Anchor Oracle Price Discrepancy on December 9th

[narco78]

So we’re suggesting paying out risk takers ‘borrowers’ at the expense of ANC holders and LLP providers who will get hit with an impermanent loss on their LP positions due to a dumped ANC price?

This is simply robbing Peter to pay Paul. It makes zero sense and sets a horrendous precedent.

Every technical issue going forward is going to result on a run on the ANC token. Other options need to be discussed.

Finally, can anyone point to any documentation that suggests community held ANC tokens will be distributed / sold to make compensation payouts for technical issues? If not, they should not be used.

[XK8]

NFT (Warrior Badge)
I Like It!

Speaking as 1of239 Liquidated on 12.09, having my investment back would be nice, but your NFT idea would certainly qualify as CryptoEthics!
-regards

[wengzilla]

I wrote a Twitter thread, but copying to the forum for further discussion:

Here are four ways Anchor can compensate users for the liquidation bug (~$9M of impact). It should be noted that these ideas can be used in combination with each other (e.g., 50% / 50%). From most preferable to least preferable:

1. TFL intervenes, uses profits from past liquidations to compensate users

Before the liquidation queue mechanic, TFL operated a privileged (but “benevolent”) liquidation bot that was extremely profitable. These liquidations, some of which happened at a non-zero liquidation discount, were never meant to be used for profit but to stop another May death spiral.

Do’s original post on the matter:

Can TFL liquidation profits be used to fund the reimbursement? To me, it feels a little weird that TFL profited off liquidations (and potentially even profited off of this bug if they placed bids into Kujira).

Pros: Does not affect the health of Anchor or the ANC token. A $9M write-off is negligible for TFL.
Cons: Requires TFL intervention, but precedent exists. Back in July, TFL stocked up Anchor’s yield reserve with $70M from its Stability Reserve Fund.

2. TFL intervenes, uses a portion of the $4.5B worth of Ozone reserves.

Ozone is a yet-to-be-released insurance protocol. Funds in Ozone were meant to be used for “coverage of technical failure risks in the Terra DeFi ecosystem”.

Given that the oracle was a technical failure, usage of the funds here seems appropriate.

Commonwealth

Pros: Does not affect the health of Anchor or the ANC token. A $9M draw-down would be negligible to Ozone reserves.
Cons: Requires TFL intervention.

3. ANC eats the loss, uses a portion of community reserves / borrower incentives / Luna staking airdrops.

Anchor has 100M ANC tokens stored in its community fund. At current price of $3.50, 2.6M ANC tokens (~3% of the community fund) would need to be sold for $9M.

According to Alpac4 dashboard, the daily ANC volume can reach ~$21M. A $9M sale would have to be carefully managed. As ANC price goes down, Anchor may need to sell more reserves.

Pros: No TFL intervention required.
Cons: ANC token price negatively impacted.

4. ANC eats the loss, uses a portion of its $79M yield reserves.

The yield reserve is supposed to be used to stabilize depositor yields if borrower demand is not sufficient to pay out the fixed depositor rate. It has been on a steady upward trajectory.

Based on current deposits of $3.5B and a rate of 19.5%, Anchor currently pays ~$1.9M of interest per day. $9M would represent 4.5 days of interest payments.

Pros: No TFL intervention required.
Cons: ANC yield reserves weakened.

Would love @PFC and @mcantieri to opine on whether #1 and #2 are realistic options. Can we actually expect to get TFL involved in restitution?

[PFC]

If someone volunteers to do the art work, I can get the rest of it up.
any other badges you can think of?

I survived the May2021 crash badge, and all I got was this lousy NFT?
‘Solid Earner’ badge for having deposits since inception

?

[XK8]

Nice!
phase2:Healing
status:Activated.

I have too much art, feel free to pm me, let’s bounce some of it around.

[fig]

@PFC I applaud the transparency here, as well on Twitter.

While it is unfortunate the loss of funds, the response and steps taken by the team are important.

It seems as if 242 total users were affected. I am interested in making use of this mechanism:

@wengzilla

Luna staking airdrops.

