued by Lido when you provide ETH to them to participate in ETH2.0. In essence, stETH IS ETH, as it can be redeemed 1:1 to ETH. Due to this fact, the stETH-ETH pool is used by Curve users as a means to increase their positions in ETH passively. By providing liquidity to stETH-ETH pools, no impermanent loss will be suffered by the providers, in theory, thus liquidity providing is a relative safe way to generate income for ETH tokens.

Introduction

With that said, less risk means less yield, so it may not be good enough for those with a higher risk tolerance. That's where Anchor comes in. The Anchor protocol is a lending platform launched in the Terra blockchain, where users can deposit bETH as collaterals for borrows. Borrows on Anchor are incentivized with ANC token yields and when combined with the fact that one can earn a constant 20% APY by depositing the borrowed UST into their Earn function and the fact that one can leverage their positions with Anchor, collateralizing bETH is a much better choice for higher risk takers in order to maximize their yields. But what does bETH have to do with stETH? You see, when one bridges stETH from the Ethereum chain to the Terra chain, the tokens are converted to bETH, thus we can say that bETH is equal to stETH.

stETH-ETH Pool Analysis

CRV allows one-sided provision/removal of liquidity, meaning that one can provide/remove liquidity using only one token instead of needing to provide/remove half of each token, which also means that the liquidity provided/removed by users aren't always 50:50. The graph below shows the preference of liquidity providers since 2021-08-12. For the first half of the this period, users seem to be more inclined to provide stETH than ETH while the later half are mostly provided in ETH. All in all, the spread of liquidity providing seems to be roughly equal for users who prefers to provide stETH and users who prefers to provide ETH.

Besides the period from 2021-08-13 to 2021-08-21 (reason this period is excluded is shown later), the volume for the stETH-ETH pool shows a much more obvious distinction between those who favor one token over another. Judging from observations, stETH is hardly ever bought, while the complete opposite happens to ETH. Well, this is understandable as stETH can always be minted 1:1 using Lido, therefore it bypasses the trade fees charged by Curve as well as the slippage occurs for swapping the tokens. In ETH's case, without any means of redeeming ETH using stETH currently, the users who wish to "redeem" their stETH for ETH must go through Curve, even if it means getting charged trading fees and getting hit by slippage.

Anchor

With the launch of bETH as collateral on Anchor on August 12, we've seen an explosion of stETHs being bridged over to Terra during the first week of its launch. This in turn increased the volume over at the stETH-ETH pool on Curve as users scramble to get some stETHs, where ETHs are getting sold/swapped into stETH tokens to be bridged to Terra.

Possible reasons why ETHs were sold instead of being sent to Lido to mint stETHs are that

1. Users do not know / want to know how to mint stETH tokens, so they do the next logical thing, which is by swapping them.
2. Users prefer Curve over Lido.

Comparing the stETHs obtained through swapping and the amount bridged to Terra, we are able to determine that about 37% of the stETHs are obtained through swapping while the rest were minted.

By comparing the amount of stETHs bridged to Terra and the number of bETHs provided as collaterals, we can safely say that most of the stETHs that are bridged over to Terra are being used as collaterals. By August 22, 76.59k out of 77.52k (98.8%) stETHs bridged over to Terra were used as collaterals, showing that Anchor's acceptance of bETH as collateral did have an effect on the stETH-ETH Curve pool. This relationship is once again proven on September 1, where the bETHs were bridged back over to the ETH chain and subsequently swapped back into ETHs. If that's not enough evidence, the spike in bETH deposited on Anchor on September 9 can also be linked to the sharp decrease in liquidity in the stETH-ETH pool on the same day, where 10k stETHs were removed from the liquidity pool while 13k bETHs were deposited into Anchor, showing that Anchor not only has influence over the volume of the stETH-ETH pool, it has influence over its liquidity as well.

Besides providing bETH as collateral for Anchor, bETH can also be provided to the bETH-mETH and bETH-UST liquidity pools to generate yield. Based on the graph below, out of the 94k stETHs bridged over to Terra, only about 607 of them are being provided as liquidity. Compared to numbers on Anchor, the amount of stETHs in liquidity pools is miniscule, where it's only about 0.6% of what's on Anchor. This shows that the effect of bETH on the stETH-ETH liquidity pool is mostly caused by Anchor.

Conclusion

1. There is no significant difference between the preferred token to be used in liquidity providing for the stETH-ETH pool.
2. When there are no external influences, the stETH-ETH pool is mostly used to "redeem" stETH tokens back to ETH.
3. stETHs are not being bought from the pool because it's not beneficial to do so, as swaps are hit with slippage and trade fees while sending the ETHs over to Lido will always result in a 1:1 exchange.
4. Anchor has a huge influence over the volume of the stETH-ETH pool, where its launch significantly increased the volume of the pool for about a week, with a few spikes here and there after that period.
5. Anchor has influence over the liquidity of the stETH-ETH pool as well, shown by the fact that the spike in total deposited bETH on September 9 coincides with the decrease of stETH of the said pool.
6. The spikes for both volume and liquidity of the stETH-ETH pool is undoubtedly caused by Anchor and not Terraswap as the amount being provided as liquidity on Terraswap is much lower than that of Anchor.