[ARFC] Increase GHO Borrow Rate


Title: [ARFC] Increase GHO Borrow Rate

Author: Marc Zeller @marczeller - Aave Chan Initiative

Date: 2023-08-23


Summary

This ARFC proposes an increase in the borrow rate of GHO from the current rate of 1.5% to 2.5%. The aim is to address the peg deviation of GHO and ensure its growth and trustworthiness in the market.

Motivation

GHO has witnessed a rapid growth, reaching a circulating supply of 22M in just a few weeks since its launch. One of the primary reasons for this growth, especially in the current adverse market conditions, is the GHO borrow rate set below the market price of stablecoin borrows. This low rate, combined with the current lack of holding venues and the inability to use GHO as collateral, has made single-side liquidity provision in stableswaps, such as Balancer pools, one of the most sought-after strategies.

While deep liquidity and growth are desired as a go-to-market strategy for GHO, the peg deviation of GHO (currently around $0.975) is affecting its growth and trust in the market. A consensus short-term solution to this issue is to slightly raise the GHO borrow rate from 1.5% to 2.5%. This rate remains attractive for borrowers and is still below the market average rates. However, by reducing the discount, it is anticipated to have a positive effect.

This proposal is particularly relevant in the context of the potential sDAI onboarding integration. This integration is expected to add short-term sell pressure on GHO, especially with the current borrow rate, which could further harm the GHO peg.

The discount for stkAave holders remains unchanged at 30%.

Specification

  • Asset: GHO
  • Contract Address: 0x40D16FC0246aD3160Ccc09B8D0D3A2cD28aE6C2f
  • Current Borrow Rate: 1.5%
  • Proposed Borrow Rate: 2.5%
  • New discounted Borrow Rate: ~1.75%

Disclamer

The Aave Chan Initiative is not affiliated with or paid by any third parties to publish this TEMP CHECK.

Next Steps

  1. Gather community feedback on this ARFC.
  2. If consensus is reached, escalate this proposal to the ARFC Snapshot stage.
  3. If ARFC snapshot outcome is YAE, implement the proposed changes to the GHO borrow rate.

Copyright

Copyright and related rights waived via CC0.

9 Likes

Aho.
Wasn’t supposed that the peg would be restored with the introduction of the PSM? Shouldn’t we wait for that release before implementing more patches?

1 Like

Hi,

Low interest rate GHO is awesome. I think it needs time and more integrations.

I don’t like that you guys are almost doubling the interest rate so soon. Seems to be a bit of a shit move. Early users of GHO are the ones that will most likely get the most hurt by this.

As @ApuMallku says wasn’t PSM supposed to help with peg? Could integrations with partners also help peg?

3 Likes

This is likely months away. So i do understand this move.

Firstly, I delegate to The ACI and welcome the discussion on GHO and its early peg struggles and welcome Marc’s proposal to debate the proposed option.

However, I’m yet to be convinced raising interest rates maybe the best possible solution although concede it could be one of the quickest and easiest to implement. Like others have already commented would not the GHO Stability Module (GSM) be a more effective tool and would it be worth waiting for it to return to Governance at ARFC stage? Which, having passed temp check and progressing though code audits for approaching two weeks now surely can’t be too far away?

My reasoning is simple, the GSM would allow the use of other stablecoins as collateral. GHO minted via the GSM can be redeemed for the underlying collateral 1:1, creating an arbitrage opportunity for part of the tokens’ supply which currently doesn’t exist for GHO. This I feel would be a far more impactful tool than raising interest rates.

Yes, I recognise that that time maybe an issue with the GSM and still needs to both progress and pass various governance checks ARFC post (5 days), ARFC Snapshot (4days) AIP vote and execution (4 days) and that if this proposal proceeds to AIP and passes it maybe quicker but I still am yet to be convinced raising interest rates by 1% would have the desired effect.

Currently, (as I see it) GHO is minted and then is either used for LP or can be sold on the open market for an approx. 0.25% loss where it is leveraged elsewhere for potentially greater rewards. To me it appears that the opportunity cost of leveraging is currently exceeding the current discounted GHO repayments, hence more mints than repayments and I don’t think raising the interest rate by 1% will be necessarily change this mindset or be anywhere near as impactful as the GSM.

