AIP - 45: Launch an Angle Gold Stablecoin

Angle Improvement Proposals
1

Hello,

This proposal aims to deploy a new decentralized Angle stablecoin pegged to one ounce of gold.

More specifically, we are aiming to do this while decreasing the potential role and impact governance can have on the stablecoin. Angle’s gold stablecoin will be governance minimized, close to how Liquity’s LUSD works.

On another note, Jim Cramer recently stated that investors should “stay away from crypto and look into gold”. As we all know informed investors should always trade against Jim, Angle Gold stablecoin vaults gives people a way to counter trade his announcement. :point_down:

Context

Angle started in November 2021 as a decentralized stablecoins protocol by launching agEUR. It quickly became the largest decentralized euro stablecoin, and the most traded EUR stablecoin altogether. The protocol was designed from scratch to be able to launch different stablecoins.

After careful considerations, we think Angle is now ready to welcome a gold pegged stablecoin alongside agEUR.

Argument #1: The gold market is a great opportunity, both on-chain and off-chain

PAXG & XAUt, the two biggest gold stablecoins, represent a close to $1B market for gold on-chain. Off-chain, gold is a ~$12T market. These two numbers represent both a great opportunity for gold stablecoins, and a acknowledgement that there is some demand for it on-chain.

As a comparison, the total euro supply (M2) was around 15T€ in November 2022 (source: ecb) for an on-chain euro market of $250M (source: Euro-stablecoins dashboard by SebVentures).

Argument #2: There is no decentralized gold stablecoin yet

One of the most prominent argument for an Angle Gold stablecoin might simply be that there is no decentralized gold stablecoin on-chain yet. This offers an interesting opportunity for Angle to lead DeFi forward.

On another front, this is also a great argument to push Angle gold even further in terms of decentralization. This means updating the current agToken and VaultManager implementation to remove multiple setters that are usually under governance supervision. This would make of Angle Gold stablecoin even more permissionless than agEUR, with no or very little governance intervention possible.

Argument #3: It creates a great opportunity to expand Angle to something bigger than a Euro-only protocol

Looking beyond this proposal, the addition of a new stablecoin to Angle allows the protocol to move from the “Euro decentralized stablecoin protocol” corner to the “DeFi products and infrastructure provider” one.

This is in line with the vision laid out for the protocol in our Beyond 2022 article: developing new DeFi products and infrastructure that provide better services and reach more people.

From a marketing standpoint, a truly decentralized gold stablecoin can therefore be quite positive for the protocol.

Design

The first debt-based gold stablecoin

Angle has two ways of issuing stablecoins: the original Core module, and the more recent Borrowing module.

Though the Core module brings a lot of value, it relies on a complex equilibrium to be sustainable. On the other hand, the Borrowing module has a different value proposition (stablecoins are minted as debt collateralized by crypto deposits), but requires much less maintenance.

We think it would be more efficient for the Angle Protocol to launch a gold stablecoin with the Borrowing module, that can be minted against permissionless assets, like ETH or LUSD. Like for agEUR vaults, people would deposit crypto assets to borrow up to a certain quantity of Angle’s gold stablecoin.

Angle Borrowing module contracts have been audited, and nothing apart from changing setters would need to be changed to the existing infrastructure for a gold stablecoin. As such with the already existing design, launch cost would be little to none.

The first permissionless gold stablecoin on-chain

Currently, the main gold stablecoins are PAX Gold and XAUt, two centralized tokens from Paxos and Tether respectively. Because of their clear centralization, the two on-chain versions are far from being permissionless. Paxos implemented a fee on PAXG transfers, and both can blacklist addresses to limit who has the right to hold these tokens.

What we propose is to make of Angle’s gold stablecoin the first permissionless version of gold on-chain. In that regard, we are planning to propose a slightly revised version of the agToken and VaultManagers implementations to reduce the impact a wrongful governance could have on the stablecoin.

