Anchor Earn vs Borrow. growing chasm

TFL will not step in because ANC is a decentralized protocol that the ideas and direction for solving this will come out of the community here with the Anchor’s team support. It can longer depend on TFL as we need to increase the strength of the community and the governance community.

Something that is in the works: adjusting the base earn rate based on market conditions and cash flow. Research is being done on a basic framework that the community can guide and agree on.

Also borrowing from the communities ideas of building a time-locked boost on this base rate is being explored.

More official write-ups summarizing things from the forum that can be further discussed are coming

1 Like

Sure, I’m just having a hard time to get the equation right. According the docs:

borrowRate = utilizationRatio * interestMultiplier + baseRate

If i want to calculate the current borrowRate, where do I get the interestMultiplier from? I can only find a hardcoded 0.42 in the docs, but I assume this one is also floating based on a function?

borrowRate = (2,067,024,812 / 5,258,196,431) * ??? + 2%

Thanks. I am surprised TFL would back out of supporting Anchor at this moment. Anchor is by far TFL’s biggest golden goose for UST demand and the gateway into the entire Terra ecosystem.

With over $11 Billion worth of TVL, is the community ready to tackle such issues without guidance? If Anchor was a PLC, it would be the 140th largest bank in the world by asset value. Surely such wide impacting decisions cannot be made by a small group of anons in a public forum?

I am looking forward to your proposals, but I fear we are going to spin our wheels reaching any sort of consensus and the clock is ticking on the yield reserve.

4 Likes

We have worked on an earn yield curve model that would stop the hemorrhaging. We suggest updating this model with the most appropriate inputs from Anchor side and putting it up to vote.

In my mind, this can be put to vote within a few days to prevent the need for capital injection by TFL.

6 Likes

One issue mentioned and being discussed in the post above is the issue with the redemption of aUST in a yield curve - that complexity could be hard to overcome… wherein aUST on one yield trajectory from date x is NOT fungible with aUST from another yield trajectory starting at time y.

Open to ideas on this or if a simpler tiered time lock would suffice given our model shows similar end states for earn with both models.

1 Like

Hey all, just joined because I believe there is another point to all of this that a lot of people dont seem to be seeing.

The Abracadabra UST leverage farm is a very convenient way for Terralabs to liquidate hundreds of millions of UST without hurting the peg a lot and they have a direct incentive to keep that one going for as long as possible.
How are they doing it?
Happy that you asked.

  1. Back in September Terralabs announced “Project Dawn” to help cover operating costs by releasing 3 million Luna a month from the genesis wallet, which I believe they have maxed out so far. Why they need 300m a month to cover operating costs, beats me.
  2. To cash out that humongous amount of money every few minutes bots are selling/burning Luna into UST (for example terra1zz2nf34fjkjygkg0kplkrr29ycxarmct6kafvj)
  3. Once enough UST is collected, TFL bridges the money over to ETH mainnet to the wallet 0xa046a8660e66d178ee07ec97c585eeb6aa18c26c
  4. Every time Daniele opens up the Abracadabra leverage farm by another 100m, what happens is that those users are selling those newly minted MIM for more UST which imbalances the MIM-UST curve pool.
  5. When the UST-MIM curve pool is sufficiently imbalanced another bot will sell as much as UST as possible into the curve pool until a perfect 50/50 ratio is reached again. In case above they are now sitting on 100m MIM in the wallet
  6. Then slowly the MIM is being sold into the MIM-3pool at a favourable peg and the USDT/DAI/USDC Stablecoins are being send to different exchanges to cash out (probably)

Over the last weeks Daniele opened up the leverage by another 500M bringing it to the currently 1.3bn it is sitting at and TFL was very fast to bridge over as much UST as possible and they have been trying to slowly sell all the MIM ever since as evidenced by the tx log from the wallet above.

