Frax: Every Decentralized Stablecoin Will Become a Central Bank
Frax is a fractional algorithmic stablecoin that appears to be on the cutting edge of becoming a central bank. To maintain widespread usage and relevance, the other decentralized stablecoins will likely have to follow suit.
How Frax Works
Frax is a partially collateralized algorithmic stablecoin and falls somewhere between USDC and UST. USDC is backed 1:1 with USD or T-bills. UST is backed by its ability to mint $1 worth of Luna, which can then be sold on the open market. In contrast, Frax is backed partly by collateral that lies within the protocol and partly by Frax Shares (FXS), which reside outside the protocol and that help absorb Frax’s volatility. In this respect, FXS is somewhat similar to Luna for UST.
Frax’s collateralization ratio (CR) varies with market demand for Frax. Currently, the CR is 85%. This means that to mint 1 Frax, one would have to present the protocol with $0.85 of USDC and $0.15 worth of FXS. Similarly, one can redeem 1 Frax for $0.85 of USDC and $0.15 worth of FXS, at which point that 1 Frax would be burned. However, as market demand for Frax increases, the CR will fall. If Frax increasingly gains adoption, then the CR may eventually fall close to zero, causing it to be fully backed on the margin by FXS, thereby mirroring UST and Luna to a substantial degree.
One benefit of this design is that the CR increases as demand for Frax falls. On the margin, redemptions will tend to place increasingly larger amounts of USDC into the protocol, leading the remaining Frax to become more collateralized. Rather than becoming weaker in the face of heavy redemptions like UST, Frax actually becomes stronger.
Protocol Controlled Value
Frax launched with a CR of 100%, but this has since declined to 85% given a Frax supply of $2.8bn. In October 2021, Frax launched strategy specific algorithmic market operations (AMOs), which automatically sweep excess collateral from the protocol’s treasury and invest it in interesting ways. Similarly, the AMOs will push collateral back to the treasury if the CR increases. Profits from these AMOs are promised to the veFXS holders, who have locked up their FXS. Frax’s primary AMOs focus on: 1) liquid collateral; 2) Curve operations; 3) liquidity provision at DEXs; and 4) lending.
The collateral AMO simply maintains liquid collateral, investing its proceeds into Compound, Aave, and Yearn. These assets are kept liquid to potentially send to the treasury if the protocol faces Frax redemptions or requires greater collateralization. A hedge AMO will eventually help maintain the price stability of these assets as well.
The Curve AMO maintains the Frax/3CRV pool at Curve, which currently has $2.8bn, including $1.6bn in Frax. Since the 3 Pool includes more established stablecoins (USDC, USDT, and DAI) and the Frax/3CRV pool ties Frax tightly to these stablecoins, this also firmly establishes Frax’s peg to $1. As an early entrant into the Curve Wars, Frax owns 19% of all DAO-controlled vlCVX and is also the top briber of vlCVX on Votium. This allows Frax to push CRV rewards to the Frax/3CRV pool, which both rewards itself (since it makes up 57.5% of the pool) and also incentives others to provide liquidity to the pool. Trading fees and LP rewards go to the veFXS holders, while the new CRV gets locked up to further entrench Frax’s influence. Recently, Frax and Luna announced that they’re going to create a 4 Pool (USDC, USDT, Frax, UST) and, along with Redcated Cartel, will concentrate CRV rewards there.
The liquidity AMO acts as a liquidity provider for Frax pairs at various DEXs involving a total of $508mn. Approximately 60% of this capital is placed at three different DEXs on Fantom, with the remainder spread across DEXs on various chains. This increases the use of Frax and also earns LP trading fees for veFXS holders.
Lastly, the protocol directly deposits 151mn Frax at various borrow/lend platforms. Over two-thirds of this amount is invested at Aave and various pools at Rari, with the remainder spread across many other borrow/lend platforms. Depositing Frax directly at these platforms allows the protocol to directly impact the utilization rates and thereby control the interest rates that Frax borrowers face. MakerDAO does the same thing with DAI at Aave. The fees from Frax deposits at these platforms goes to veFXS holders as well.
Since AMOs were introduced in October 2021, Frax has accrued $104mn in profits. The profits create excess collateral, and the protocol mints new Frax against this excess collateral, using it to purchase FXS on the open market to then distribute to veFXS holders. Though it’s speculative on my part, I wonder if the protocol might one day instead reinvest some of that newly minted Frax into the AMOs to generate greater long term profits for veFXS holders. Similarly, as Frax receives broader adoption, the CR will fall and create excess collateral, which would also allow the protocol to create new Frax.
Borrow/Lend: Death & Renewal
Borrow/lend platforms like Compound and Aave typically set up their interest rate curves to quickly spike up after 80% utilization (i.e., when the amount of assets borrowed is more than 80% of the amount of assets deposited). They do this to create an asset reserve so that depositors can redeem their holdings at any time. Otherwise, there might be a bank run as people fear not being able to pull out their money.
However, what this means is that the borrowing rate must be higher than the deposit rate, since the borrowing rate occurs on a smaller asset base (i.e., the interest accrued from borowers must cover the interest paid out to the larger deposit base at a lower interest rate). Within this context, Morpho is set to completely upend the borrow/lend platforms.
