Preliminary Risk Profile of
Adding FRAX to Aave’s Ecosystem
Introduction
In this post, we perform a preliminary analysis of the risk in adding FRAX as collateral on Aave V2.
The general approach we took was to draw comparisons between existing tokens used as collateral on Aave versus FRAX.
The findings of the analysis presented in this post show that adding FRAX to Aave would not increase the risk profile of Aave but in fact, leave it unaltered and potentially even lower it. The benefit of adding FRAX is clear: This represents a potential boost of liquidity and revenue to the Aave ecosystem, a lot of new DeFi users, and an improved Aave that supports the most innovative tokens in DeFi today. We are excited to bring forth this proposal and opportunity.
Analysis Setup
Gauntlet Network performed a comprehensive Market Risk Assessment of Aave here. In this post, we argue that, since FRAX’s price fluctuates on a similar scale, with similar volatility as existing tokens used for collateral, adding FRAX will not increase Aave’s risk profile.
There are other components that could decrease the quality of the risk profile, however, even if prices behave consistently: Wild swings in specific markets may occur. Additionally, there is a risk of having a low amount of liquidity available to liquidate loans. Thus, we include data in the post to show that FRAX’s swings during volatile periods are comparable to existing collateral types, and that there is consistent liquidity for FRAX in the market.
Why FRAX?
We are bullish on FRAX because it balances the stability of centralized stablecoins with the utility of decentralized ones. From this source: Many stablecoin protocols have entirely embraced one spectrum of design (entirely collateralized) or the other extreme (entirely algorithmic with no backing). Collateralized stablecoins either have custodial risk or require on-chain overcollateralization. These designs provide a stablecoin with a fairly tight peg with higher confidence than purely algorithmic designs. Purely algorithmic designs such as Basis, Empty Set Dollar, and Seigniorage Shares provide a highly trustless and scalable model that captures the early Bitcoin vision of decentralized money but with useful stability. The issue with algorithmic designs is that they are difficult to bootstrap, slow to grow (as of Q4 2020 none have significant traction), and exhibit extreme periods of volatility which erodes confidence in their usefulness as actual stablecoins. They are mainly seen as a game/experiment rather than a serious alternative to collateralized stablecoins.
Frax attempts to be the first stablecoin protocol to implement design principles of both centralized and decentralized models to create a highly scalable, trustless, stable, and ideologically-pure on-chain money. The Frax protocol is a two token system encompassing a stablecoin, FRAX, and a governance token, Frax Shares (FXS). The protocol also has a pool contract which holds USDC collateral. Pools can be added or removed with governance.
How does FRAX keep its peg? FRAX can always be minted and redeemed from the system for $1 of value. This allows arbitrageurs to balance the demand and supply of FRAX in the open market. If the market price of FRAX is above the price target of $1, then there is an arbitrage opportunity to mint FRAX tokens by placing $1 of value into the system per FRAX and sell the minted FRAX for over $1 in the open market.
Qualitative risk assessment
First, as examples of tokens already supported by Aave and part of the risk environment, we find USDC, FEI, TrueUSD, and DAI. We can start by making some comparisons with FRAX from a risk perspective.
The risk of USDC originates from it being pegged to a centralized system. DAI is much less risky from this perspective since it is a decentralized stablecoin but unfortunately it is still backed by a lot of centralized assets. FRAX, being a mix of centralized and decentralized sources, is also much less risky than USDC from this angle.
FRAX’s price cannot really deviate a lot more than other stable coins because when it does, easy lucrative arbitrage opportunities will be created in turn re-stabilizing the price: When FRAX is trading below $1.00, arbitrageurs are incentivized to buy FRAX at a discounted price, and potentially pay off some of their debt at a cheaper price than if FRAX were $1.00. If FRAX is trading above $1.00, arbitrageurs are incentivized to borrow more FRAX and sell it at a premium. In fact, FRAX is typically more stable doing volatile times than other stable coins — this obviously is the very behavior we seek in regards to minimizing unintended liquidations, etc. on Aave.
Another fact to mention is that for Aave having a lot of on-chain liquidity is important in order to perform liquidations. As measured by market cap and trading volume — Frax is a top-10 stablecoin with a high level of on-chain liquidity. This further lowers the risk of being unable to liquidate when necessary when compared to other supported coins.