Additionally I also think as an alternative to both the GSM and raising of interest rates consideration should be given to letting Aave users use GHO as collateral when borrowing other stablecoins which would provide a quick and effective arbitrage opportunity and certainly be a very effective mechanism to help GHO maintain its peg. I don’t know why this isn’t currently an option for discussion.

Now, I’ll readily admit I’m no giga-brain DeFi genius, economic wizz, or dev guru so am happy to be corrected if I’ve either misunderstood or spoken complete twaddle but to me both the GSM and/or using GHO as collateral would be better mechanisms to pursue than raising interest rates and should be part of the discussion.

2 Likes

Hello @MrKris and firstly thanks for your trust delegating to the ACI, we’re honored to be your voice in the DAO.

GSM helps when GHO is overpeg, not underpeg. GSM allows to swap stablecoins for GHO at 1:1, but as gho is currently at 0.975/0.98$ on secondary markets, there’s zero benefit using the GSM today.

this was the wGHO plan, but @Gauntlet raised concerns about it and onboarding wGHO was abandoned in the short-term.

2 Likes

If GSM only helps when GHO is overpeg, why were many people (that are part of AAVE DAO) “screaming” that the peg will be restored as soon as the GSM is out there? The release of GHO without the GSM was a flaw, and now it seems that offering a low fixed borrow rate was also a design mistake. Early borrowers will get rekt if this proposal gets implemented.

No, they won’t. Likely, they sold the borrowed GHO (haven’t checked the addresses), hence why GHO is not pegged. If we increase the interest rate they will have to buy back and repay the debt (i.e., re-peg GHO). It’s a simple supply and demand. Both proposals (GSM and raising interest rates, which in turn should lower the supply) do have their merits. The only other tool left out of the discussion is having something like savings rate (to influence the demand for GHO).

Though I do believe that releasing the GSM would help faster, as it does actually work both ways (you can check stability modules for DAI or USDD or any other stable. not as @MarcZeller says). It opens the doors for arbitrageurs to buy/sell GHO on open market (at market) and via the GSM (at 1:1 no matter the market).

PS: This ofc assumes that there is something to swap the discounted GHO against in the GSM, but I believe that GSM would need to be either bootstrapped by the treasury or in some other way.

3 Likes

That is correct, currently, There’s no market reason to “fill” the GSM with other stables, so implementation now would not help underpeg situation. that was my main point.

The ACI is against spending stables in the treasury to fill the GSM as it’s an inefficient subsidy, if the DAO wants to support the GHO peg it can do so by directly buying GHO on the secondary market. when peg is restored the DAO will earn the discount defined by the purchase price.

This option is not something we’re keen to support with the ACI.

It might be a good experiment to buy GHO (with DAO’s funds) at a discount as soon as the PSM is live. We prefer that type of patch instead of tweaking interest rates within such a short span of time

2 Likes

Yes, and I do agree that the treasury funds should be used wisely (even more so given the current market conditions). Just wanted to point out that 100bps raise in interest rates (though “free” tool from a financial perspective) is most probably not enough to kickstart a repaying of existing debt, but only stop/slow down further demand for new debt creation. Thus, the adverse impact (mainly the halt in the expansion of GHO’s supply and TVL) might make the decision not worth the risk.

My point is that GSM will need to be bootstrapped either way and without any clear demand-driving tool for GHO such bootrstrapping will be hard to do organically. The current amount needed to re-peg GHO is not very high and I would argue that using some portion of treasuries (or reserves even) in the GSM (ideally more than is needed to re-peg) would bear better R/R than raising interest rate arbitrarily. Of course, I must note that my thesis is driven purely by my observations and I have not done any “deeper” analysis (like modelling the demand/supply and knowing the exact $$$ amount needed to re-peg).

1 Like

Increasing the utility of GHO is critical for adoption and peg stability, while on the other hand, peg stability is crucial to achieving confidence that will drive more usage. At the same time, seeing the recent drop in liquidity in balancer pools over the past couple of days, we are not sure if the suggested borrow rate increase will have a substantial effect. As @MarcZeller has noted, both current and suggested borrow rates are below market average rates. While it is hard to establish that the borrowing rate increase will have a strong impact on the GHO peg, we do not see this introducing additional risk.

