Thread by @anchor_protocol

1/ Over the weekend, Anchor underwent another round of duress as markets plunged deeper. We’ve received a ton of valuable community feedback: what worked well, what didn’t, what can be improved, and even cool new features developed by community members

https://twitter.com/terra_money/status/1396780930736410629

2/ Besides the market turmoil, we will cover some exciting developments ready to deploy and just over the horizon. Let’s take a look at how Anchor performed and where it’s going 👇

3/ Over the weekend, sell-offs across crypto markets accelerated. LUNA’s sharp price decrease was compounded by Anchor liquidations, which triggered cascading LUNA/bLUNA liquidations of open borrower positions with LTV ratios above the required threshold.

medium.com/@terra-bites/l…

4/ A popular Anchor strategy to capture larger returns during positive price momentum (borrow UST, buy more LUNA, mint more bLUNA; rinse & repeat) induced a deluge of liqs as leveraged positions were queued up by the Liquidation Contract - triggering further market sells of LUNA.

5/ Liquidations were accelerated by the max premium rate of 30% making it profitable for bots to liq. open borrow positions & sell LUNA/bLUNA on the open market. The 30% rate was intended to support fluid liqs at launch (when bLUNA was scarce) & will be reduced as Anchor evolves.

6/ High liquidity of the bLUNA/LUNA Terraswap pair during volatility acted as a gateway for liquidators to easily sell bLUNA tokens to UST, incr. slippage b/w bLUNA & LUNA. With ⬆️ slippage due to volatility, users faced swaps at ⬆️ rates, forcing sells at excessive losses.

7/ Concurrently, large chunks of liquidation transactions began filtering into the mempool, which squeezed out some oracle transactions. However, as we adapted Anchor’s back-end infrastructure to the duress last week, the web app remained stable throughout the volatility.

8/ Finally, although the deposit rate held steady at the threshold rate (18%), where the yield reserve was supporting the rate, we believe it’s in the best interest to slightly raise the threshold and target rates via community vote.

9/ So, how do we address the above issues?

10/ First, augmenting the Anchor liquidation mechanism.

Second, the Anchor team will be initiating a series of governance proposals on the Anchor community forum to perform the following:

11/
1. Raise the deposit threshold rate to 19.5% and target rate to 20.5% to maintain a closer peg to the 20% APY. This will enable the yield reserve to support a threshold rate closer to the protocol’s initial target rate during periods of volatility.

12/
2. Increase liquidation threshold above 500 UST (exact number undefined yet), which will decrease the amount of concurrently triggered liquidations, since all positions below this number will be liquidated instantly -- helping with network congestion.

13/ The above will also help curtail oracle vote transactions being squeezed out of the mempool, something that will also be allayed by Tendermint’s next upgrade feature in Columbus-5 -- mempool prioritization.

14/
3. Adjust max premium rate to below the current 30% (the exact number will be available on the governance poll) -- make the liquidations less lucrative so that even if liquidations do happen there is less of a negative impact on the system from these liquidations.

15/ Note -- the liquidation premium is capped to avoid zero-bid auctions of borrower collateral like what happened with MakerDAO in March 2020.

blog.makerdao.com/the-market-col…

16/ We’re currently outlining the proposals and will release more details soon.

Third, community feedback indicates that even though the web app was functioning properly, manual pay downs of outstanding positions is cumbersome during market turmoil.

17/ We'll be addressing this in 2 ways:

1. Auto-repayments of loans to avoid liquidation. Instead of a direct integration into the web app, a URL will be added to a community-built auto-repay software.
2. Introducing WalletConnect - mobile functionality for paying down loans.

18/ In regards to the auto-repayment of loans, a community member designed an elegant solution -- coding a loan auto-repay script that withdraws aUST and pays off the loan back to a percentage that you want.

19/ In fact, it helped the community member avoid liquidation when Anchor’s web app went down last week since the smart contracts were still functional. We’re currently in contact with this community member and will release more details soon.

20/ Concerning #2, WalletConnect is a type of protocol that was originally designed for inter-device connectivity -- web to mobile. We have already applied it to our mobile Anchor web app version, and it is set to go live later this week.

21/ Basically, WalletConnect functionality will allow users to use Station mobile with Anchor web app — allowing them to access all functions of the web app, including paying down loans from a mobile device.

22/ Finally, after a significant number of positions were liquidated during the volatility, borrower demand has subsided noticeably, with the utilization ratio currently around 26%.

23/ Yield reserve can support threshold deposit rate, but for Anchor's long-term robustness, borrower demand must recover and remain strong. Flushing of leveraged positions is natural in adverse market conditions, so borrowing demand will return, but explicit steps are necessary.

24/ Critically, the inclusion of more staking derivatives as collateral on the supply-side, such as bETH, bATOM, bDOT, bSOL, and more, which are currently on the Anchor team’s roadmap.

25/ A wider selection of staking derivatives as collateral on Anchor reduces its dependence on a single asset (bLUNA) & expands the demand for borrowing by making Anchor a locus of cross-chain liquidity for previously illiquid, bonded staking positions.

medium.com/anchor-protoco…

26/ Subsequently, the utilization ratio of the protocol will replenish to healthy levels, supported by the improvements iterated above that help minimize the impact of liquidations on users -- reducing downward reflexivity of the LUNA price during market downturns in the process.

27/ In the longer term, we will also be researching an improved liquidation contract that will take into consideration the premium rates, and therefore will reduce the caliber of premium rates and the amounts of liquidation TXs being generated.

28/ You may be wondering -- why so much granular transparency into the inner mechanics of Anchor during market duress? A few reasons.

29/ Not only is it good for the community to have insight into what’s happening but it helps the community and prospective builders to learn, assist, and improve on Anchor’s current implementation.

30/ Building and iterating on top of Anchor just like we’ve built Anchor on the shoulders of other ideas is the cycle of innovation -- it fuels progress. Without it, Anchor cannot adapt to the wild theater of crypto markets.

31/ It’s the beauty of open protocols & communities behind them. They’re evolving organisms, adapting to their environment in ways that allow them to remain robust in the face of turmoil. Thanks to everyone for their constructive feedback; let’s keep up the momentum.

Onwards.