These perspectives support the case that adding FRAX to Aave will fit entirely within the existing risk profile.
Quantitative risk assessment
Our qualitative look supports the case that FRAX fits well into the Aave ecosystem, standing up to even some of the longer-standing assets supported by Aave. Let us next turn to a more quantitative angle, analyzing pricing data.
Our approach is the following. What should matter most in terms of the risk of liquidation is price spikes in the stable coin during volatile times. Ideally, the coin stays pegged at $1 of course, but fluctuations happen giving rise to volatility in the stable coin leading to risk from Aave’s perspective. We now show that FRAX has a comparable and at times smaller, risk profile than existing coins already supported on the platform.
We obtained historical data for a subset of Aave stablecoin tokens supported now, namely USDC, FEI, TrueUSD, and DAI, and of course pricing data for FRAX as well. Then, we formed the band between low and high prices for a basket of “non-FRAX” stablecoins (USDC, FEI, TrueUSD, and DAI). Then we found the overall minimum value of those lows. This was repeated for each day. The same was done for the daily high prices. The band thus created each day between the low and high of the coins in this set gave the “non-FRAX” pricing band. If FRAX is not consistently the extremes of this band (forming the envelope) this supports the case that its risk profile is similar to the other supported coins. Let us plot all pricing data of Aave stablecoins (with FRAX) together:
Next, we plot FRAX and the non-FRAX basket all grouped together for better visual clarity:
We see that compared to the current price fluctuations of the non-FRAX basket (formed by the blue band) FRAX does not add more volatility to the risk profile of Aave’s collateral stable coins. In fact, the price is more stable around $1 with a generally narrower band than the existing non-FRAX basket.
Liquidation Analysis
If FRAX behaves like existing Aave stablecoins, adding it to Aave will return a similar (and in some cases more favorable) liquidation profile as Aave stablecoins. Next, we take DAI as an example since both DAI and FRAX belong to the decentralized stablecoin family, we analyze the on-chain data and find the number of liquidations from DAI and USDC — both used for collateral on Aave:
Since December 2020 ~73 liquidations have occurred for DAI on Aave v2. This was against a total of 18,904 loans created. For these loans, this amounted to $4,027,593 in lost funds to the users. This is in contrast to a total of $6,889,982,674 in loans made.
Liquidations originate from two sources: the collateral token’s volatility being the cause (this is unintended and scenarios to be minimized from Aave’s perspective, we can call them “unintended liquidations”) and the borrowed asset’s volatility being the cause (“natural liquidation”). Assuming, however, that all of these liquidations are unintended, we estimate the liquidation percentage of such DAI liquidations to be less than or equal to 0.4%. In terms of amounts, we see that 0.058% of value was lost in said loans when held in contrast to how many funds were borrowed. In other words, more than 99.9% of borrowed funds are not liquidated for DAI loans.
From this, we extrapolate that adding FRAX would leave a liquidation percentage of similar magnitude — and most likely less. In other words, taking out 1,000,000 loans against FRAX on Aave should result in no more than 4,000 unintended liquidations. And on $1,000,000 of borrowed funds from using FRAX as collateral we expect less than $580 to be liquidated.
Suggested Risk Parameters for FRAX
Keeping with DAI as a comparison stablecoin, we suggest the following risk parameters for FRAX:
Conclusion
We find from both a qualitative and quantitative perspective that adding FRAX to Aave will greatly support and benefit the platform. We purely look at the risk profile in this work, analyze it from a qualitative and a quantitative angle, and in addition, we can make the case that the benefit/reward to Aave will be a massive influx of capital to the system along with a large number of users all benefiting and strengthening Aave as a critical lending platform.
About Bribe
Bribe creates DAO infrastructure tooling to incentivize protocol participation. Depositors stake their governance tokens in the Bribe pool to earn income. Bidders borrow the staked votes to support or reject governance proposals. Bribe V1 introduces Voter Extractable Value (VEV) to coordinate and auction powerful coalitions of DAO votes. Bribe is best used as part of a well-balanced and active delegation strategy.
