Increase ANC value by whitelisting it as borrowing collateral

A pump is not the point - see the first sentence in my first post. ANC being more valuable → more people want to borrow (to farm the now valuable ANC) → more revenue for Anchor and easier to achieve solvency.

Could you explain in more practical terms why borrowing against ANC is a bad idea? I don’t get how the 51% scenario is relevant to ANC specifically being allowed as collat. If you want 90% LTV and low fees you can already do that if you vote using enough ANC in gov and then swap for bLuna.

The issue is a general one on Terra Dapps that do not lock staked governance tokens long enough to require an alignment of interests between voters and the future of the protocol.
Look at the most successful dapps out there: Curve locks voted capital and distribute staking rewards based on the length of lock, AAVE has a cold down period for staked AAVe.

I do retract my previous worry about governance and borrowing from the money market. Perhaps my worries can be fixed by a simple lock on voting or a cooldown.
However, the premise of anchor is to accept stake assets from blockchains, not governance tokens. this would not align.

Near Term Priority should be:

  • reduce the liquidation premium so people are so scared of borrowing. (the team is working on queued liquidation bids)
  • more assets so that not only Lunatics borrow.

Medium Term Priority:

  • Figure out a more competitive borrowRate vs Utilization curve to promote borrowing (will require adjusting DepositRate) however this is not important right now as ANC distribution makes borrowing be free.

Long Term Priority:

  • Now Anchor is a self sustaining business, how do we redistribute some of that to ANC holders.

@Kamil I think increasing the price of ANC to promote more borrowing (from distributions) would work but minimally compared to the above. For example if liquidations meant a 5% haircut as opposed to a 30% haircut, people would borrow a lot more.

1 Like

Would you personally use your ANC or purchase some ANC to use as collateral in the state it is in?

Yes I would, I work on the PC most of the time and when i’m not i’m checking my phone, it depends on the individuals risk management, the type of person they are. High risk = high reward etc.
I would like to be able to utilise my ANC rather than just have it staked. Maybe create some aANC like they have with UST (aUST) which enables you to have it staked (earning 9%) while at the same time putting it in borrow as collateral.
Just a thought.

Finally, some active dialogue! It seems like there is some consensus that 1 & 3 are good ideas and 2 might have merit (but also may do nothing for ANC price stability if implemented on it’s own). Can we just put the three to a vote individually to see what the entire community thinks?

Are there any reasons we shouldn’t begin to vote on these options (separately) immediately?

Thanks,
bM

From what I’m seeing there seems to be some that just don’t want to deviate from the original plan. Aka just relying on increasing borrowing and keep all mechanics the same.

But I’m ready to vote.

This is a nice recap of the potential improvements.

As for whitelisting ANC as collateral, I think a better option would be to whitelist Bonded ANC, or bANC. Since now there’s somewhat of a better framework for creating bAssets, creating bANC shouldn’t be too difficult.

Also adding in what has been in discussed within the Anchor team to address ANC’s value (arranged from easiest to implement to the hardest).

Short-term

  • Whitelist bETH as collateral → in final steps, increases cashflow via more borrowing demand.
  • Whitelist bANC as collateral → recaptures borrower ANC incentives back to the protocol, decreasing selling pressure.
  • Increase buyback ratio from 10% to 30% → doable if bETH allows for a net-positive cashflow

Mid/long-term

  • A flat fee of e.g. 0.5% when repaying borrowed UST → capture additional cashflow from active use of the protocol.
  • Instead of having gated yield for ANC stakers, offer lower yield to smart contract-based depositors (perhaps whitelist a few contracts to give them the full Anchor yield) and use the yield difference to buyback ANC & distribute as ANC staking rewards. → applies a protocol fee to third-party DeFi applications building on top of Anchor without hurting the yield for savings applications
  • Require ANC stake to efficiently participate in liquidations (the more ANC you stake, the more bids you can submit to the liquidations queue) → increase utility for ANC

Would love to hear how the awesome community thinks about the ideas above

3 Likes

Interesting ideas & good to hear that the Terra team is also looking at the problem!

  • bANC instead of ANC - sounds good. Let’s use whichever format makes it easiest for the protocol to capture and make use of staking rewards.
  • 0.5% repayment fee - my immediate impression was that this is not quite in the spirit of DeFi (lowers the flexibility, e.g. if someone wanted to borrow only for a few days for whatever reason). But tbh considering the borrower APRs are in double digits anyway, this might work as long as it’s clearly communicated in the UI.
  • Whitelisting a few smart contract-based depositors sounds like political trouble :thinking: . Are there any criteria you already have in mind for how to choose who gets whitelisted in a fair way?

All else sounds great!

I like these ideas. All have potential, generally as long as we acknowledge there’s a problem here and act QUICKLY to try to solve that problem - we’ll be fine. I think the flat repayment fee will not see support from borrowers… but do they even get to vote? Because if not, they’re not supporting the protocol anyways (and generally contributing to the downward pressure on ANC).

