MakerDAO vs Frax: Two Different Theories of What a Decentralized Stablecoin Should Be

The question "which decentralized stablecoin is better" gets asked a lot. It's usually the wrong question. MakerDAO and Frax aren't variations on the same design — they started from different premises about what a decentralized stablecoin actually needs to be stable, and those premises led to very different architectures.

MakerDAO's premise: trustlessness requires overcollateralization. If you want a stablecoin that doesn't depend on anyone's promise, the only credible design is one where the collateral backing it is always worth more than the stablecoin itself. DAI has operated on this logic since 2017.

Frax's premise: full overcollateralization is capital-inefficient. If you could algorithmically maintain peg stability with less collateral, you'd unlock a more scalable model. Frax launched in December 2020 as a "fractional-algorithmic" stablecoin — partially collateralized, partially maintained by market incentives around its governance token.

Both premises have been tested, hard. The mechanisms that emerged from each are worth understanding carefully.

How MakerDAO's DAI Works

DAI is minted by locking collateral in a Vault (originally called a Collateralized Debt Position, or CDP) via the Maker Protocol. You deposit ETH, WBTC, or other approved assets, and mint DAI against them at a specific collateralization ratio. The protocol requires overcollateralization — you can't mint $1 of DAI with $1 of ETH. Ratios vary by asset risk, but 150% is a common floor: you lock $150 worth of ETH to mint 100 DAI.

The peg is maintained through several mechanisms working together. Stability fees are interest charged on DAI debt — MKR governance votes on these rates, and higher fees reduce DAI supply by making minting more expensive. Liquidations kick in when collateral values drop below the required ratio: the protocol auctions off collateral to repay debt, maintaining solvency. And MKR token burning — when stability fees are paid, MKR is used to settle them and then burned, which is the protocol's primary value-accrual mechanism.

The Peg Stability Module (PSM), added in late 2020, changed the peg dynamics meaningfully. The PSM allows 1:1 swaps between USDC and DAI, essentially letting arbitrageurs eliminate peg drift at scale. It worked: DAI's peg stability improved dramatically. The tradeoff is that DAI became substantially USDC-backed — at times, more than 50% of DAI collateral was USDC — which introduced centralization risk that MakerDAO governance has been trying to reduce ever since.

The DAI Savings Rate (DSR) lets users deposit DAI and earn yield, funded by protocol revenue. When the DSR is elevated (it hit 8% in mid-2023 during a period of high real-world interest rates), it effectively becomes a policy lever — a way to increase DAI demand and tighten supply.

MakerDAO's governance system runs on MKR tokens. MKR holders vote on collateral types, risk parameters, stability fees, the DSR rate, and major protocol changes. This concentration of governance power is both MakerDAO's strength (professional risk management at protocol level) and its recurring controversy (MKR whale influence on risk decisions).

How Frax Works — and How It Changed

Frax's original design was genuinely different. FRAX launched as a fractional-algorithmic stablecoin with a dynamic collateral ratio. If the collateral ratio was 85%, minting 1 FRAX required $0.85 of USDC and $0.15 worth of FXS (Frax Share, the governance token) burned. The ratio was market-determined: if FRAX traded above $1, the protocol lowered the collateral ratio (trusting the algorithm more); if below $1, it raised it (requiring more collateral). The FXS token's value was tied to protocol confidence — when the algorithm worked, FXS captured that efficiency.

The Terra/LUNA collapse in May 2022 ended the fractional-algorithmic experiment. UST, which had run a similar (but more extreme) algorithmic design, lost its peg and collapsed in days, destroying roughly $40 billion in value. The shared architecture made Frax's risk suddenly visible in a way it hadn't been before. Frax's founder, Sam Kazemian, moved the collateral ratio toward 100% over subsequent months. By 2023, Frax was effectively fully collateralized — the algorithmic component was gone.

What emerged from Frax v3 is a different kind of differentiation. Rather than competing on capital efficiency through algorithms, Frax moved toward Real World Asset (RWA) backing. sFRAX — a yield-bearing wrapper for FRAX — earns a return tied to the Federal Reserve's interest on reserve balances (IOER), routed through FinresPBC (a regulated financial entity Frax partnered with). The pitch: a decentralized stablecoin that earns real yield backed by US Treasuries.

