Here we go again. An innovative new product on the horizon should often lead me to a new article. This week, I came across the Whitepaper that describes the concept of a new stablecoin on Ethereum, $GHO. Aave is in charge.
It's not every day that you see a serious lending protocol creating a stablecoin. I’m hyped so I dug a little bit. It's never easy but as always I will try to be simple and precise in expressing my ideas. Let’s GHO.
First things first, what’s GHOing on here?
According to the whitepaper, “GHO is a decentralized, over-collateralized crypto-asset intended to maintain a stable value.” It doesn’t necessarily say that it will be pegged to $1, maybe it's just for regulatory reasons.
GHO can be exclusively mint by the facilitators. Different entities belonging to different market segments will be allowed to mint a fixed GHO amount at a previously set interest rate. The main stabilization mechanism will be the over-collateralization. The idea is to have several stabilization mechanisms working at the same time.
The role of the facilitators is to diversify the assets collateralizing GHO while maintaining some control over its issuance. Different types of facilitators are cited:
Aave himself, Risk-weighted assets and delta neutral positions are the collateralized facilitators of GHO. I won’t focus on the under-collateralized facilitators because we never saw effective integrations of these types of products in crypto, they are the next development stage for GHO. They should enable better capital efficiency for the stablecoin in the future. But I’m still going to talk a little about RWAs.
Diversification, decentralization or both?
We saw previews of a working model in the Maker Governance forum. The protocol can mint $DAI to lend it to entities who will invest it in various RWAs (SOPR, Gov Yield, fixed and floating per eg).
The model works like this: each loan allocated to RWA will be represent by an ERC20 token which correspond to the value of Maker's asset and the accumulated return. This allows more transparency on the operation and its follow-up for Maker DAO members. Lending decisions will be handled by on-chain governance.
It is a smart way to expand the supply and type of assets supporting $DAI. But there are downsides.
These type of strategies require off-chain maintenance and trust assumptions. Maker need human monitoring of off-chain data to make sure everything is working according to chosen agreements (performance of the collateral, debt ratios). This is a resource-intensive task requiring deep expertise as transaction sizes increase.
They also need to work with an operator that will purchase the RWAs for them and a trustee who will manage the custody of these assets.
Some will argue that leaving the chain isn’t necessary to diversify $DAI backing, that these investments risk attributing too much dependence to centralized entities. Others will say that it’s the price to pay to grow and that perhaps it is better for Maker to trust the American Government rather than DeFi protocols. The truth is probably between the two. This paradox is amusing, decentralization which was the image of progress not so long ago, could become a drag on growth.
Philosophical values must not be drowned out by the rationality of investment because this distances us from our identity even more quickly. Maker has shown an interesting path that can make both ways of thinking coexist.
In conclusion, RWAs are a relatively safe bet for treasuries to generate low-risk yield out of DeFi and Maker is right to experiment with it. If Aave want to implement similar strategies, they must be ready at all levels because it can create frictions, dysfunctions and philosophical differences if not properly implemented.
Now all eyes on the first facilitator of $GHO, Aave protocol.
Facilitators will have the power to trustlessly mint and burn GHO. They each have a bucket with limited capacity coupled with a variable interest rate determined by the Aave governance.
Aave can control the GHO supply by balancing liquidity between buckets to maintain overall system collateralization and stability in a decentralized and permissionless way.
Logically Aave is the first and lead facilitator of GHO.
How it works?
A la Aave, users will provide collateral (unknown to date) to borrow $GHO at a specific collateral ratio and interest rate. Available collaterals should be ETH/BTC, a few defi blue chips and majors stablecoins. We saw recently that deep liquidity on tokens that act as collateral is essential to prevent market manipulation and exploits. Relatively low volatility on collateral is preferred to make the dollar more stable. I will keep an eye on the presence of yield bearing assets as collateral (wrapped staked ETH tokens).