While there would still be some parameters that can be updated by governance, like the debt ceiling or the possibility to add new collateral assets, most of the parameters would be immutable, and contracts would not be upgradeable. In particular, it would be impossible for governance to add new minters that are not specific VaultManager contracts, and the governance will not be able to engage into AMOs with Angle gold stablecoin or mint it through other means than borrowing. Governance will not be able to pause contracts or to add repayment fees.

Collaterals

Making Angle’s gold stablecoin permissionless also means only accepting assets that fit this criteria as collateral. As such, we propose launching the stablecoin with wETH and LUSD as collateral to mint the token. Though other assets can technically be added, the plan is not to add as many as possible, but rather to be restrictive and only keep safe permissionless assets to stick to the promise of building censorship resistant gold on-chain.

We propose to use Chainlink oracles to estimate the gold value of the collateral assets.

Growth

In terms of growth, the main use cases for Angle’s gold stablecoin should be for holding gold on-chain in a permissionless way, and to trade gold against other crypto assets. Beyond that, borrowing the gold stablecoin from Angle could be used by traders to short Gold against the vault collateral.

The first step for growing it will be to create a Curve pool with PAXG and push to have a gauge, and a Uniswap V3 pool with agEUR to have access to all of agEUR’s liquidity. Once the stablecoin is launched, we will launch a vote to add an Angle gauge and send ANGLE incentives to the pool.

With time, PAXG and XAUt holders looking for a more trustless alternative should progressively move towards Angle’s gold decentralized stablecoin.

Risks

The main risk for the protocol is that it accumulates bad debt due to a failure in liquidations of vaults having issued the gold stablecoin. On our side, we will adapt our keeper infrastructure so that it is ready to liquidate these vaults and repay gold debt like it does for agEUR.

Other than that, there is the reputation risk that the stablecoin doesn’t maintain its peg. Though this would be negative, there would be no direct consequence for agEUR or other working parts of the protocol itself outside damaging the reputation.

As always, the goal of this post is to foster discussions and hear everyone’s opinion about this. Please comment the post and share any thoughts or feedback!
Next steps would be to have this proposal voted to officially launch the efforts for an Angle Gold stablecoin. This would need to be followed by another governance vote to validate the implementation and the launch parameters.

9 Likes
2

Hello there, I like where this proposal is heading: to my knowledge, that would be the first gold-pegged CDP stablecoin on the mainnet.

I also like the focus on making sure the whole system is as permissionless as possible and the governance-minimized approach. Going with ETH + LUSD as sole collaterals is an apparent great fit in that perspective.

In terms of DeFi strategies, it further broadens the scope of possibilities offered by Angle, such as shorting gold using USD or ETH collateral.

My main concern would be concerning the liquidity of the agGOLD (or whatever its name is) - as the markets for XAUT (Tether), for instance, are not very liquid nor active, despite a sizeable TVL: not even 1M volume today for > $460M worth of XAUT circulating.

image
image1628×1362 315 KB

Is there any idea as to what the liquidity structure would look like? Having pairs against agEUR of course makes sense, but you will likely need more. Pairing against other Gold-stable? Or trying to grow an agGOLD/USDC pair, for instance, with the IL risks implied? Since liquidity structuring has become a bit of my specialty of late, I’d be happy to lend a hand on this topic if desired.

5 Likes
3

This is such an interesting turn and it really demonstrates the platform value of the ag protocol beyond agEUR. Also, if you resist the temptation to onboard tokenized gold in custody and stick to ETH and LUSD you will be making quite a committed play on decentralized gold-tracking tokens. All in all a super interesting framework for a product that has broader potential than just gold.

Do you have a 12 or 24mo expected schedule of liquidity expenses?

2 Likes
4

I might be missing a key technical feature–why wouldn’t you be able to have AMOs in a governance-minimized contract?

1 Like
5

Interesting thoughts, thanks for sharing it here!