Frax also launched frxETH and sfrxETH, a liquid staking derivative for ETH. This expanded Frax from stablecoin into a broader DeFi infrastructure protocol, using ETH validator revenue to generate yield, and cross-subsidizing FRAX's peg stability with ETH-denominated revenue.

Where the Constraints Differ

MakerDAO's binding constraints are governance and collateral mix. MKR governance has proven capable of navigating crises — the March 2020 ETH crash ("Black Thursday") nearly broke the protocol, but the response ultimately strengthened risk parameters. The harder constraint is USDC exposure: reducing centralized stablecoin collateral without degrading peg stability is an unsolved optimization problem. The "Endgame" restructuring (2022–2024), which involved rebranding DAI to USAI and MKR to SKY, along with creating SubDAOs like Spark Protocol (a lending frontend), is MakerDAO's attempt to evolve the architecture without breaking the core.

Frax's constraints run in a different direction. The pivot to RWAs introduced regulatory exposure — sFRAX's yield depends on a partnership with a regulated entity (FinresPBC), which is a centralization point. If that relationship changes, or if regulatory pressure targets RWA-backed stablecoins, Frax's differentiation collapses. The FXS token's value accrual also depends on protocol revenue staying positive, which requires sustained FRAX demand.

Both protocols share a deeper constraint: the market for decentralized stablecoins has a clear dominant incumbent (DAI/USAI has been live longer and is more deeply integrated across DeFi), and breaking through requires either significantly better capital efficiency or significantly better yield — and both of those are now heavily contested by centralized alternatives like USDC and USDT.

What's Changing

MakerDAO's Endgame transition is the largest governance reorganization in DeFi history. Spark Protocol (the lending frontend, similar to Aave) launched in 2023 and has accumulated meaningful TVL. The rebranding of DAI to USAI and MKR to SKY in 2024 is live but adoption has been uneven — much of the market still uses the old names, and the SubDAO structure is still being built out. Whether this restructuring strengthens or fragments the protocol's coherence is genuinely open.

Frax v3's RWA yield is real and currently operating. sfrxETH has been accumulating ETH validator revenue consistently. The question is whether "decentralized stablecoin backed by US Treasuries" is a stable product category or a temporary niche before regulatory clarity forces a choice between decentralization and yield source.

What Would Confirm Each Direction

For MakerDAO: Spark Protocol TVL growth alongside reduced USDC dependency in DAI collateral. Endgame SubDAO launches that don't fragment governance or fragment liquidity. Sky (USAI) adoption in major DeFi integrations without losing Maker's existing integrations.

For Frax: sFRAX AUM growth with stable FinresPBC relationship. sfrxETH maintaining competitive yield versus other liquid staking options. FXS value accrual sustaining even during low-volatility periods when DeFi revenue contracts.

What Would Break It

MakerDAO: A systemic collateral failure (a large WBTC or RWA position going bad faster than liquidations can clear) would test whether the MKR backstop — the protocol's last resort where MKR is diluted and auctioned to cover bad debt — is credible at scale. Governance capture by a small group of MKR holders making a catastrophically bad risk decision is the other scenario, and it's not hypothetical.

Frax: Any regulatory action targeting the FinresPBC relationship or RWA-backed stablecoins broadly would remove sFRAX's yield source. A loss of FXS market value significant enough to undermine governance participation would also weaken the protocol's ability to respond to a peg attack.

Timing

Now: For DeFi builders choosing a stablecoin integration, the practical choice is usually DAI (deep liquidity, years of Aave/Compound/Uniswap integrations) or FRAX (if you need a yield-bearing stablecoin with a smaller integration surface). They're not direct substitutes.

Next: The Endgame restructuring and Spark Protocol's growth are the things worth watching at MakerDAO over the next 12–18 months. At Frax, the regulatory trajectory for RWA-backed stablecoins.

Later: Whether "decentralized stablecoin" as a category can hold market share against USDC, USDT, and increasingly sophisticated bank-issued digital dollars is the long-horizon question. Neither MakerDAO nor Frax has solved it. Nobody has.

This post covers the mechanisms and architectural differences. It doesn't address tax treatment of DSR yields or sfrxETH rewards, which vary by jurisdiction. It's also not a commentary on whether either protocol is appropriately valued — that requires different analysis than what's here.

The designs are distinct. Whether either represents a durable competitive position depends on factors still unfolding.