In the end you can’t make everyone happy, so the “spirit of DeFi” might have to take a backseat to the practicalities of the real world… like the law, like human nature to game incentives and take advantage of others, etc. 0.5% sounds quite high considering how often borrowers repay part of the loan (to avoid liquidation) so we might need to think up a more appropriate number but as a borrower, I do think the current fees are probably too low.

I like these ideas very much!

The only point I don’t quite agree is the flat fee of repaying borrowed UST. This fee would negatively affect users that control their LTV more actively, and lower total borrowed amounts in the protocol.

For example, I try to keep my LTV between 40%-45%, depending on market volatility. If I had to pay 0.5% fee to repay loans, I would keep my LTV at a lower level, as the marginal gain from keeping a higher LTV would be offset by the higher costs of repaying loans.

Another side effect of a flat fee is that they would discourage the development/usefulness of tools that help users to manage their LTV in anchor.

1 Like

I was thinking more on the suggestion to lower yield to smart contract-based depositors, and I think that it would be detrimental to the ecosystem in the long term by giving an advantage to centralized/custodial solutions vs decentralized applications.

For example, if we give lower yield to contract depositors, an exchange like Binance could offer higher ANC auto-compounding yields in their centralized platform than what a user could find in the Terra ecosystem dapps.

I agree with the above comments. Maybe instead of a flat 0.5% fee, the fee could be scaled depending on what portion of the loan is being repaid? That way it wouldn’t disproportionately penalize people who are paying small amounts back, while maybe you could justify a fee even bigger than 0.5%.

Hmm I think it’s already implied that 0.5% applies only to the portion being repaid rather than the entire borrowed sum every time. Worth explicitly confirming ahead of implementation.

The idea is to have the 0.5% fee applied on the amount being repaid, not the total amount borrowed

This might have the perverse effect of lowering the overall borrower’s LTV because people would want to reduce the number of times they have to repay in order to maintain a satisfactory LTV.

Why not just deciding to increase the interests paid on borrowed UST by 0.5% instead (as long as the Net APR is positive) ?

So you would higher the interest rate by 0.5% and buy back ANC with it and distribute ANC again?
That sounds like the same - but more complicated.

The fee applied to those who use the system more frequently might have a negative impact on the overall transaction volumes, right?

If people use it less, they might pay less fees for the overall network.

The ideas are awesome. Being able to borrow against bANC would give a really good reason to hold it. It’s the Anchor, right?

I use Borrow a lot and would certainly not use it if I was having to pay a fee every time I repay. Not a good idea.

Even so still not a fan. Better to reduce the pay to borrow % and its clearer to the borrower what they are getting than adding hidden fees or fees that you don’t calculate till later, leaving borrower with a bad experience.

While I think this is a great idea in creating more value for ANC, it might have the effect of making ANC vol much higher.

Based on the current ANC buyback mechanism where 10% of value flowing into the yield reserve is used for value accrual of ANC, the value of ANC should theoretically be heavily affected by whether yield reserve is increasing or decreasing (value of future cashflows into ANC increase = higher net present value/price). This is essentially what we’ve seen over the past few weeks.

If bANC is added as a collateral type and assuming that bANC ends up becoming a respectable proportion of the total collateral pool, this means that if the collateral pool decreases (for whatever reason) → cashflow into yield reserve falls → buyback of ANC falls → ANC value falls → liquidation of bANC → collateral pool falls & the cycle repeats.

Of course, the true relationship isn’t so straightforward as there are other factors such as diversification of collateral pool and investor sentiment.

The reverse of the above is also true which would amplify the pump in ANC price as the loop would be self-reinforcing. However, as we have seen with bLuna, the upside loop is weaker because people are less willing to raise their borrow rate vs. when prices are going down.

All things considered, while this does create a good use case for ANC (which I’m fully supportive of). However, it also introduces more volatility/cyclicality/beta to the ANC token price which I’m not sure is ideal for the design of the token.

I believe the aim is to have Anchor be counter-cyclical; having the prices of ANC crash hard due to this feedback loop would be the opposite of what is intended.

Repay fees are good at boosting the counter-cyclical nature of ANC, but I agree that its not great user experience.


On a separate note - is it possible to take a larger slice out of liquidations and direct it to the yield reserve?

The money market sets the execution fee address to its yield reserve, transferring 1% of the bid value to the yield reserve.

Is this yield reserve the same as the anchor yield reserve? We can probably be more aggressive with the split by making this 10% of the premium for example (if premium is 10%, yield reserve gets 1%, if premium is more aggressive at 30%, yield reserve gets 3%). This will also be less punitive for liquidators looking to execute at a <10% premium.

Keen to hear your thoughts! Sorry if this has already been raised in other parts of the forum…

Good points on trying to have counter-cyclical mechanisms for the anchor protocol/token.

I think increasing the % of liquidation going to anchor is a great idea. We have already seen that current liquidation incentives are way more than enough to make liquidators massively profitable. Taking a higher share of this cash flow source in market downturns sounds like a good idea.

1 Like