Author: James Wright | Stablecoin Researcher | Former Quantitative Analyst at Bridgewater Associates

Terra's collapse in May 2022 wiped out $40 billion in value and killed algorithmic stablecoins. At least, that's what everyone believed. Three years later, the category is quietly rebuilding — with new mechanisms, harder questions, and maybe some actual answers.

The stablecoin landscape today

Let's establish the baseline. Stablecoins now represent over $160 billion in market cap. They're the most successful crypto use case — actual utility, not speculation. People use them for remittances, trading, savings in unstable economies.

The market is dominated by collateralized stablecoins. Tether holds roughly $115 billion. Circle's USDC sits around $35 billion. Together they control over 90% of the market.

The model is simple: for every stablecoin issued, equivalent dollars sit in a bank account. You trust the issuer to maintain reserves. You trust auditors to verify them. You trust banks to not fail.

It works. But the trust assumptions are significant.

Why algorithmic stablecoins mattered

The original promise was elegant: a stablecoin that maintains its peg through code, not collateral. No banks, no reserves, no counterparty risk. Pure decentralized money.

Terra's UST attempted this with a dual-token model. UST could always be redeemed for $1 worth of LUNA. Arbitrage would maintain the peg. When UST traded below $1, traders would buy it and redeem for LUNA profit. Supply decreases, price rises. Elegant in theory.

The flaw was reflexivity. In a crisis, everyone redeems simultaneously. LUNA supply explodes. LUNA price crashes. The collateral backing redemptions becomes worthless. Death spiral.

$40 billion learned this lesson the hard way.

The new generation

Post-Terra builders didn't give up. They got smarter.

Ethena's USDe is the most interesting experiment. It's not purely algorithmic — it uses delta-neutral hedging. The protocol holds crypto assets and shorts equivalent perpetual futures. The funding rates from shorts generate yield. The hedge eliminates directional exposure.

It's clever. In bull markets with positive funding, yields can reach 20-30%. The peg is maintained not by faith in redemption but by actual market-neutral positions.

The risks are different. Exchange counterparty risk — what if the exchange holding your short positions fails? Funding rate inversion — what happens when funding goes negative for extended periods? These are real concerns, but they're not death spiral risks.

Frax evolved differently. It started algorithmic, then progressively added collateral. Today it's nearly fully collateralized but uses protocol-controlled value to optimize capital efficiency. Hybrid approach — pragmatism over purity.

The overcollateralized middle ground

Between fully backed and algorithmic sits a category that's quietly won: overcollateralized crypto-backed stablecoins.

MakerDAO's DAI is the prototype. Deposit $150 of ETH, mint $100 of DAI. The excess collateral absorbs volatility. If collateral value drops, positions get liquidated before the system becomes undercollateralized.

It's capital inefficient. You need more collateral than the stablecoins you create. But it's resilient. DAI survived every market crash since 2017. No death spiral. No bank runs. Just math enforced by smart contracts.

Liquity's LUSD pushed this further — no governance, immutable contracts, only ETH collateral. Maximum decentralization at the cost of flexibility. It works, but it's never scaled beyond a few billion.

What I actually think about each model

Centralized stablecoins will dominate for practical reasons. Institutions need them. Regulators understand them. The trust assumptions are familiar — we trust banks already.

But centralized means controllable. Tether can freeze addresses. USDC has blacklisted wallets under government pressure. For censorship-resistant finance, they're insufficient.

Overcollateralized stablecoins fill the decentralization gap. DAI and its successors offer genuine censorship resistance. The tradeoff is capital efficiency — you lock up more value than you create. For DeFi-native users, it's acceptable. For mainstream adoption, it's friction.

Pure algorithmic stablecoins probably can't work at scale. Every model requires some source of external value to maintain peg in crisis. The question is just how that value is structured and what happens when it's tested.

Hybrid models like Ethena and Frax are the frontier. They use real mechanisms — hedging, yield, collateral buffers — rather than faith in reflexive redemption. Whether they survive the next crisis remains unproven.

The regulatory shadow

Everything I've discussed exists under regulatory uncertainty. US lawmakers want stablecoin legislation. EU's MiCA imposes reserve requirements. Both favor traditional, bank-like models.

Algorithmic and crypto-collateralized stablecoins don't fit regulatory categories. They might get banned outright. They might get pushed offshore. The innovation is happening, but the legal ground shifts beneath it.

My prediction: regulated fiat-backed stablecoins will dominate compliant markets. Everything else will thrive in permissionless DeFi, serving users who prioritize censorship resistance over regulatory clarity.

Two parallel systems, serving different needs. That's probably the future.

The metrics that matter

If you're evaluating stablecoins, here's what I look at.

Peg stability during stress. Not average deviation — maximum deviation during market crashes. Any stablecoin can hold peg in calm markets.

Redemption mechanics. Can you actually get your dollar back? How fast? At what cost? Theoretical redeemability means nothing if practical redemption fails.

Reserve transparency. For collateralized coins, how often are reserves verified? By whom? Can you see them on-chain or do you trust attestations?

Centralization vectors. Who can freeze your funds? Who can change the rules? What jurisdictions have power over the system?

No stablecoin scores perfectly on all dimensions. Understand the tradeoffs, choose accordingly.

James Wright researches stablecoin mechanisms and monetary policy in crypto. He previously built quantitative models at Bridgewater Associates and holds a PhD in Economics from MIT.