Untitled - 2023-08-24T111752.899

Decrease in Balancer pool liquidity - TokenLogic Dashboard

Alternative Solutions

As mentioned by other community members on this thread, the GSM could introduce a vital peg stability aid in the long term, but to increase confidence and adoption in the immediate term, other measures need to be taken. Increasing GHO utility, such as enabling onboarding wGHO as collateral, can support the re-peg of GHO. We are aware of the concerns raised by Gauntlet here, but given risk controls on V3, primarily listing wGHO in isolation mode with a community-approved debt ceiling, this risk could be managed continuously, considering the evolving liquidity and market circumstances.

5 Likes

Yeah, I think driving demand should be the main concern right now. Be it by LP subsidies or by using GHO as collateral.

1 Like

Quickly catching up on the discussions.
Re-pegging, even with the GSM live will not solve the issue as there is currently no mechanism to drive the price upwards back to peg.

Raising the interest rates is a no brainer, but I wonder if we shouldn’t eventually move up to a dynamic rate at some point. The low interest rates are really good for rapid adoption, but as we’ve seen with the past month, there aren’t enough demand sinks for GHO yet, so users swap it out for other assets in a more excessive fashion than the liquidity can handle.

As it stands, repegging is not even a costly operation as ~1.5M$ would be sufficient to do so.
image

Basically GHO needs more integrations, that scale with minting, as well as interest rates that prevent peg deviation.

5 Likes

I agree that a dynamic interest rate is a good solution combined with liquidity incentives that can drive demand up.

The repegging operation you just described is not very representative of the real situation though because given the recent balancer vulnerability there has been a significant drop in liquidity in the gho boosted pool there which was the most liquid venue for gho.

Token logic has a good analytics dashboard for that.

Proposal escalated to Snapshot stage.

1 Like

With AIP-308, the GHO borrow cap will be cut to 35m, limiting potential depeg capacity. The effect of Increasing the GHO borrow rate to 2.5% is unclear for the growth trajectory for GHO, but does not add risk.

  • Increasing borrow rate to 2.5% may encourage GHO minters to repurchase GHO on market to repay GHO debt
  • Balancer incentives of 10% are sufficiently higher than the increased borrow rate, suggesting that liquidity flight from minters that chose to LP is unlikely.

On GSM capabilities

A well designed and sustainable GSM is critical to maintaining GHO peg, since it will facilitate continued decentralized GHO-USDC/USDT 1:1 swaps when GHO < $1.

With the current envisioned GSM, users can mint GHO 1:1 without liquidation risk or DEX swap - a capability that does not exist on Aave v3 - to participate in incentivized LP strategies. Although it is unclear as to the composition and size of exogenous assets in the GSM upon launch, this new capability should spur organic GHO mint via GSM. Utimately, GSM sustainability is dependent on significant use cases for GHO.

2 Likes

Glad to see the Snapshot passed!

I wanted to provide some perspective on the GHO peg, based on prior experience working on DAI. Apologies for the long-ish post :smiley:

Current State

As @MarcZeller mentioned, GSM will only provide peg support if overall GHO demand is greater than supply (this pushes price up to $1 or slightly above, which creates incentives to mint GHO with the GSM). Introducing a GSM is useful but not really a short term fix.

This leads to a related point, which is that stablecoin growth can only be led sustainably via the demand side, not the supply side.

If incentives to increase supply (eg low fixed GHO borrow cost) are greater than incentives driving demand (liquidity incentives, other organic usage drivers / supply sinks), the peg will tend to break downwards. This ends up curtailing growth because at a certain point, borrowers will not be willing to mint and sell GHO at a discount given the risk that price will return to peg later and they will lose money on the spread between when they opened their loan and when they eventually repurchase GHO to close out. We’ve seen other cases of supply side growth strategies and TLDR is they never work unless there is an equal or greater level of demand (idk if anyone remembers GUSD minting incentives from last cycle, which backfired severely when these tokens were quickly returned to Gemini to be redeemed or dumped on the market).

