DAI, UST, FRAX — Three Stablecoins. Three Different Answers to the Same Problem.
Every stablecoin is trying to solve one thing: how do you keep a token worth exactly $1 without a central bank backing it?
Three protocols answered that question differently. One collapsed. Two survived. Here’s why.
1/ The problem with stablecoins
A stablecoin needs to hold its $1 peg even when markets go crazy. That’s easy if you just hold $1 in a bank — but then you’ve rebuilt the traditional financial system inside crypto. Pointless.
The real challenge is: how do you maintain the peg in a decentralized way, without depending on banks or governments?
Three different philosophies emerged.
2/ UST — Pure algorithm. No collateral.
Terra’s UST maintained its peg through an exchange mechanism with LUNA. $1 of UST could always be swapped for $1 worth of LUNA, and vice versa.
When UST dropped below $1 → arbitrageurs bought UST cheap → swapped for LUNA → sold LUNA → UST supply decreased → price recovered.
Elegant in theory. Fatal in practice.
The problem: LUNA was Terra’s own token. When UST lost confidence, people sold UST → minted LUNA to cover → LUNA supply exploded → LUNA price crashed → UST collateral worth less → more UST sold.
A circular structure. Once confidence broke, there was nothing to stop the spiral.
$40B gone in a week.
3/ DAI — Overcollateralized. External assets.
MakerDAO took the opposite approach. To mint $100 DAI, you need to deposit $150+ of ETH. Real, external collateral — not a token MakerDAO controls.
If ETH crashes, there’s still buffer before DAI becomes undercollateralized. Automated liquidations kick in before the system breaks.
And if the system takes a loss anyway? MKR token holders are the insurer of last resort. The protocol mints new MKR and sells it to cover the gap — diluting MKR holders in the process.
Get 0x365home’s stories in your inbox
Join Medium for free to get updates from this writer.
MKR holders accept that risk in exchange for the upside: when the system runs well, protocol fees are used to buy back and burn MKR — increasing its value.
The tradeoff: overcollateralization means capital inefficiency. You lock up $150 to borrow $100. A lot of value sitting idle.
4/ MakerDAO’s recent evolution — RWA
To solve the capital efficiency problem, MakerDAO started accepting US Treasury bonds as collateral.
Treasuries earn yield (~4–5%). ETH just sits there. Suddenly the protocol has real revenue. MakerDAO’s income exploded after RWA adoption.
But there’s a catch — and it’s a fundamental one.
If the US government sanctions MakerDAO, the Treasury collateral gets frozen. A supposedly decentralized protocol becomes dependent on Washington’s goodwill.
That’s why MakerDAO rebranded to Sky. They’re wrestling with this tension out loud.
5/ FRAX — The middle path
Frax looked at both extremes and built something between them.
To mint $1 FRAX: $0.85 USDC (real collateral) + $0.15 FXS (Frax’s own token).
Not 100% algorithmic. Not 100% collateralized. The ratio isn’t fixed either — it adjusts automatically based on market conditions. When FRAX is stable and demand is high, the collateral ratio drops. When it’s unstable, the ratio rises. More conservative, more capital returned.
The excess USDC doesn’t just sit there either. It gets deployed into Aave, Curve, and other protocols to earn yield — making Frax a revenue-generating machine, not just a stablecoin issuer.
6/ The core lesson
All three were trying to do the same thing. What separated them was one question:
What happens when confidence breaks?
UST → nothing to fall back on. Circular collapse. DAI → overcollateralized buffer + MKR backstop. Survived. FRAX → real buffer + yield-generating collateral + auto-adjusting ratio. Survived.
The lesson isn’t that algorithmic stablecoins are impossible. It’s that stablecoins without real external collateral will always be one confidence crisis away from zero.
MakerDAO and Frax both understood that. UST didn’t.