ATokens and GHO debt token will be distributed to the borrower and can be used on the Aave Ethereum Market. Liquidations work in the classic way and the liquidated GHO is burned.
1 GHO minted by the Aave facilitator will always worth 1$. Accordingly, this allows the usual arbitrage when GHO value on the market differs from 1$. When the price is lower than 1$, it encourages buying GHO on the open market to repay a loan. When the price is higher than 1$, it encourages to borrow GHO to sell it on the open market and make a profit.
Interest Rate
When a new stablecoin comes to town, one often wonders what its yield is and then what its interest rate is.
In the GHO model, the interest rate can’t be determined algorithmically based on the dynamics of supply and demand because supply does not exist. The Aave governance will adjust the interest rate according to the demand for GHO.
Now my favorite part, where the innovation lies.
In my opinion, there are three main goals for GHO :
Increase Safety Module funds to build trust and increase resilience to a shortfall event
Increase utility for AAVE holders
Generate a new source of income
The staking design of AAVE is in line with the interests of the protocol as it contribute to its safety. Users can lock AAVE or AAVE/ETH Balancer pool tokens in the Safety Module to earn protocol fees and rewards in AAVE tokens.
As I write this, about $355M are locked in the Safety Module. The yield is higher for staking ABPT rather than AAVE (16.3% vs 6.7% APR). It’s not a surprise as Aave recognizes deep liquidity is needed to limit the downward pressure theoretically created by the AAVE token auction sale to cover the incurred deficit.
Indeed stakers can be slashed up to 30% of their stake in the case of a shortfall event.
If it’s not enough to cover the deficit, the system will issue AAVE tokens that can be purchased at a fixed price against ETH and stablecoins.
I think that Aave needs more funds in the safety module, there are $5.72bn in assets in the Ethereum Market and only $118M available (2.06% of total assets at these prices) to pay off any deficit.
But how to incentivize AAVE holders to stake in order to grow the Safety Module?
The current staking rate stands at 21.3% of the circulating supply. It seems that a 6.6% APR is not enough to attract stakers when there is a chance that they lose 30% of their stake. Aave doesn’t want to create more inflation by giving more AAVE incentives to stakers (already a 2.5% annual inflation due to safety incentives).
Generating more revenues for Aave mainly depends on the market environment, as it requires a higher amount of borrowing. The need for leverage is low because prices have fallen all year. That could change soon.
At the moment, Aave does not have enough leverage to make staking more appealing so they have chosen to diversify utility for AAVE stakers. Enter GHO.
StkAAVE holders will receive a discount on the interest rate of GHO.
Here is the example in the whitepaper :
The assumptions required to use this formula are as follows:
stkAAVE > 0, if there are no funds staked in the Safety Module, there will be no discount.
The total amount of GHO borrowed by a user is > the maximum GHO amount a user can mint at the discount (per stkAAVE held). If ≤, the discount will be directly deducted from the interest rate (1% instead 2% of with 50% discount)
What is the game theory here?
We will assume that the discount will be 100% and that at least $1 of GHO can be mint at a discount per $1 of stkAAVE held. If AAVE at $80, 1 stkAAVE can mint $80 of GHO at 0% interest. This is where it gets interesting.
Zero-interest borrowing of GHO is a nice feature for the AAVE token but GHO needs to generate a return to be more attractive. Competition is tough in the stablecoin landscape. Aave doesn’t have CVX/CRV/FXS voting power and therefore can’t guarantee a competitive APY in a GHO Curve pool. I think they’ll just do the minimum to have decent liquidity to perform swaps. So there might be something.
What if stkAAVE holders had a new kind of power, the zero-interest borrowing power. This power can potentially create value, especially for facilitators. (FED interest rate will soon be at 4%).
The facilitators will be in charge of providing use cases to GHO.
Access to borrowing an “homemade” stablecoin for Aave users
Lend GHO to invest in RWAs to generate yield
Long the volatility with delta neutral positions. Wen variants straddles on Aave?