GHO needs to drive demand to achieve sustainable growth, and ideally we want to create organic unincentivized demand (people who are holding GHO mainly for utility, rather than to purely for earning yield). Organic demand is where the true benefits of making a stablecoin emerge, as the issuer can earn a spread between the average yield earned on GHO backing (primarily GHO borrowers but potentially also GSMs and RWAs in the future) and yield demanded by GHO holders.

How to get organic demand? There are a lot of factors here, but I believe the most important are trust and liquidity.

When users trust a stablecoin, they assign it a lower implicit risk premium. And if a stablecoin is highly liquid, it provides users a greater level of convenience yield (basically, a way of thinking about the utility value of an asset, like how people are willing to carry some cash around in their wallet earning 0% for convenience) as they can use it for trading or commerce purposes more easily. Both of these factors (lower risk premium and higher convenience yield) improve the all in value of holding a stablecoin like GHO, without the issuer needing to pay for it.

GHO’s current under peg state is negative for trust, as users have uncertainty on when the price will return to $1 and are exposed to volatility. And trading under peg is also negative for liquidity, as the swap market impact per unit TVL is much greater within stableswap pools like Curve and Balancer when an asset is trading off peg. With this being said, returning GHO to peg should be a top priority even if it leads to a contraction in supply in the short term, as this is a necessary precondition for building up the organic demand that will make GHO succeed over the long term.

Action Items

The current proposal to increase GHO minting costs from 1.5% (1.05% with stkAAVE) to 2.5% (1.75% with stkAAVE) is a great starting point. But I think Aave DAO can consider additional actions if this change alone is insufficient to bring GHO back to peg (as I suspect will be the case).

Forward Guidance and Primitive Rate Automation

If users expect borrowing costs may increase in the future, they should be less likely to contribute to over-supply by minting and selling GHO while it remains under peg. Creating credible expectations of maintaining GHO peg can even help achieve supply demand balance without needing to raise borrowing rates as high as otherwise might be necessary.

While this could be automated in the future, for the intermediate term Aave governance could commit to a set rate adjustment plan when GHO is under peg, such as raising GHO borrow cost by 0.5% each month until GHO price is >= $0.995 for 90% of the previous monthly period. (Specific parameters and thresholds are up for debate of course)

Liquidity Incentives

While I don’t think Aave should directly spend tokens to incentivize GHO liquidity, the DAO does have significant amounts of CRV / CVX / BAL etc that can be used more effectively to direct liquidity incentives towards GHO pools. This should help create some demand pressure to bring GHO back towards peg.

Protocol Owned Liquidity

GHO trading below peg presents an opportunity for Aave DAO to purchase some at a discount using existing stablecoin reserves, and then provide into LPs on Balancer/Curve/etc to improve GHO market depth and earn fees and liquidity incentives.

Final Thoughts

I think Aave DAO should refocus the GHO growth strategy on the demand side, as this (combined with introducing a GSM to prevent over peg situations) will allow GHO to achieve the trust and liquidity needed to begin building a base of organic users.

While I know this may not be a popular view, I personally expect that GHO won’t be able to find a sustainable supply demand balance until GHO borrow costs are close to as high as the DSR, as otherwise there will be too great of incentives to participate in looping strategies which push GHO price down below peg (supply sDAI, mint GHO, sell for DAI, deposit to sDAI, repeat).

It’s time to put the supply side growth strategy out to pasture, so GHO can achieve its true potential.

LFGHO

16 Likes

how is this man here, giving free, gold-level 1000$/hour advice?

why instead of him and other stablecoin professionals, are we trusting some kind of Risk-focused organization to grow GHO? They aren’t a growth team. They manage risk (duh).

I remember at the start of GHO there was talk about a GHO growth and stability working group. What happened to that? IMHO we should start one.

Here is an example from Arbitrum:

1 Like

Great post @monet-supply

Additionally, it can be used the extra spread (increased minting costs) to bribe veBAL, vlAURA, veCRV or vlCVX holders; and direct emissions in the form of rewards to the main LP of GHO; this would help discourage minting and dumping, while also incentivizing the main LP.

2 Likes