Dynamic Anchor Earn Rate

Could this dynamic responsiveness cause a death spiral?
Lower rates causes lower demand > Lower demand causes lower LUNA price > Lower price causes further reserve depletion (denominated in LUNA).

Would a better way to improve the reserves be to increase the cost of borrowing, rather than the reducing the compensation for lending? Or a combination of both?

Luna is now the #6 crypto by marketcap. UST #14. We’re big enough now. It’s time to mature a bit and realize 20% on stables is excessive. Cutting the rate in half wouldn’t lead to significant flight of capital. Brand is established, IMHO. Still, increasing the borrowing costs along with reducing the savings Earn yield makes most sense towards competitive sustainability.

This solution is unworkable, IMO. There will always be a way to create a synthetic asset and additional leverage. Someone will always try to take maximum advantage of a massive UST faucet. Appropriate design execution is needed to set the incentives such that we don’t need to rely on good-faith of users. The most obvious solution here is reduction in Earn incentives and increase in borrowing costs. Terra has such a massive industry lead we can now afford to be less generous with the advertising costs. Congratulations to us for reaching this next milestone!

1 Like

YR needs to be addressed since at current run rate it has less than 100 days left. However, making it dynamic without fixing the income side means that Anchor earn rate will drop 1.5% monthly towards possibly less than 10%. Current unsubsidized rate is approx 6%.

Anchor makes up 11.2B of total 15.5B outstanding UST. Anchor is the Terra linchpin. Dropping earn rate will cause exodus from the Terra ecosystem and trigger the death spiral. It doesnt matter that LFG has $3B. Money will still leave UST.

As Do tweeted less than 2 months ago: “Anchor is still in the growth phase, and maintaining the most attractive yield in DeFi stable will strengthen that growth & build up moats.”

There will be a time to bring down the Anchor earn rate to something sustainable. That time is not now.

Instead, what we should do is for TFL to double down and drop another $1B in the YR. In the mean time, work on means of increasing the income side by adding more collateral types and drive more borrowing, steps that are more aligned with the ethos of Anchor and the moats mentioned by Do.

7 Likes

@bitn8 whats the process/timeline for implementation if this proposal is passed? There will be time to test/audit the code? thx

1 Like

I think that has been everyone’s expectation. Even the LFG council members were suggesting it, but seems there’s been a change of direction somewhere.

2 Likes

I don’t understand why people are voting for this horrible proposition. A dynamic deposit rate is the worst thing that could happen to Anchor. It’s against the principle and original idea of the protocol. The ONE big competitive advantage that Anchor holds over all other yield farms is that it offers a STABLE rate - even during a bear-market. Dynamic Rates exist everywhere! Make it 15% or 10%, whatever. Admit that the experiment of 20% stable rate went wrong and adjust it. Or think of other ways to fix it.
Giving up on the initial idea during such a long sideways/bear-phase is like selling your bags at an ATL.

Maybe we should start to solve the problem by first asking:

  1. Why was the original deposit rate set at 20%?
  2. What was the original plan?
  3. What went wrong?
    @ryanology045

The big fallacy is that the earn rate will be cyclic to bull and bear markets. Bull market → rate goes up (lagging), Bear market → rate goes down (lagging). But the core strength of Anchor is exactly its anti-cyclical nature.

4 Likes

posted yesterday, but it was deleted as spam :man_shrugging: - so reposting

2 Likes

Exactly, the Terralytics idea is fantastic, and could even negate the need for something like Prop 20

It’s really scary to me that Prop 20 is going to pass, without even any changes from this thread, and the reason is because Quorum is so low (10%) and some people with big Twitter followings voiced their approval for it. I haven’t seen any big Twitter mentions of the Kash proposal:

I also think there are a few other issues with ANC governance that should be changed to encourage more engagement, though I can’t even put them up for Proposal because of the issues themselves and my wallet size

EDIT: Less concerned after reading the conversation between both Proposal creators

3 Likes

So I did some math… this isn’t going to work.

  • Average daily deposit growth rate over the last 50 days has been +0.99% per day (TVL went from $8.54B to $16.98B);
  • Average daily protocol deficit (YR depletion) over the last 50 days has been -1.37%/day and is currently at -$4.2177M (deficit is increasing because yield from borrows is not growing anywhere near that rate);

With Prop 20 rates will most certainly be reduced to 15% (or lower) because the protocol is currently losing waaaaay more than the projected 1.5% per month APY cuts. Deposit demand is unlikely to slow down while interest rates are >= 15%. At that rate they are still some of the best “stable” yields in DeFi.

Once the threshold of 15% is crossed, it will be interesting to see how the market reacts. The schedule for the APY reduction is:

  • April - 18%
  • May - 16.5%
  • June - 15% (will be interest to see how market reacts here and below, will keep decreasing if borrow demand does not match deposit demand, I think it’s unlikely we see borrow demand match with levered deposits)
  • July - 13.5%
  • August - 12%
  • September - 10.5%
  • October - 9% (around here is where I think the wheels come off because <10% you can get as good or better APY with the same or less risk. If we see mass redemption of UST to something else, UST will depeg, even with BTC backstopping.)

Without either:

  1. Another significant YR back stop ($1B+, keeps getting bigger);
  2. Immediate reduction in the levered deposits; or
  3. significant borrow demand increase or deposit drop (very unlikely)

The YR will run out before we even hit the scheduled 15% rate.