Therefore TFL indirectly uses Anchor to cash out hundreds of millions of UST using MIM as a intermediary and given that Project Dawn keeps continuing releasing tokens I would guess they will continue to go on like this.
They have to keep this going for as long as possible and without knowing whats going on in the background I am sure all the parties involved here are aware of this.

Please let me know if I interpret something wrong, but the on-chain data is pretty clear in its purpose.

9 Likes

That makes a lot of sense. I was wondering how so much UST was finding itself over in the UST-MIM curve pool. No wonder TFL have been so passive and silent on this issue.

Sadly those being encouraged to take on the leverage are going to be the sacrificial lambs in the next flash crash.

2 Likes

TFL could also just deposit a bunch of bLUNA into the protocol. This is has better optics and would help the reserve. Not that this is for us to decide, but clearly TFL has large incentive to keep the party going.

At this point, we are seeing -$1m from the YR daily. I suggest (as we are all working out the solution) to limit earn deposits >90% collateral ratio. If Luna flash crashes before we have some reasonable limits, it may be catastrophic and result in TFL funding the YR again.

I would also suggest raising borrow APY higher than earn ASAP so that it’s not a free money loop.

2 Likes

To be honest, I don’t think we should do anything.

Let the yield reserve burn to zero until TFL come out and explain their master plan of trashing Anchor yields themselves…

2 Likes

I actually meant the “Distribution APR” of ANC tokens instead of “Borrow APR”.

Since we have a growing chasm between Deposits and Borrows and we want to stimulate borrowing, shouldn’t we increase the ANC “Distribution APR”?

What mechanism controls the “Distribution APR”? The docs say its done via poll but I haven’t seen any polls, and I see the Distribution rate dropping every day. The docs also state that there is a “target deposit rate”. What is that target?

I have a feeling these are noob questions. Sorry in advance.

1 Like

That’s actually covered here in the docs - see Borrower ANC incentives. Simply put your thinking is correct. ANC borrow emissions are calibrated based on the deposit rate threshold. It’s a multiplicative increase / multiplicative decrease feedback control algorithm. It has a modulated lag that increases or decreases over a week period.

Should we not have expected a boost at this point in that case? It’s been weeks since the yield started to fall but the APR on the borrow incentives has remained steady. Is there a variable(s) that I’m not taking into account that means it’s not ready to increase to provide more incentives on the borrow side?

The boost is not based on the provided yield, but rather on the utilization ratio of the deposits, and that’s actually held pretty steady near the 40% mark, since two days ago it finally started to drop and incentives are starting to kick in.

This seems to be a flaw in the design of the incentives imo, it’s being used solely to control deposit utilization (as to avoid all deposits being borrowed and depositors unable to withdraw), but doesn’t consider the yield being provided by that utilization.

I think pausing deposits past a certain threshold would be very interesting. It would create a secondary market for aUST tokens, where they would trade at a premium to underlying. This has two desirable properties

  1. New people could indeed come into anchor if they are willing to pay a premium on the secondary market for the aUST for a lower implied interest rate after taking into account the premium they paid. (Maybe that loop pool isn’t so silly after all!).

  2. It doesn’t affect current deposits so there is no capital flight.

1 Like

$15m USD from reserve depleted in 1 month. 4 months of runway left. And no response from TFL. Am I supposed be paranoid?

1 Like

No need to be paranoid in the short-term. Longer term, Anchor has to attract more borrowers

There have been some Anchor AMA’s on Twitter talking about multichain integration with bAssets and a tweet by Matt Cantieri regarding his plan. I personally feel better about things but it has certainly been a dramatic ride!

As a user of the platform I am not worried, but I am closely watching to see if we get a yield reserve top up or a rate cut.

If the rate is cut I will likely be moving my funds out. You can get 20% in Convex or Yearn.

It’s just business at the end of the day and that’s how everyone should look at it.

2 Likes

Does Yearn guarantee 20%? To my knowledge Yearn yield rate is all based on market conditions and is not predictable.
There is no yield reserve such as on Anchor.

Not sure about Convex

1 Like