Crypto Gloss

Stablecoin Design

Stablecoins are among the most important and most dangerous instruments in crypto. A well-designed stablecoin enables DeFi, facilitates global payments, provides inflation hedges, and allows traders to hold value without leaving crypto. A poorly designed stablecoin — as TerraUST demonstrated in May 2022 — can collapse catastrophically, wiping out tens of billions in market cap in days and triggering systemic contagion across the entire crypto ecosystem. Understanding the design architectures underlying different stablecoins explains both their utility and their risk profiles. There is no free lunch: every stablecoin design trades capital efficiency, decentralization, robustness, and trust in different ways.

The Stablecoin Trilemma

Before examining individual designs, the stablecoin trilemma frames the fundamental tradeoffs:

  1. Capital efficiency — Can you hold $1 of stablecoin for less than $1 of collateral?
  2. Decentralization — Is the peg maintained without centralized custodians?
  3. Stability/Robustness — Can the peg survive extreme market conditions?

No design fully achieves all three. Each architecture sacrifices at least one dimension.

Design 1: Fiat-Backed (Custodial)

Examples: USDT (Tether), USDC (Circle), FDUSD, PYUSD

How it works:

  1. User deposits $1 USD with a custodian (Tether, Circle)
  2. Custodian mints 1 stablecoin token
  3. Custodian holds the $1 in a bank account or short-term US Treasuries
  4. User burns the stablecoin to redeem $1

Collateral ratios: ~100% (or close to it)

Pros:

  • Perfect capital efficiency (1:1 ratio)
  • Most stable — directly redeemable for dollars
  • Simplest user experience

Cons:

  • Fully centralized — custodian can freeze accounts, face regulation, or fail
  • Counterparty risk — Tether’s reserve composition has been repeatedly questioned
  • Circle’s USDC depeg in March 2023 (when SVB bank held Circle reserves — USDC briefly depegged to $0.87 before US government backstopped SVB)

Risk profile: Low volatility, high centralization risk. “Bank run risk” exists if everyone tries to redeem at once with insufficient liquid reserves.

Design 2: Crypto Overcollateralized

Examples: DAI (MakerDAO), LUSD (Liquity), crvUSD (Curve)

How it works:

  1. User deposits crypto (ETH, WBTC) as collateral in a smart contract vault
  2. Smart contract mints stablecoin at 150-200% collateralization ratio
  3. If collateral falls below minimum ratio → liquidation (collateral sold to repay debt)
  4. Burning stablecoin + paying fee returns collateral

Collateral ratios: 150-200%

Pros:

  • No custodian — smart contract holds collateral
  • Transparent — anyone can verify reserves on-chain
  • Censorship resistant (soft version — MakerDAO has since added USDC collateral)

Cons:

  • Capital inefficient — must lock $1.50 of ETH to mint $1 of DAI
  • Liquidation risk — if ETH drops quickly, vaults liquidate, suppressing price further
  • Black swan risk — in extreme crashes (March 2020 “Black Thursday”), liquidation mechanisms failed and MakerDAO accrued bad debt

Risk profile: Medium stability, lower centralization risk, economically inefficient.

Liquity (LUSD) innovation:

Liquity improved on MakerDAO’s design with:

  • No governance (fully algorithmic parameters)
  • 110% minimum collateral ratio (more capital efficient)
  • Stability pool mechanism: LUSD holders earn from liquidations
  • One-time fee model (no ongoing interest)

Design 3: Algorithmic (Uncollateralized)

Examples: Terra UST (failed), Basis (failed), IRON/TITAN (failed)

How it works (Terra model):

  1. $1 of LUNA can always be burned to mint 1 UST
  2. If UST > $1: arbitrageurs burn LUNA → mint UST → sell → profit → brings UST down
  3. If UST < $1: arbitrageurs buy UST → burn → mint LUNA → sell → profit → brings UST up
  4. Mechanism is self-reinforcing on the way up, self-destructive on the way down

Collateral ratios: 0% (reflexive only)

Why they fail:

The algorithmic design has a fatal flaw: it depends on the value of the volatile asset (LUNA) to stabilize the stablecoin (UST). In a death spiral:

  • UST depegs → arbitrageurs mint LUNA → LUNA price falls → less collateral value → UST depegs more → repeat
  • “Bank run in a funhouse mirror” — the more the stablecoin falls, the more the “collateral” falls

Terra collapse (May 2022):

$40B UST market cap went to near-zero in ~72 hours. LUNA fell from $80 to $0.0001. Adjacent stablecoins, lending protocols, and VC funds suffered billions in losses.

Rule: Pure algorithmic stablecoins with reflexive collateral are considered largely unworkable. The Basis, Iron Finance, Terra failures represent multiple attempts with the same outcome.

Design 4: Delta-Neutral (Synthetic)

Examples: USDe (Ethena), GHO (Aave, partially)

How it works (Ethena USDe):

  1. User deposits ETH (or stETH)
  2. Ethena opens a short ETH perpetual position of equivalent size on a derivatives exchange
  3. Result: Long ETH (spot) + Short ETH (perp) = delta-neutral position worth ~$1 regardless of ETH price
  4. Ethena mints 1 USDe

Yield mechanism:

  • stETH generates staking yield (~3-5% APY)
  • Short ETH perp earns funding rate when market is in positive funding (perp buyers pay shorts — bullish markets)
  • Combined yield often >10% APY — passed to USDe stakers as sUSDE

Pros:

  • Capital efficient (near 1:1 ratio)
  • Yield-generating
  • No bank account risk

Cons:

  • Funding rate risk: in bear markets, funding goes negative (shorts pay longs) — yield disappears or USDe costs fees
  • Counterparty risk on centralized exchanges holding the short positions
  • Complex operational risk (Ethena must manage positions across multiple exchanges)

Risk profile: Innovative but with structural CEX dependency and funding rate cycle exposure.

Design 5: RWA-Backed

Examples: ONDO USDY, Maker’s DAI (partially), Mountain USDM

How it works:

  1. Issuer buys short-term US Treasuries or money market instruments
  2. Mints stablecoin backed by these real-world assets
  3. Yield from Treasuries distributed to stablecoin holders

Pros:

  • Low risk (Treasury-backed)
  • Yield-bearing (Treasury rates = ~4-5% in 2024)

Cons:

  • KYC/accredited investor requirements (US regulatory constraints)
  • Not fully permissionless — access may be limited to certain geographies/entity types

Comparative Summary

Design Capital Efficiency Decentralization Stability Example
Fiat-backed ★★★★★ ★ ★★★★★ USDC, USDT
Overcollateralized ★★ ★★★★ ★★★★ DAI, LUSD
Algorithmic ★★★★★ ★★★★★ ★ UST (failed)
Delta-neutral ★★★★ ★★★ ★★★ USDe
RWA-backed ★★★★ ★★ ★★★★★ USDY

Related Terms

Sources

Eichengreen, B. (2019). From Commodity to Fiat and Now to Crypto: What Does History Tell Us? NBER Working Paper 25426.

Clements, R. (2021). Built to Fail: The Inherent Fragility of Algorithmic Stablecoins. Wake Forest Law Review, 11.

Moin, A., Sekniqi, K., & Sirer, E. G. (2020). SoK: A Classification Framework for Stablecoin Designs. FC 2020.

He, D., et al. (2022). Stablecoins: Risks, Potential and Regulation. BIS Working Papers No. 905.

Qin, K., Celik, E., & Gervais, A. (2023). Towards a Theory of Maximal Extractable Value III: Stablecoins Under Adversarial Conditions. arXiv.