Why Stablecoins Need Reserves

Stablecoins are claims. Whether it's USDT, USDC, or DAI, holding a stablecoin means holding a promise that this token is worth $1. What makes that promise credible — or exposes it as hollow — is the reserve structure sitting behind it.

This isn't a theoretical concern. TerraUSD was supposed to be worth $1. It collapsed to fractions of a cent in May 2022. The distinction between stablecoins that survived that event and those that didn't comes down to whether there was real backing that could be redeemed.

Understanding why reserves matter is really about understanding how the $1 peg holds mechanically — and what happens when the mechanism breaks.

How the Peg Holds

The basic mechanic is arbitrage. If a stablecoin trades at $0.98 on a secondary market, a rational actor can buy it cheaply and redeem it with the issuer for $1.00. The $0.02 profit per token exists until enough buying pressure restores the peg. This loop keeps the price close to $1 — but only if the redemption step is credible.

Redemption requires reserves.

For fiat-backed stablecoins — USDT, USDC, PYUSD — the issuer holds roughly one dollar in real-world assets for every token in circulation. When you redeem, they send you the dollar and burn the token. Circulating supply contracts. The assets held are typically cash, US Treasury bills, and short-duration government securities: liquid instruments that can be quickly converted if a large redemption wave hits.

Crypto-backed stablecoins use a different mechanism but the same underlying logic. DAI requires more collateral than debt — historically 150% or more. If you want to mint 100 DAI, you deposit at least $150 worth of ETH (or other accepted collateral). The excess absorbs price swings. If your collateral ratio falls below the minimum because ETH dropped, an automated liquidation closes the position before DAI becomes undercollateralized. The reserve surplus is what gives the liquidation mechanism room to operate.

Both approaches ensure each circulating token has a claim on something real.

What Happens Without Real Backing

Algorithmic stablecoins tried to replace reserves with incentive design. TerraUSD (UST) maintained its peg through an arbitrage mechanism with LUNA, Terra's native token. If UST fell below $1, holders could burn UST to mint LUNA at a discount. If UST rose above $1, holders could burn LUNA to mint UST. The mechanism worked while confidence held.

When confidence broke — triggered partly by large-scale withdrawals from Anchor Protocol — the incentive loop ran in reverse. Selling UST required minting LUNA. More LUNA inflated the supply. LUNA's price fell. The collateral backing each UST shrank even as more was being printed. The "reserve" was circular: UST was backed by LUNA, and LUNA's value came primarily from UST adoption.

There was no floor because there was nothing external to sell. No Treasury bills, no ETH, no cash. The peg was an equilibrium that depended on the loop working — and loops can break.

Where the Constraints Live

The hard constraint for fiat-backed stablecoins is the transparency gap. Reserves have to exist, but they also have to be verifiable. Attestations — periodic reviews where an accounting firm confirms the reserve figure — are not the same as audits. An audit involves independent verification of accounting controls and underlying asset claims. Tether publishes attestations. It hasn't completed a full independent audit.

This distinction matters in a stress scenario. USDC depegged briefly in March 2023 when Silicon Valley Bank collapsed. Circle held roughly $3.3 billion of USDC reserves at SVB. The peg recovered quickly — SVB deposits were guaranteed by regulators — but the event demonstrated that reserve quality includes counterparty risk. A dollar in Treasuries is different from a dollar at a bank that might fail.

For crypto-backed stablecoins, the binding constraint is liquidation efficiency. Black Thursday in March 2020 exposed an early DAI vulnerability: ETH dropped ~50% in hours, the mempool clogged, and some auctions cleared at near-zero bids because liquidators couldn't get transactions through in time. Undercollateralized positions weren't closed quickly enough. MakerDAO covered the resulting shortfall through emergency MKR minting — but the episode revealed that overcollateralization only works if the liquidation mechanism actually runs.

What's Changing

Reserve requirements are becoming regulatory requirements, not just design choices.

MiCA in Europe is most advanced. Issuers with more than €5 billion in average outstanding value face requirements around reserve composition, redemption rights at par, and ongoing supervision from the European Banking Authority. Tether announced in early 2025 it wouldn't pursue full MiCA compliance for the EU market. Circle's USDC and EURC are designed to comply.

In the US, the GENIUS Act and STABLE Act have moved through committee stages as of early 2026. Both would require 1:1 reserves in cash and short-duration Treasuries — essentially codifying current USDC practice. If passed, this creates meaningful pressure on issuers with more varied reserve compositions.

On-chain stablecoins are also evolving. MakerDAO's successor protocol (Spark/Sky) holds tokenized US Treasuries as part of DAI's collateral base. The line between fiat-backed and crypto-backed stablecoins is blurring as on-chain systems incorporate the same instruments that back off-chain ones.

Confirmation Signals

US legislation with explicit reserve standards taking effect. Tether publishing a full independent audit. Continued growth of tokenized Treasury-backed stablecoins (BUIDL, OUSG, USDM) as accepted collateral in DeFi protocols. Successful large-scale redemptions during market stress without peg breaks.

Invalidation Signals

A major fiat-backed stablecoin failing to honor redemptions despite apparently adequate reserves — suggesting the real risk is operational or custodial, not the reserve framework itself. A crypto-backed stablecoin failing despite overcollateralization because oracle failures or liquidity crises prevented timely liquidations at scale.

Timing

Now: Reserve quality is a live differentiator. Tether and USDC are not the same product — this distinction is increasingly reflected in institutional usage patterns. USDC is dominant in regulated contexts.

Next: US and EU regulatory frameworks will formalize reserve standards within 12–24 months. Compliance costs will reshape who can issue stablecoins at scale.

Later: Real-time, on-chain proof of reserves remains an open technical problem. Periodic attestations are a floor, not a ceiling.

What This Doesn't Cover

This post explains the reserve mechanics. It doesn't assess the safety of any specific stablecoin or recommend one product over another. Reserve quality is one factor in stablecoin risk; custodial structure, regulatory exposure, smart contract risk, and counterparty risk are others. The mechanism described here applies under normal market conditions — extreme events stress any reserve structure in ways this analysis doesn't fully capture.