I’d be curious to see how much of the repayment this covers - seems to be the least critical effect on the ecosystem. LUNA airdrops are a good reward but often benefit the wrong people.

I would advise pursuing a solution that does not involve TFL, as this is a privileged relationship. The protocol should operate as if TFL will not be a crutch in the future.

For those curious about the Anchor Yield Reserve, here is a visualization and previous bounty by Flipside:

Anchor Protocol Flipside Analysis

[Salt7]

AnchorProtocol is using “governance” to their advantage.
They know few people faced the erroneous liquidations. You admit this is an “erroneous liquidation” yet submit it for votes. I’m only hoping 10% quorum will be reached with a YES.

I’VE BEING “ROBBED” IN MY HOME. AND I’M AT THE POLLS VOTING WHETHER THE “THIEF” SHOULD GIVE IT BACK OR KEEP IT🤦🏾‍♂️

[paletas]

That’s a poor analogy, you’re frustrated I get it, but when you decide to invest into Web3 protocols you should know that you’re not a client, you’re part of the protocol. Want a better analogy?

You decide to open up a store with thousands of your closest friends, everyone buys the items they want to sell and leave it at the store, one day the lock on the door gets broken and nobody noticed, that night a thief comes in and steals some of the items. It’s not your items, or their items, it’s everyone’s items. Now, because you’re all part-owners you must decide how to handle the situation, should you ignore it? Should you compensate whoever spent money to buy the items? How will you compensate them?

That’s what we’re doing here, deciding on how to compensate our close friends who got affected by a problem that was no ones fault, but everyone’s responsability. If you can’t handle the responsability of being in charge of your decisions, perhaps this isn’t the place to be.

[cryptograk]

I think this is a key point here.

What is the role of insurance?

1. Did insurance options exist at the time of the incident? I believe so.
2. If the users had insurance, would the insurance have covered them? Again, I believe so.
3. So, if some users had chosen to pay for the insurance, then the insurance would have covered their losses. But we are going to have the “community” pay for the losses of those that chose not to pay for insurance?

This is tough, and I understand the difficulty and don’t know that there is a clean answer. I was not personally liquidated, so my perspective is different. Also, I did not have insurance, so were I liquidated, my perspective would likely also be different (although, I should probably look into it now…). That’s why I’m trying to be as fact based as possible and not be driven by personal experience.

Maybe a fair compromise in this growing ecosystem is as follows:

1. Reimburse those affected from a community pool, using a mix of assets as others have suggested in order to not create downward ANC price pressure for others.
2. Have the team address the oracle issue to prevent this from happening again (which I think they are).
3. Firmly take a stance that future technical issues like this will NOT be reimbursed and encourage everyone to be responsible and insure their positions.

[dragonish-undertow-v]

So just to reiterate what was said further up. Anchor Oracle Price Discrepancy on December 9th - #6 by codehans

This was not just an issue with Anchor, the terra system has a built-in oracle that generates prices for Luna based on validator votes (https://docs.terra.money/Reference/Terra-core/Module-specifications/spec-oracle.html#concepts)

Recently there has been a number of occasions that have resulted in validators votes not being successful and in this situation, the vote price for luna dropped from 66.41usd to 58.44usd then back to 66.28usd in less than a minute and a half. (see images below sourced from Terra validators discord)

I can only assume the contract anchor uses to calculate the price of luna is based on the open-source oracle feeder terra has available, and they too use the validators voted price to determine the current luna value in USD.

So in my opinion this vote is too early and more information is required to validate the best approach to somehow reimburse the affected individuals.

So what information are we missing, mainly how is the anchor luna price calculated?

If via the oracle built into terra, is it the responsibility of the terra system to cover the cost of reimbursing the affected individuals, or is the assumption that the terra system oracle is only for reference?

Which I don’t believe it’s as I believe its price is used to burn luna into ust (this I’m not 100% on as I haven’t managed to find information on it).

I hope this helps.

[jbbasics]

Glad to see movement towards reimbursing. Above all, confidence in the platform is top priority… you can definitely feel a drop in confidence from the premature liqs, and this will help restore that confidence.