Here is what I personally have in mind for the liquidity structure:

  • agEUR/agGOLD
  • agGOLD/PAXG → XAUt is not the biggest Gold stablecoin, PAXG is the one that is truly leading in volume.

I don’t think that an agGOLD/USDC pair would be helpful if we already have agEUR and if agEUR is used as a connector token in other venues.

The way I see it is that when launching a stablecoin on a chain, you need one pair with another asset that connects (so agEUR here) and one pair with a pegged asset to help support peg (so PAXG here).

Happy to explore other ideas, because I’m maybe thinking only one step away

4 Likes
6

Ok we should have probably put this a bit differently.
Implementation idea is to force all the contracts with a minting right on agGOLD to have the same byte code, so we could probably make some AMOs, but they’ll need to be framed differently than the current implementation.

2 Likes
7

No, this is something that could be worked on.

Thoughts are: if the protocol has an agEUR-agGOLD gauge, it won’t induce any extra expenses for the protocol, just a re-allocation of it.

Several expenses that could be planned though:

  • Curve incentives for an agGOLD / whatever other gold stablecoin pool → could also be a re-allocation from what is currently used for agEUR-EUROC or an increased budget
  • We’re considering to make another proposal once deployment is done to take USDC from the protocol’s surplus swap to LUSD and borrow agGOLD to then seed a pool with agEUR-agGOLD, and so this would be an expense in the sense of an opportunity cost from using the surplus.

But you’re right and we should probably think further before deploying incentives.

1 Like
9

Mhn, I have mixed feelings where the only source of “outside liquidity” (liquidity not against other gold stables) is through agEUR. It means you are essentially capping this agGOLD liquidity to the one of agEUR: what if this takes off more than agEUR itself?

Also, this is an excellent path for those already holding agEUR/other euro stable, but when you put yourself in the perspective of an ETH/USD-stable holder, the path to purchasing/selling agGOLD starts looking a bit painful:

Swap ETH->agEUR (gas fees, slippage, swapping fees)
Swap agEUR-> agGOLD (gas fees, slippage, swapping fees)

Or swap USDC-> agEUR → agGOLD

As a European, it pains me to say this, but most DeFians’ reference is the dollar, and USD stables are the most commonly held. I’d strongly advise catering an optimized swapping path for USD-stable holders. It could even be an agGOLD/3crv Curve pool with some customization work (like what RAI did).

But of course, USDC/DAI/USDT pairing clashes with the wish for a more resilient and trustless system, so an ETH pairing could also be explored. Here, you’d be in a situation highly similar to ETH-LUSD, a situation I know very well. Several options are possible here, but I’d recommend mixing two pool models for best results:

  • A wide range ETH/agGOLD, even maybe UNIv2 LP to make sure a baseline of liquidity is always available
  • A more price-focused UNIv3 LP to ensure sizeable liquidity at -15% +15% around market price.

This is the ideal now, but the DEX game is about to be completely shaken up this year, so depending on the release schedule, you might have much better options available.

1 Like
10

Indeed! I see your point, to start with at least a no brainer to do agEUR-agGOLD, but more long term yes you’re right, might make sense to pair with ETH or LUSD.

That being said, regardless of the pool incentivized, I would not focus on a specific range for the pool. Idea would be for me to use a generalization of this mechanism that guarantees that LPs can provide liquidity in the range they want while still being incentivized.

As for the release schedule, depending on the vote to launch or not agGOLD, I think it can go pretty fast, like my expectation would be to have this ready in 1 month.

1 Like
11

Lots of constructive idea!

I agree that agEUR should’t be a limitation to agGOLD, as both products should be independent. But if one could help the other (while not jeopardize it) we should go for it. We should monitor whether most are going through USDC → agEUR → agGOLD routes over XXX → agEUR → agGOLD or agEUR → agGOLD, and adjust incentives accordingly.

