Liquidity Pricing for Stablecoins

The premise of a stablecoin is a tokenized demand deposit. The ability to realize a determinate amount of value at any time, freely transferable, and possibly interest bearing. The specific denomination is less relevant to this discussion, although so far the clear majority of demand has been in dollars.

Holding a large idle reserve of non-yielding centralized stablecoins is undesirable from the pespective of Ethereum-native stablecoin protocols like Volt and Maker. It creates unnecessary counterparty risk and generates no revenue.

Maker is pursuing an intermediated real world asset strategy, with the goal of accumulating enough ETH to back a large portion of the DAI supply with a Protocol Owned Vault. I’ve considered the problems with this idea elsewhere, and I’d also suggest reading Luca Prosperi on the topic:

Conceptually, the Surplus Buffer would go from being a cash-like cushion to protect against unexpected losses (in line with the concept of Common Equity Tier 1 capital of traditional banks—we had discussed it already here) to some sort of DAO-owned hedge fund with yield and volatility. The intention is then to reinvest in this way most of the yield generated by using the dollars in the PSM. The shift from being a permissionless set of smart contracts providing ability to leverage $ETH and $BTC to a Fei-like protocol-controlled fund that uses the $DAI stablecoin product to aggregate value is clear to me.

Yikes. Surplus buffer management aside, it’s quite unclear in this framework how much MakerDAO needs to keep on hand in the PSMs to manage redemption demand, how much is safe to deploy in RWA of a given maturity, and so on.

Beyond the Peg Stability Module

The original premise of a CDP stablecoin was a set of loans against collateralized positions with native debt unit, whose price stability to an external peg was based on the threat of collateral liquidation and periodic interest rate adjustments.

These indirect mechanisms for managing liquidity were crude and insufficient, especially as demand for stablecoins grew vs demand for leverage against ETH and other cryptoassets. When demand for the stablecoin is high price will trend above peg, making it more expensive for borrowers to repay their debts. When there is insufficient demand to repay debts vs to sell the stablecoin, it can easily drop below peg.

The PSM offered a quick and easy fix, but as mentioned above and widely lamented, introduced centralization risk and is an unproductive form of capital.

RAI introduced an alternative, the use of a Price Controller. Instead of unpredictable volatility around the peg, the controller makes predictable adjustments that allow easier coordination. Despite this improvement, the stablecoin supply is still constrained by the demand to take on debt against the accepted collateral. In a multi collateral system, there is no clear way to decide which rate should be charged on each collateral, and it’s not viable to use a controller to do things like adjust collateral factors.

A viable replacement for the PSM model should maintain the property of reliable liquidity on demand for stablecoin holders at peg, while not requiring idle centralized stablecoins to be held within the system. The loans or other assets backing the stablecoin must be callable by the stablecoin holders on demand. There must be some price to call these loans to prevent griefing.

You may have already read my recent idea for a callable loans primitive – if not please have a look. While callability is essential for lender-safety and strong peg defense, it’s bad UX for borrowers. Mathis from Morpho Labs suggested that as an alternative to calling a loan, the lender could attempt to auction off the loan itself. This is not efficient for a widely diverse set of nonfungible loans, but with a common debt unit, coordination is easier.

We then have a more complete model of sell-or-call, where anyone willing to accept a discount on the stablecoin less than the cost-to-call can do so. For example, assume the fee to call a loan were 0.05%. There might exist a pool VOLT-USDC, where upon deposit the pool offers a swap at the current VOLT index price (a system value based on yield accrued over time), and gradually decays to a 0.05% discount, after which loans are called to satisfy the pool’s redemption demand. The discount would be reset to 0% either if a swap occurs filling the pool’s demand or if the loans are called.

If you like this sort of thing, please also have a look at my recent article about auction-style liquidity provision.

This model still can’t handle a stablecoin demand larger than the available collateral base. Rather than idle centralized stablecoins, it should be applied to loans against high quality and credibly neutral real world asset collateral.

Real, hard asset collateral backing is surely required to satisfy demand for stablecoins at scale. To make it work, we must have:

• scalable approach for setting interest rates
• ability to call loans, with fair call-time and calling cost to prevent griefing of borrowers
• no need for dependence on trusted oracles
• robust mechanisms for rejecting bad collateral, or offboarding existing collateral
• strong right-to-exit and right-to-veto for varied stakeholders
• ample first-loss capital + incentive aligned governance

The onchain protocol does not need to be opinionated about types of legal vehicles and jurisdictions, and there is a concerning amount of insider dealing within MakerDAO. Our preference at Volt Protocol would be to have access to a variety of asset tokenizers and borrowers. I anticipate substantial short-term progress in this sector, as much groundwork has already been laid fueled by the market exuberance of the last two years.