Stablecoins in Depth: Reserves, Algorithms and De-Pegging Risk
Analyse how fiat-backed, crypto-backed and algorithmic stablecoins hold their peg, why they de-peg and what MiCA requires in the EU.
What you will learn
- How the three main stablecoin designs try to hold a peg.
- How to read a reserve breakdown and calculate reserve coverage.
- How redemption and arbitrage pull a price back to the peg, and when they fail.
- What the EU's MiCA rules require from stablecoin issuers.
Explanation
The Bitcoin course introduced stablecoins as tokens that aim to track a currency. Here we look at the mechanics behind the peg. A peg is the target price, for example $1.00 or €1.00. A de-peg is a lasting move away from it.
1. Fiat-backed stablecoins
An issuer holds reserves (bank deposits, short-term government bills, repurchase agreements, money-market funds) and promises to redeem tokens for money at par. Key questions:
- Quality: short-dated government bills are safer than commercial paper or loans.
- Concentration: how much sits at one bank?
- Reporting: an attestation is an accountant's check of reserves at one point in time; a full audit examines controls and accounts more broadly. They are not the same.
- Who may redeem: often only verified, large customers. Everyone else sells on the secondary market (exchanges).
Case study: in March 2023 the issuer of USDC disclosed that about $3.3 billion of its reserves were held at Silicon Valley Bank, which had failed. USDC traded around $0.90 for a time over that weekend before recovering when US authorities guaranteed the bank's deposits. Even good reserves carry bank risk.
2. Crypto-collateralised stablecoins
Users lock crypto in a smart contract and mint stablecoins against it, with overcollateralisation: e.g. $150 of ETH for $100 of stablecoins. If the collateral ratio (collateral value ÷ stablecoins minted) falls below a minimum, the position is liquidated (collateral is sold to repay the debt). Risks: fast crashes, oracle errors and, in some designs, heavy reliance on fiat-backed coins as collateral.
3. Algorithmic stablecoins
These try to hold the peg with incentives instead of full reserves, often by letting users swap one stablecoin for $1 worth of a second, volatile token. In May 2022 TerraUSD (UST) lost its peg: as holders redeemed, the protocol minted huge amounts of its sister token LUNA, whose price collapsed, which destroyed confidence further. This feedback loop is called a death spiral. The lesson: a peg backed mainly by the value of the system's own token can fail when confidence fails.
How the peg is defended
If a token trades at $0.98 and redemption at $1.00 works, arbitrageurs buy cheap tokens and redeem them, pushing the price up. Arbitrage fails when redemption is paused, slow, limited to a few firms, or when people doubt the reserves.
MiCA in the EU
The EU Markets in Crypto-Assets Regulation (MiCA) rules on stablecoins have applied since 30 June 2024. It defines e-money tokens (EMTs), which reference one official currency, and asset-referenced tokens (ARTs), which reference other assets or baskets. Issuers need authorisation; EMT holders have a right of redemption at par value at any time; issuers and crypto-asset service providers must not pay interest to holders; and reserves must meet custody, liquidity and investment rules. Details and supervisory lists change, so check the current registers of ESMA (European Securities and Markets Authority) and BaFin. Lending protocols that pay "stablecoin yield" are a different product with different risks (Lesson 6).
Worked example
Hypothetical example: an issuer has 1,000 million tokens outstanding at $1.00. Reserves: $600m government bills, $300m deposits split evenly across three banks, $100m in a money-market fund. Fees and taxes are excluded.
- Reserve coverage = reserves ÷ tokens = $1,000m ÷ $1,000m = 100%.
- One bank fails and $100m is frozen. Available reserves: $900m. Coverage: 90%.
- If the market expects to recover half of the frozen money, the expected value per token = ($900m + $50m) ÷ 1,000m = $0.95.
- Secondary-market prices may trade near this value until redemptions resume.
Hypothetical crypto-backed example: you lock ETH worth $1,500 and mint 1,000 stablecoins. Ratio = 150%. Minimum ratio: 130%. If ETH falls 20%, collateral = $1,200, ratio = 120%, below 130%: liquidation. The ratio hits 130% when collateral is $1,300, i.e. after a fall of ($1,500 − $1,300) ÷ $1,500 = 13.3%.
Common mistake
Practice activity
On paper: (a) A stablecoin trades at $0.97 on an exchange. Redemption at $1.00 is possible with a 0.1% fee. You have access to redemption and buy 10,000 tokens. What is the gross arbitrage profit before gas and trading fees? (b) You mint $1,500 of stablecoins against $3,000 of ETH. Minimum ratio: 150%. How far can ETH fall before liquidation?
Show solution
(a) Cost: 10,000 × $0.97 = $9,700. Redemption: 10,000 × $1.00 = $10,000 − 0.1% fee ($10) = $9,990. Gross profit: $290 (about 3.0% on $9,700).
(b) Liquidation when collateral = 150% × $1,500 = $2,250. Fall: ($3,000 − $2,250) ÷ $3,000 = 25%.
Limitations: in a real de-peg, redemption may be paused or delayed, and the price can fall further while you wait.
Key takeaways
- Fiat-backed coins depend on reserve quality, bank concentration and working redemption.
- Crypto-backed coins depend on overcollateralisation and liquidations.
- Algorithmic designs backed by their own token can enter a death spiral.
- MiCA requires authorised issuers, redemption at par for EMTs and bans interest to holders.
Next step
Next, you will learn how decentralised exchanges let you swap tokens such as stablecoins without an order book.
Self-check
Answer the questions, then check your answers. You can retry as often as you like.
Automatic self-check for your own revision. Your answers are checked in your browser and are not seen or assessed by us. It is not an exam and not a qualification.
Your progress
“Complete” only means you marked the lesson as done. It is not a test of your understanding. Progress and notes are saved only in this browser on this device — not in an account.