A House Built on Code
To understand the survivors, you first have to remember the spectacular failure. TerraUSD (UST) was the poster child for the purely algorithmic model. It wasn’t backed by dollars in a bank, but by a complex relationship with a sister token, LUNA. The
idea was simple: if UST’s price fell below $1, users could burn it to mint $1 worth of LUNA, creating a profitable arbitrage that would push UST’s price back up. If it went above $1, the reverse would happen. For a while, it worked, making UST the third-largest stablecoin. But the system had a fatal flaw: it all relied on faith in LUNA. When a wave of selling hit UST in May 2022, confidence evaporated. As people rushed to exit, the system minted astronomical amounts of LUNA, flooding the market, crashing its price, and sending both tokens into a death spiral that wiped out an estimated $45 billion.
The Myth of the 'Pure' Algorithm
The Terra crash exposed the “real reason” many other algorithmic stablecoins didn’t suffer the same fate: they weren’t really algorithmic in the first place, at least not in the same reckless, uncollateralized way. The survivors were actually hybrids, blending algorithmic mechanics with a concept crypto skeptics can appreciate: actual collateral. The projects that endured the market panic were those that didn't rely on faith alone, but had a backstop of tangible assets to absorb the shock. This marked a quiet but critical evolution away from the purist model toward something more pragmatic and resilient.
Survival Strategy 1: Over-Collateralization
The most prominent survivor is DAI, a stablecoin that is often lumped into the algorithmic category but operates very differently. DAI is better described as a crypto-collateralized stablecoin. To mint DAI, users must lock up other cryptocurrencies, like Ethereum, as collateral in the MakerDAO protocol. Crucially, the system is over-collateralized, meaning for every $1 of DAI created, more than $1 worth of other crypto assets must be deposited. This creates a substantial buffer. If the value of the collateral starts to fall, the protocol automatically sells it off to ensure DAI remains fully backed. While Terra’s founders once criticized this model as not being scalable enough, its resilience during the crash proved that safety, not just speed, was paramount.
Survival Strategy 2: The Hybrid Model
Another successful approach is the fractional-algorithmic model, pioneered by Frax (FRAX). As the name suggests, Frax is only partially stabilized by an algorithm; the other part is backed by hard collateral, similar to fully-reserved stablecoins like USDC. The protocol features a dynamic collateral ratio. When the market has high confidence in FRAX and it trades above $1, the system can reduce the percentage of collateral required. But if the price falls, the protocol increases the collateral ratio, adding more tangible backing to restore stability. This hybrid design gives it more capital efficiency than a fully over-collateralized coin like DAI, but far more security than a purely algorithmic one like UST. Following the Terra collapse, the Frax community even voted to move toward a fully collateralized model, underscoring the market's newfound preference for tangible reserves.
