Algorithmic Stablecoin

How did UST collapse? Explain the death spiral in the simplest possible way.

UST's collapse is a textbook 'death spiral' case. Three steps:

Step 1: Trigger. In early May 2022, large holders began mass-selling UST (later confirmed as deleveraging by a major fund), causing UST's market price to drop below $1 — a slight depeg.

Step 2: Mechanism response. To get UST back to $1, the protocol's arbitrage mechanism activated: allowing holders to swap $1 of UST for equivalent LUNA. This meant large amounts of LUNA were minted, LUNA's market supply surged, and prices began falling.

Step 3: Spiral accelerates. LUNA's price decline caused the market to lose confidence in the entire Terra ecosystem. More UST holders rushed to swap UST for LUNA then sell, LUNA continued minting and falling, UST's peg collapsed further. This loop repeated countless times within days, until LUNA fell from approximately $80 to near zero, and UST followed to zero.

Core lesson: this mechanism is self-reinforcing during market panic — the code itself becomes an accelerator in crisis, not a stabilizer. Systems without external real assets to 'backstop' face total loss when confidence collapses.

What are the main differences between algorithmic stablecoins and fiat-backed (USDC) or crypto-backed (DAI) types?

The most fundamental difference: 'if confidence collapses, what's left?'

USDC (fiat-backed): if Circle and USDC lose market confidence, holders can demand 1:1 redemption of real dollars from Circle (as long as reserves are sufficient). Confidence collapse doesn't mean total loss because real assets backstop.

DAI (crypto-backed): if DAI loses confidence, holders can use DAI to trigger collateral liquidation and recover some crypto assets (though potentially with liquidation losses). Real assets also backstop here — just volatile crypto assets.

UST (algorithmic): if UST loses confidence, holders can swap for LUNA, but LUNA's value is built on market confidence in the entire Terra ecosystem. Confidence collapse → LUNA minting → LUNA depreciation → more people lose confidence → cycle. No real assets independent of 'confidence' to backstop — result is zero.

One-sentence summary: fiat-backed and crypto-backed types' floor is 'real assets'; algorithmic types' floor is 'market confidence' — and confidence is the most unreliable floor.

After UST's collapse, did the industry develop safer algorithmic stablecoin designs? Are there still such attempts?

Yes, but the design direction fundamentally shifted: from 'no reserves needed' to 'let algorithms optimize efficiency on top of real reserves.'

FRAX's evolution is the most representative case: FRAX was originally 'partial algorithmic + partial USDC reserves.' Post-UST collapse, FRAX governance voted to shift to 100% real reserves. FRAX abandoned the 'no reserve' algorithmic component, retaining algorithms' role in reserve management and rate adjustment.

Currently ongoing directions: stablecoins with tokenized real-world assets (RWA) as reserve bases — if underlying reserves are low-volatility assets (like Treasuries), algorithms can more safely manage lower overcollateralization ratios approaching 1:1 efficiency while still having real asset backstops. These designs seek better positions within the trilemma triangle rather than claiming to break it.

Regulatory impact: MiCA and GENIUS Act require 1:1 real reserves, directly closing off 'zero-reserve pure algorithmic' viability in mainstream compliant markets. If these regulations take effect, truly 'algorithmic stablecoins' can only exist in regulated markets as 'academic experiments' or 'DeFi edge scenarios.'

If someone tells me their new stablecoin 'has no reserves but is very safe,' how should I evaluate it?

Any stablecoin claiming 'no reserves but stable' deserves evaluation through these questions:

Question 1: If everyone wants to redeem simultaneously, where does your money come from? If the answer is 'other buyers in the market' or 'algorithmically minted paired tokens,' this is an algorithmic design — no backstop mechanism exists independent of market confidence. Historically, all such designs have failed under stress.

Question 2: Can this stablecoin exist within MiCA or GENIUS Act compliance frameworks? If not, it has no long-term space in mainstream markets, only operating where regulation is weak. This itself is a risk signal.

Question 3: Why does it claim to be safer than UST? UST's core failure was 'confidence collapse causing vicious cycles' — if the new design hasn't fundamentally addressed this (introducing real reserves or external liquidation support), just adding complex code layers to the same problem isn't a substantive improvement.

Conclusion: without sufficient market stress test records and transparent reserve mechanisms, treating 'zero-reserve stablecoins' as a safe place for primary capital is unreasonable. Trying them as a small proportion position you can fully accept losing is more rational.