I agree with @sogipec’s remarks on incentivization, and would like to add that the mechanism is designed for concentrated liquidity, not just for UniswapV3. Even though DEXs may evolve, current Uniswap contributors appear satisfied with their CL, so future developments may not abandon it. I believe other competitors will also use CL, so we should be good on this.

1 Like
12

If you built in some way of doing AMOs you could essentially have agEUR / agGOLD pairs for free, capped at the size of the protocol surplus I guess. No sense in overpaying for what you already have.

13

Gold is a very interesting asset class as it is so far the only one hitting product market fit beyond crypto-currencies and stablecoins. With >$500M market cap, 26k holders and 700k transfers, PAXG is the leader. Nevertheless, Paxos make very little effort, if any, to promote it.

Gold sits between the fiat stablecoins in term of relative stability and the crypto-currencies in term of hard money. Most likely an interesting niche, if not more.

We are obviously here to see ETH becoming the base money in the future :smile: but until then, gold is most likely better than fiat currencies, $ mainly, on which DeFi is being built on.

Therefore, extremely supportive of this launch.

The question is more on whether to support a immutable design (ungoverned) or a more flexible one (governed and extendable). I will call the former the purist way and the second one the financial way.

The purist case business case is based on the expectation that some people will prefer a synthetic version of gold versus a physical one. Paxos custody is indeed a counterparty risk that is removed by having it generated against ETH and LUSD collateral (and let assume that ETH and LUSD are pristine collateral, which might not be the case depending on the Oracle solution and I’m not that convinced by Liquity one).

While PAXG is having transaction fees, those are quite muted for an investor. On the other way, minting agGOLD by borrowing will have some interest rate and will have some peg deviation risk. On the peg I wrote quite extensively on the DAI design issue here. Look also no further than LUSD to see the same problem. Look also at LQTY price action to see how there is no sustainable business model, only dilution to try to keep an fictious market cap.

The purist market is loud, but almost insignificant. No one wants to pay for decentralization as decentralization max realized. Additionally the current audience of Angle is not the purists. It would blur the brand to have no different (opposed) narratives. I would rather launch dGOLD (decentralized Gold) and avoid branding confusion.

The financial way business case is to leverage the ability to mint (AMO) to enable more “gold tokens” liquidity as a value proposition. As @adcv mentioned Angle could provide agEUR/agGOLD LP for no funding cost. Same would apply to provide agGOLD credit lines to lending protocols.

Any agGOLD used as collateral or sitting idly in a LP has no cost of funding for Angle, which is not something possible for PAXG. That’s the business model.

agGOLD could also provide elasticity and liquidity in the provision of tokenized gold being a credit system based on it just like old bank were under the gold standard

The peg issue wouldn’t go away and would probably require a PSM solution based on gold-backed stablecoins. But contrary to agEUR we could have a wider mint/burn fees (1%?) to avoid ending up with too much PAXG on the books. Governance should be more active and ensure that no exposure to counterparties will be too big. The aim here is that agGOLD will be the key token for a gold based financial system.

It is my view that for a DeFi gold stablecoin becoming mainstream it should be governed and be flexible to have a value proposition. Much innovation will be needed along the way which isn’t working with immutable contracts.

3 Likes
14

Regarding the tradeoffs for decentralized collateral, a few more thots–

  1. why LUSD and ETH vs just ETH?
    • as LUSD is essentially a net long ETH position, you may as well just do only ETH assuming the main constraint to solve for is decentralized collateral
  2. why spot ETH and why not stETH?
    • outside of considerations for decentralization, relative to stETH spot ETH has a negative cost of carry (albeit a positive convenience yield before withdrawals)
    • => Assuming withdrawals are activated and work, using spot ETH would create a permanent upwards bias on the price of agGOLD indexed to the stETH APR
    • The net impact of which could be to create a giant reserve of PAXG if you ever open a PSM, or a highly unbalanced LP against it if you don’t, neither of which are attractive features of decentralized gold
15

Thanks for your thoughts here!
So your main arguments for the more flexible version of a gold stablecoin is that:

  • the protocol would be able to provide agEUR/agGOLD liquidity for almost nothing (hence getting a competitive advantage with respect to PAXG).
  • we could also be freer when it comes to having a price stability module based on PAXG
  • more globally, you need mutable contracts if you want to adapt to the future evolutions of the market

For the first point, this is something that could be done indeed, with the big caveat that it could expose the protocol to a big risk of impermanent loss, so I’d be kind of wary and conservative if doing this.