This chart is showing ~96 days left of YR but it’s only taking the total YR divided by the current deficit. With a -1.37%/per day deficit trend the YR will run out well before that.

If we keep on the same deposit growth trend, by June there would be ~$41.21B UST in Anchor. Since Anchor accounts for ~72.5% of UST in circulation, we’d have a market cap of ~$56.84B UST. Even with $10B in BTC, unless there is significant utility/demand for UST by this time, UST will only be ~17.5% backed. With a massive redemption as demand for UST dries up you could rationally expect UST to potentially depeg to as low as that value (heh, at least it’s not $0?).

I don’t think UST would actually depeg that hard but the main issue here is this Prop 20 doesn’t solve anything, introduces more complexity, and the math just doesn’t work.

In reality we’d better have another backstop or the YR is going to $0 quick with much more mass redemption risk than the last time just two months ago. We just keep kicking the can down the road and it becomes riskier and riskier.

If we assume that 50% of UST deposits on Anchor (40% of total UST market cap) leave when yields dip then even a ~$20B dump of UST is still gonna be painful.

:warning: This is needs a rethink! The biggest problem is levered deposits. The second biggest problem is borrow demand. :warning:

At this point it would be better to do a hard drop to a fixed 10-15% APY and just have LFG backstop it again when the YR runs low. This would likely bring in a short term shock but is much less risky in the long term and retains a fixed rate that is still market competitive. This kind of APY drop would actually makes a meaningful dent to extending the YR now so that future subsidies aren’t so expensive.

Then the focus needs to be on how to generate massive borrow demand and non-Anchor UST demand.

@ryanology045 @bitn8 happy to DM to discuss. :slightly_smiling_face:

If this Prop goes through it will make more sense to just join the degens until the yields are no longer attractive and then move on. I think there is huge risk here. :disappointed:

6 Likes

Interesting stuff, but, reading that, I think Anchor has way more issues than what we are thinking.

Using your calculation, if you were able to remove the degenbox, it would only extend the yield reserve funds for less than 10 days.

Close to 7B have been deposited since February, where did they come from? All Terra protocols may be depositing their funds to get yields, people waiting in Kujira, people coming from other chains?

That’s too much money coming in to the Earn side (even excluding leveraged deposits) for this to be sustainable.

Would be nice to get more insight on where are the funds coming from.

1 Like

Agreed. I haven’t looked into that. Could probably be done reasonably well with some on chain analysis but determining the leverage ratio vs. organic deposits would take some time.

Frankly, I think this is a case of growing too much, too quickly. Probably should have been some gating put in the place to the deposit side to ensure that deposit growth stayed closer to borrowing demand. Increasing the cap every time new collateral was available.

Right now the imbalance is so far out of whack because of the subsidies that I have a hard time seeing how this recovers without some unwinding, an epic macro crypto bull run or continual subsidies.

But at some point you gotta pay the piper…

4 Likes

Hey @bitn8, after this passes, does anything else need to happen or is this going to production immediately?

Has the code been audited?

2 Likes

The best bang for the buck will be releasing saUST (locked/staked aUST) and drop the native aUST yield to 10%.

This will immediately buy time for the Anchor team to rollout cross chain borrowing and cut the leveraged yield outflow significantly. It will also take a while for users to migrate over.

Right now we have no idea how much leverage is being looped into Earn (via aUST as collateral), or no way to investigate it. This proposal is the only viable solution.

5 Likes

Degenbox is the only known large external supplier of leveraged UST (about $750m). We have no clue how much is coming from other sources, but we do know for certain it’s going to escalate rapidly as Earn goes cross chain.

We can assume that Mirror is the largest entity due to leveraged short farm/ delta neutral strategies. It’s probably time to put a nail into that one as well. If Mirror cannot survive with aUST at 10% then they have far greater problems to solve.

When deposit APY drops, leverage demand is going to go up until eventually the entire thing is packed full of leverage. This is where we’re heading.

1 Like

Agreed. We are now too deep in the hole. May need to do 5% for aUST and 3X - 15% - for saUST if Anchor is to be saved. Otherwise, even if this was implemented today (realistically will take months to code, test and audit before release), too little time to make a difference. Now it’s time for drastic measures, if by next year Anchor is still to be around.

2 Likes

Hi @ekryski, I have made a dashboard with this exact purpose to model both linear scenario and an accelerating scenario.

From my calculations, the 21-day-avg depletion rate is around 3.4m and it’s increasing at about 2% a day. With this in mind, we have less than 2 months.

3 Likes

Yep, and the most perverse part is that dropping the yield might actually have an adverse impact.

As yields drop, investors will be driven into the arms of lending platforms like Yeti Finance who will happily print their shitty CDP stablecoin in exchange for aUST as collateral.

3 Likes

I absolutely agree with this - by my calculations the current sustainable apy is ~8%. The YR is going to be dried up long before it reaches that value as per this proposal.

2 Likes

There is a mechanism for what @mariano247 proposed–we do it all the the time when providing bAssets as collateral. Collateralized bAssets cannot be traded/wrapped/transferred/etc (if they can please tell me how!).

The same thing could be done with aUST. Keep the token, but provide different yields in Earn: one (lower) yield for aUST as we know it now, and another (higher) yield for aUST that is in effect “collateralized” the same way that can be done with bAssets for Borrow (obviously that’s the wrong term…maybe “cemented” as it is locked/cemented in a smart contract like collateralized bAssets are).

2 Likes