Remember that facilitators will be submitted to ceilings on the GHO minting and a variable interest rate. Therefore if the demand for their product or some products built on top of them is high, they might need additional GHO to fill it. How to acquire GHO effectively if minting parameters don’t allow it due to supply control?
They will have a way to borrow GHO at 0%. By having a big bag of stkAAVE.
Firstly, Aave governance should logically favor GHO minting by stkAave holders through the Aave facilitator rather than through other facilitators. Attracting new stkAAVE holders is in line with GHO’s primary purpose of increasing the funds available in the Safety Module. Thus GHO supplied for all operations linked to stkAAVE holders should be greater than that allocated to the other facilitators. This should drive the other facilitators to borrow GHO on Aave in certain scenarii.
Hum, a stkAAVE war is brewing you said?
StkAAVE tokens should be more valuable once GHO is launched. Let’s take a LTV of 75% for the GHO market and a 100% discount on 100 GHO mint (100$) per stkAAVE (83$). An user want to borrow 1M GHO. It requires $2.83m ($2m in collateral + $830k in stkAAVE) to borrow 1m GHO at 0% with 50% LTV ; $2.02m ($2m collateral + $20k interest paid) without holding stkAAVE and with an interest rate at 2%. The choice looks like a no brainer.
But what if the borrower can access a zero-interest rate for less than $20k?
The operation would look like this:
The borrower bribe stkAAVE holders with $10k to buy their discount borrowing rate power, they will vote to apply their discount for one year to the loan that bribed them. Keeping the same numbers, $10k represent 1.5% APY for $840k of stkAAVE or 10$ per stkAAVE bribed. For stakers that don’t particularly need to use the discount rate, that’s a pretty competitive return for only one loan. For the original borrower in our example, this is a savings of $10k on their loan and, more importantly, a more predictable interest rate of 1%.
Bribing stkAAVE can be even more lucrative for speculators who want leverage because borrowing costs add up faster, making a cheaper solution even more desirable. It will be profitable as long the bribing cost is lower than the borrowing cost without any discount. Borrowers will always seek the cheapest interest rate, whether they are protocols, institutions or individuals.
What could correct this arbitrage?
It won’t be stkAAVE holders, but the GHO system itself firstly. The amount of GHO generated per stkAAVE held will likely increase to make it more expensive for users to borrow GHO interest-free. In this context, the discount should stay the same because the total GHO amount borrowed will not exceed the GHO amount available at discount for stkAAVE. However, this situation should not last.
Thus some users will start borrowing again at the current interest rate, afterwards the system will increase the amount of GHO available at discount, that will bring back the arbitrage to bribe stkAAVE holders.
it could end with a large entity controlling a majority of stkAAVE to make the bribing cost more predictable and more efficient. It would be the end of the war.
In a convex-like situation, borrowers will bribe CVX holders to qualify for interest-free loans.
AAVE holders can earn staking yield added to an extra yield paid in a governance token and bribes, all this without locking AAVE (a 10-day lock-in is required to stake).
I think this is an unexplored market with great potential that can create a some demand for the GHO stablecoin, especially if the appetite for leverage becomes high again.
Regarding the other facilitators, the RWA’s facilitator can be quickly operational with a low and controlled GHO supply allocated in the right integrations (e.g Centrifuge). For now, it is difficult to see other clear use cases for GHO because there is still little information about it and its facilitators. Anyway, the featuring between GHO and stkAAVE is exciting.
Let's not forget that Aave belongs to the DeFi blue chips’s old guard by being the leader in lending markets. This gives it an image of a battle-tested and growing protocol, which it is. We know that the success of a stablecoin is largely based on trust in its creator, as long as everything technically works in the smart contracts. From this perspective, Aave should be able to look Maker in the eye.
I’ll probably do a part 2 of this paper when GHO launches.
NFA. DYOR.
Thanks for your time and take care, Daesu.