For the second point, I’d say that if launching around such a model, people should be aware of the fact that the peg of the derived gold stablecoin will be a loose peg (and not as tight as what we’ve had for agEUR). Question is: is it really an issue?

For the last point, it all comes down to whether the narrative of a decentralized gold stablecoin will be powerful enough. I see an Angle gold stablecoin as an experiment thrown out there, and am genuinely wondering to what extent having mutable contracts could hinder the adoption. RAI is really small in market cap, but there’s still a lot of noise around it. There could be some externalities around Angle and agEUR by adopting this positioning for gold.

1 Like
16
  1. LUSD is to have a USD stablecoin (even though it has some correlations to ETH, it is still loosely worth 1 USD)
  2. I am not against stETH as a collateral personally, and this is sth that could be added, even with the proposed decentralized design. Like it’s important I believe to keep the possibility to add new assets even in a decentralized model
17

Thanks for all your comments !

About outside liquidity @TokenBrice

As I see it, if there is that much demand to trade a gold stablecoin against USD, which isn’t the case yet, then this will either push agEUR/USDC liquidity to grow, or an AngleGold/USD-stable pair to be created. In any case, holding is currently the main use case for Gold-stables, not trading, so I don’t think this is a very pressing issue.

Immutable or flexible design

I disagree on a few of your points here @Sebastien .

Decentralization cost

While decentralization / immutability bears a cost, which is what Ameen points out in the tweet you linked, there are some people ready to pay for it as LUSD’s 200M+ market cap suggests.

The case for decentralized gold

I think there is an even bigger use case & potential for a decentralized gold token than for decentralized USD stable. Gold has been used as a savings asset for a long time, and not as a transactional one as on-chain activity suggests. As a holder of such a token, I would expect as little counterparty risk as possible, whether it be from a centralized entity or a decentralized governance.

Issues with AMOs

Going the AMO way (token minting through governance approval) has very different risks than decentralization:

  1. Much bigger reliance on governance for peg stability
  2. Reliance on governance for “strategic choices” which could end up impacting negatively the stablecoin
  3. Significant development time, which is a cost in time for contributors, and money and opportunity cost for the protocol.

In my opinion, PSM solution is completely hinders the decentralization, as the protocol would then be fully reliant on the centralized versions of gold accepted to stabilize price.

Is the financial way really worth it compared to the purist way when it comes to a gold stablecoin?

While that might be true for fiat denominated stablecoins, I think that to get the most out of a gold stablecoin right now, we should go for the more decentralized & immutable path.

Ps: note that we are not advocating for a fully immutable token. Specifically, new collateral could be added as long as they rely on the same bytecode as the others as @sogipec points out.

1 Like
18

To add on @sogipec 's reply, I personally think stETH adds a counterparty risk that needs to be considered, and that is widely overlooked in DeFi currently.

However, it would be possible to add it!

2 Likes
19

Thanks a lot for your comments @Sebastien @adcv @TokenBrice, lots of food for thought in there

3 Likes
20

Seb has the right angle here and implementing his guidance will make for a successful agGold launch

otherwise this will just be another ageur / lusd instead of something revolutionary

1 Like
21

Thank you all for your thoughts on the topic! From everyone’s comments, it seems that there are 3 main different options at this point. Feel free to suggest if you see other points or pros/cons for each option, but that’s how I would summarize the problem.

