LUNA Black Swan: Terra’s Do Kwon Gives AMA on What We’ve Learned & Anchor Action Plan
What a week. May 19 through May 23 saw a spiral downward in LUNA price and Anchor Protocol borrow utilization. Cascading liquidations forced more and more users to have their bLUNA sold off, adding to the pain.
As a result, LUNA reached a low of $3.99 on May 23 before rebounding, sitting at $6.47 as of this article. Congestion brought the Anchor Protocol site to a halt for short periods of time, making it difficult for people to cover their positions.
The Anchor website wasn’t alone. Coinbase, Binance, Kraken, Gemini, and more experienced downtime, as well.
Ethereum DeFi became so congested that gas fees made it economically unviable for many people to save their positions.
Disclaimer: Terra Bites is for informational and educational purposes only. Nothing in this article is advice — no financial, tax, legal, or any other kind of advice. Nothing we say is an inducement to make a particular investment or follow a particular strategy. Anything you invest in is at your own risk; do your own research and consult with professionals.
This situation is still developing and some information is new; some material may be subject to correction or change.
Anchor Liquidation Action Plan
The main piece of the AMA was an action plan for improvements to Anchor Protocol.
Liquidations are essential for Anchor Protocol to fulfil its core mission: to provide stable, attractive yield on stablecoin deposits.
However, the 30% discount on liquidations which was intended as a starting point has generally been the final discount. Bidders have no competition. And whales can acquire large amounts of bLUNA at a discount and make significant profits without much of a balancing mechanism.
The team’s action plan for Anchor is to:
- reduce the bLUNA<>LUNA pool, making it impractical to use the liquidation strategy at such a large scale
- reduce premiums in general and force liquidators to bid for the premium they want to get the liquidation at
- work with the community to integrate auto-repaying loans
- work with infrastructure-focused node operators to add nodes for increased stability on Anchor Protocol
- add liquidation history/transaction history view to Anchor, displaying cost of liquidation and other stats
- ensure that the lowest liquidation premium will always win the bid
- release a mobile webapp deep linked to Terra Station mobile app, scheduled to ship Wednesday, allow Anchor positions to be managed on the go. (In addition, Terra Station app updates coming soon enable Anchor Deposit and introduce mAsset swaps)
- consider removing the Instant Burn option so that liquidators have to take on beta risk. The largest problem revealed in the liquidation system design was that arbitrators were all bidding 30% for liquidation, so they were incentivized to sell LUNA at a discount. $300 million in LUNA was market sold. Making liquidators bid for collateral will tighten the discount.
- continue collecting data. The community asked whether the deposit interest rate on Anchor should be decreased, so that the Anchor reserve is larger and/or lasts longer. But, as Do responded, yield is set by governance. Do thinks any proposal should be informed by trend lines based on utilization over time and advises more data collection.
- add bUSD, USDT, USDC to Anchor Protocol, within a estimated time of a month.
- add Lido Finance’s stETH and stSOL soon after. Collateral can be added for multiple assets to a single position, meaning a single loan can be taken out against multiple kinds of collateral.
General Terraform Labs response to the Black Swan
Overall, the team pointed out that the protocol worked much better than they had expected in an event like this.
It’s worth repeating that this was the worst crypto crash, by numerous metrics, in recent memory. Across the board, sell-offs, liquidations, and congestion were hitting virtually all assets and protocols. Conditions were changing rapidly, and it was not always clear what kinds of actions would be beneficial.
Still, the team could have been better prepared, and Terraform Labs commits to formulate and release responses to events like this more quickly in the future.
UST Re-peg and Proposal 90
Despite the chaos, UST’s peg rebounded. It is currently trading nearly at the exact on-chain swap spread. Since only 20 million UST can be minted a day under current parameters, this spread sometimes increased dramatically, and UST was occasionally trading at rates like $0.92.
The core issue with the UST peg is this swap spread, which diverges from zero. Proposal 90 will significantly increase minting and redeem limits, up to more then 100 million, so spreads will be more resilient, and when nudged will recover much more quickly.
It should only take a few days for the peg to be fully restored. It’s already very close to $1, and in any case, Proposal 90 is passing currently and ends in 12 days, increasing stability of the peg.
Side note: UST listing on more exchanges is coming, but no more public comment can be made on this, for obvious reasons.
LUNA Price Action: is there Systemic Risk?
LUNA’s price movement was a result of UST dynamics, liquidity limitations, the Anchor premium parameters mentioned above, and market panic.
Total LUNA sells of $300–350 million came from Anchor liquidations, and UST redemption was low, only in the tens of millions. The drop had multiple factors, creating a feedback loop into each other. The trigger was macro downtrends, as we saw across the markets, but liquidations sustained the spiral.
Spending time focused on LUNA’s current price makes it hard to focus on work, Do said. He is a holder, not a trader. And Terraform Labs is not a buyer of last resort on LUNA, as that would be tantamount to market manipulation. Anyone concerned about LUNA’s price should hedge appropriately.
The AMA included some more general questions not related to the crash.
Chai is a payments application that uses Terra stablecoins on the backend. Chai progress in Vietnam was mentioned last month, but international expansion is slow due to COVID19.
Terraform Labs had its beginning as an enabler for Chai. Payments work is continuing actively, with more cards being issued and a couple of more countries being added this year.
However, Do’s primary focus is larger companies working with Anchor Protocol.
Should governance changes have shorter timeframes?
Do doesn’t think so. The base layer takes 2 weeks to pass or fail governance proposals, so that it is more stable, less prone to many sudden changes.
Mirror and Anchor proposals take 1 week. Anything shorter would, Do argues, disenfranchise users who are unable to check these protocols very often.
This crash was a difficult time for many people. Thanks to all those on the team and off it who supported the community during these events. Fundamentals continue to improve, lessons to be learned, and improvements to be made. “We’ll make it all back,” Do said towards the end of the AMA.
More open AMA sessions will follow every month or two.
Thanks to gorgeous danton, Perfect Legato, and other community members who helped a great deal with this post.