Morpho is a meta platform that sits on top of borrow/lend platforms like Compound and Aave. Let’s use Aave as a specific example. Instead of initially depositing or borrowing ETH from Aave, a potential investor would instead interact directly with Morpho. If an investor wants to deposit 10 ETH on Morpho but there isn’t a corresponding borrower on Morpho, then the protocol will send that 10 ETH to Aave to receive the prevailing deposit rate. The same would occur on the borrow side if an investor wanted to borrow 10 ETH on Morpho but there’s no one who has deposited that same asset at Morpho. However, if depositors and borrowers do match up on Morpho, then they’d both be quoted the mid-rate between Aave’s deposit and borrow rates. This would represent a higher deposit rate versus Aave and also a lower borrow rate versus Aave. Matching lenders and borrowers up on a peer-to-peer basis (rather than a pool-to-peer basis) should enable Morpho to siphon away all of the matched liquidity on the underlying platforms. This will severely undermine the position of existing borrow/lend platforms.
One solution to this problem would be to create a system akin to certificates of deposit (CDs) where depositors are locked up for a fixed period of time and receive a fixed interest rate. Borrowers on the other side would pay that fixed rate over the same time period. However, this would get complicated since the market clearing interest rate will differ for CDs of different tenures. Also, some secondary mechanism might need to be created to allow depositors and borrowers to leave the arrangement if they’re willing to exit at a discount. This may still end up occurring for non-stablecoin borrowing and lending.
However, a cleaner solution exists for stablecoin lending. Instead, Frax could create its own borrow/lend platform and create separate pools for Frax deposits and loans based on the kind of collateral provided. The protocol can then deposit Frax directly into each pool, so that it sets the market clearing interest rate (where the amount of deposits equals the amount of loans). The collateral ratio of each pool may differ substantially depending on the riskiness of the collateral (e.g., 1.05x for USDC but 2x for volatile tokens), but the overall interest rate targeted by Frax could be the same for each pool. This would allow Frax to create a unified interest rate independent of collateral type.
Just like depositors in commercial banks no longer fear bank runs due to the backing of the FDIC/Fed, so Frax could guarantee that any depositor can redeem by paying them in Frax directly, if need be. Frax would build up a large amount of excess collateral from the interest on its direct deposits, and if the CR is low enough, this excess collateral could stand behind the potential issuance of Frax necessary to engender confidence in such a system. While I don’t know the details of Frax Lend, which is set to launch this summer, my sense is that the structure will be at least somewhat akin to what’s been described so far.
Liquidity for Every Asset
Currently, if a new protocol or DAO only holds its native token in its treasury, there’s no good way to create liquidity for the token. The protocol would need to acquire ETH or stablecoin to pair with its native token at an AMM, but no market yet exists to acquire that ETH or stablecoin with its own token. On the other hand, if the protocol sells its native token directly to the public to build up ETH in its treasury, then a substantial portion of that ETH may get tied up at AMMs, instead of being used to pursue the protocol’s actual mission. This approach also faces the possibility of impermanent loss (IL). Lastly, if the protocol uses LP rewards to motivate others to provide liquidity, then that increases the native token’s circulating supply and is inflationary.
Tokemak helps solve this issue. Tokemak sells TOKE in exchange for ETH or stablecoins, which then become protocol controlled value (PCV). Outside protocols or DAOs can deposit their native tokens in “reactors”, and Liquidity Directors then stake their TOKE at reactors to push the ETH and stablecoins to specific reactors, pairing their liquidity up with the native tokens and sending both to AMMs. TOKE emissions reward both sides (the depositors and the Liquidity Directors), and the TOKE APRs are set for each side such that the same portion of deposits in each reactor gets matched with liquidity. Outside protocols don’t have to worry about IL, with TOKE holders making money over time as the LP fees outweigh the IL (plus, they receive TOKE emissions as Liquidity Directors).
However, for a stablecoin, an even easier solution exists. Frax could create a DEX where outside protocols deposit their native token at AMMs, as this then gets paired with Frax directly from the protocol. If Frax collects all trading fees and absorbs all IL, then it has taken on Tokemak’s role without having to raise any initial funds from outside the protocol. Also, if L1s develop to the point that reliable central order books are possible, then Frax could create such an exchange and use its own liquidity to market make across the various listed assets. The trading fees from such a DEX would also help build up excess collateral and enable the further issuance of Frax for both this and other purposes. FraxSwap, which launches next month, may or may not follow these contours. But perhaps this might be the long-term goal.
Decentralized Stablecoins: the New Central Banks
At this point, say that Frax can:
- Substantially expand its money supply at its discretion — due to excess collateral from profits and a falling CR
- Directly set a single unified borrowing rate for Frax across every collateral type
- Create liquidity for literally any digital asset (cryptocurrencies, NFTs, etc.)
If Frax can do all three of those things, then my sense is that it should be considered a central bank for the crypto economy. Interestingly, Frax is also developing the Frax Price Index (FPI), a token that will increase in price based on inflation, effectively creating a new kind of protocol specific currency that’s not tied to the U.S. dollar.
Frax will face stiff competition from other decentralized stablecoins. For example, perhaps UST, DAI, and UXD will try to integrate a borrow/lend platform and a DEX. Perhaps FEI, MIM, and alUSD will attempt to incorporate a DEX. The chance that the other decentralized stablecoins just lay down and let Frax take over seems low.
However, Frax could experience a flywheel effect as an early mover. If Frax is the first to establish an integrated borrow/lend platform and a DEX, then the associated profits and greater adoption of Frax would boost excess collateral, which would allow more Frax to be issued. This increase in protocol-controlled Frax could then be used to strengthen the borrow/lend platform and the DEX, minting even greater profits. That would in turn build more excess collateral, which would lead to more Frax issuance, and so on. While this may not preclude other decentralized stablecoins from competing, it could provide Frax with a dominant position. But to compete, every decentralized stablecoin will have to become a central bank.