1. Immutable contracts

This is the vision of the initial proposal, which consists in saying that we could make Angle Gold stablecoin truly decentralized and backed only by collateral assets that cannot be censored in any way.
In this setting, contracts would be immutable (non upgradeable) and there would be far less setters in the smart contracts.

Pros:

  • Absolutely no counterparty risk or exposure to another gold stablecoin
  • Real decentralized use case to push, like RAI or Liquity (even though in this case I’d still be possible to add new collateral assets), and the marketing impact + narrative it creates for the protocol is not to neglect
  • It’s a guarantee that it won’t take too much development time to protocol contributors once it’s launched since it’s going to be hard if not impossible to change the protocol

Cons:

  • Based solely on the proposed CDP model, Angle Gold stablecoin peg would be a soft peg with respect to gold. While statistical arbitrageurs may come to the protocol, there will be no direct incentive to bring the gold stablecoin back to peg (a bit like Liquity today which trades 2/3% over peg)
  • There will probably be more frictions immediately when launching the stablecoin to grow liquidity for it as there will be no means for the protocol to directly mint on Curve or on Uniswap for instance. And over the long term, protocol would not have the full flexibility to expand this Angle Gold stablecoin by taking advantage of one power of decentralized stablecoins: direct deposit modules to for instance bootstrap liquidity in lending protocols (like Aave)

2. Same flexibility as with agEUR

This means that contracts for Angle Gold stablecoin would be the same as for agEUR, in the sense that they would be upgradeable, and it’d be possible to add new minters to the contracts. This could make it easy for instance to do direct deposit modules with the protocol, like the protocol could for instance mint agGOLD in a Curve pool with PAXG whenever there are more PAXG than agGOLD hence stabilizing agGOLD price.

Pros:

  • Possibility to add a form of price stability module on Curve for the stablecoin hence improving the stablecoin’s peg
  • if it ever takes off, then governance would have the means to engage in AMOs, and the protocol would be flexible enough to take any action that maintains the flywheel for the product
  • It’d probably be more coherent in terms of the Angle Protocol brand if all stablecoins work roughly the same way, otherwise we’d have to be clear on agGOLD is more decentralized than agEUR which could create confusion

Cons:

  • Still a decentralized stablecoin governed by the Angle governance, but there is a significant risk in this setup that agGOLD ends up being backed by a majority of PAXG (through Curve deposit modules) which would increase the counterparty risk of the stablecoin
  • Mutability means there could be more incentives to spend time amending it/developing the protocol even if gold may not be the highest ROI thing the protocol should focus on

Note that time to develop smart contracts for setup 1 and 2 is the same, just like the time to do the first integrations for the stablecoin will be the same regardless of the setup that is used.

3. Do nothing

This is another option as well obviously. One of the main arguments that led us to interest ourselves to an Angle gold stablecoin is that it could be a great marketing push for the protocol, changing the protocol image, so that it’s not only seen as a € only stablecoin, but more generally as a global stablecoin protocol.
While launch efforts to put agGOLD to the market should be pretty limited (smart contract infrastructure already exists, app and analytics are designed to handle multiple stablecoins), and maintenance for the Borrowing module is none, it may take some time once it’s launched, time that will not be spent on agEUR or potentially higher value added products.
As such, one option still remains to do nothing, and keep focusing on the rest.

Pros:

  • No time spent and investment effort
  • No opportunity cost

Cons:

  • Not possible to change Angle image beyond a Euro-only stablecoin at this point
  • No way to take advantage and test Angle as a multi-stablecoin infrastructure

In essence, these options are tantamount to choosing a value proposition for an Angle Gold stablecoin and what it’ll stand for. Like is it a fully decentralized reserve currency that can only be borrowed from other assets? Or is it a more flexible DeFi tool that can potentially grow more easily in the first place? Or is it simply a distraction we should not worry about?

5 Likes