Just realized the proposal that this link came from is to utilize the funds from ozone vice the recommended use of the community pool(a little misleading). I do think Anchor needs to be self sufficient, and yes it will likely result in some temporary downward pressure on the Anc price…but think this is the best route.

[Salt7]

“not the place to be” huh? For that you are right, I deserve better “place to be”. Fortune has shined on me, and $LUNA can be used as collateral elsewhere.
Now to your fractured analogy, not all the part-owners are obilgated to vote, yes? And the unaffected have significantly higher percentage.
Have you voted??? I can’t tell
Can we reach over 10% quorum? I hope so.

But what if we don’t, our close friends will be forever fearful to trust the store.

Incoming partners, will change their minds about bringing their things to the store.

This is not about 239 affected persons, it’s about @anchorprotocol and maintaining it’s good image.

I hope the proposal gets the YES votes and passes quorum; that’s all I need for the community. But my analogy still holds.

We are in this together

[bitn8]

Prop 110 isn’t viable or implementable. Ozone is now under Risk Harbor management, an independent protocol not in any way controlled by the Anchor community. The funding for Ozone was secured through the Luna Governance process, not Anchor governance. The funding is now under the management of Risk Harbor and is earmarked for issuing coverage on contract bugs to earn aUST deposits and provides no coverage on the borrow side (when it goes live).

As for compensating those who lost funds, @paletas has submitted a post detailing possible options that are actually viable:

@paletas

There’s two major ways to do it, imo:

1. Use the $ANC allocations from the community pool as suggested;
2. Use funds from the yield reserve, at least I believe that to be a viable option;

3. The option to do nothing

[Alf]

I am not in favor of this vote at this time. How does a vote on Anchor impact funds on Ozone? Is it even possible for anchor protocol to access ozone funds like this? In my opinion, there needs to be a series of votes with multiple options and the highest voted solution gets implemented. These votes should take place only after a full accounting of what happened, who was effected, and exactly how many funds were affected?

[codeNoodle]

Yea, it seems someone prematurely fired off the proposal, not sure why. Discussions are still ongoing on this forum post, and there are good reasons offered that declare this specific solution (Ozone funds) not the best route to take. Not only that, but we have yet to receive the full report on the situation.

[wengzilla]

The core problem, IMO, is that the Anchor team has not opined on the feasibility or preference towards any of the 4 options that I listed. If they do not help us understand what is actually on the table, it’s going to be hard for the community to propose an optimal solution.

Funding it entirely from the Anchor side (especially with ANC tokens) is damaging to the token, so it really should only be considered if it’s not possible to have TFL refund some of the earlier liquidations.

@ryanology045 and @mcantieri, can you please opine on your preferences from the options above?

[narco78]

@bitn8

3. The option to do nothing

I am in firmly the do nothing camp. Unless TFL or the community are willing to donate their own funds then I don’t see any other viable alternative.

Borrowers and savers on Anchor are risk takers using a decentralized platform that offers no guarantees. There needs to be a risk disclaimer on the website. The crypto market is also rife with flash crashes regardless of feed glitches so borrowing is clearly a higher risk strategy.

Selling millions of ANC tokens just hurts everyone else inside the platform (stakers and LP providers) and damages the value and integrity of the token. This idea/option needs to be taken off the table immediately.

[busterscruggs]

What is frustrating is that the team isn’t really engaging on this discussion or letting voices be heard (been a week). Like you said, the community needs guidance/discussions with the team in order to see what’s even feasible.

[narco78]

I agree, who is “The Team”? It seems we’re running without any leadership structure. I see the same thing over in Mirror Protocol where nobody can made a decision and we just go around in circles.

Can somebody with some clout / authority just pick this up and own the issue?

[somethingelse]

Interesting idea. So you are proposing we use the ANC airdrops for LUNA stakers to compensate the affected users. We can take this one step further in fact, we could use the ANC airdrops for bLUNA’s underlying staked LUNA to compensate the affected users, there is a massive amount of ANC currently sitting unclaimed in the Overseer contract.