Algorithmic stablecoins are designed to track a target price, usually one US dollar, without holding the full set of cash-like reserves that back traditional stablecoins. Instead, they rely on code-driven incentives that expand or shrink supply, encourage arbitrage, or use a second token to absorb volatility. In calm markets, some designs can look surprisingly stable. In stressed conditions, however, the same mechanics can unravel quickly, especially when confidence drops and everyone tries to exit at once.
This article explains how algorithmic stablecoins work in practice, the main models you’ll see, and the typical reasons they fail.
What is an algorithmic stablecoin?
A stablecoin aims to keep a steady value by referencing something external, such as a fiat currency. Most well-known stablecoins try to hold their peg using reserves (for example, cash and short-dated government debt) or over-collateralised crypto assets locked in smart contracts.
An algorithmic stablecoin takes a different approach. It tries to maintain a target price mainly through programmed market incentives rather than relying on a fully redeemable reserve of low-risk assets. These incentives usually aim to do one of two things:
- Increase supply when the price is above the target, pushing it down.
- Decrease supply when the price is below the target, pushing it up.
That sounds straightforward, but it assumes that markets will keep cooperating during volatility. That assumption is where many failures begin.
The core mechanisms that try to hold the peg
Supply expansion and contraction
Most algorithmic models treat the stablecoin like a system that can “print” or “remove” coins to steer the market price. If demand is high and the token trades above the target, the system increases supply. If demand falls and the token trades below the target, it tries to reduce supply by rewarding people who remove coins from circulation.
Arbitrage incentives
The stabilisation logic usually depends on arbitrageurs. When the stablecoin is below the target, the system offers a profit opportunity (for example, buy discounted stablecoins and redeem them for something valued at the target). When it’s above the target, it offers a different profit opportunity (for example, mint new stablecoins at the target value and sell them for more).
If arbitrage is slow, expensive, or risky, the peg can drift. If arbitrage becomes impossible, the peg can break.
Oracles and price references
To know whether it should expand or contract supply, the system needs price data. That often comes from oracles pulling in market prices from exchanges. If oracle data lags, is manipulated, or becomes unreliable during chaos, the stabilisation mechanism can react too late or in the wrong direction.
Common types of algorithmic stablecoins
Dual-token mint and burn models
This model uses two assets:
- The stablecoin that targets the peg.
- A second token that absorbs volatility and acts as the balancing asset.
In simple terms, the system allows conversions between the stablecoin and the second token at a reference value near the target. When the stablecoin falls below the peg, holders are encouraged to swap it for the second token, reducing stablecoin supply. When the stablecoin rises above the peg, the system encourages minting more stablecoins, increasing supply.
Bond or coupon models
Some designs try to reduce supply when the stablecoin is below the target by issuing “bonds” or “coupons”. Users give up stablecoins now (removing them from circulation) and receive a claim that can be redeemed later when the peg is restored, typically at a profit.
This only works if the market believes the peg will return. When confidence is low, people don’t want future promises.
Rebase models
Rebase stablecoins adjust the number of tokens in holders’ wallets according to a rule. If the price is above the target, balances increase; if it’s below, balances decrease. The idea is that changing supply changes the price.
Rebase systems can keep a price reference, but they often fail the everyday expectation people have of stablecoins: that one unit should stay one unit. Even if the price behaves, your balance can change.
Hybrid and partially collateralised designs
Some projects describe themselves as “algorithmic” while also holding reserves or using over-collateralised crypto. In practice, many systems that survived longer did so by leaning more heavily on collateral and less on pure incentives, especially after major market shocks.
Why algorithmic stablecoins fail
Most failures trace back to one problem: the peg relies on confidence and continuous market participation. When those disappear, the mechanism can turn from stabilising to self-destructive.
They can behave like a bank run without a lender of last resort
If many holders try to exit at once, the system needs deep liquidity and willing buyers. Unlike a bank, there is no central backstop. Without credible reserves, the stablecoin can only offer incentives to stay in the system or swap into something else that is also losing value.
Reflexive “death spirals” in dual-token systems
In dual-token designs, the second token often needs to keep a meaningful market value for the peg mechanism to work. In a crisis, the system may mint large amounts of the second token to meet redemptions or incentives. That extra supply can crash the second token’s price, which then makes the stablecoin look even less credible, causing more selling. This feedback loop is the classic death spiral.
Liquidity disappears when it’s needed most
Pegs are easiest to defend when trading is active and spreads are tight. In stressed markets, liquidity on exchanges can thin out, slippage increases, and large traders can move the price sharply. A design that looks stable at small scale may fail abruptly at larger scale because the exit door is not big enough.
Incentives depend on belief, not guarantees
Bonds, future redemption promises, and “eventual recovery” narratives require buyers to trust that the system will survive long enough to pay out. When confidence breaks, promised future value is discounted heavily or ignored entirely.
Oracle and governance risks can amplify shocks
If price feeds are attacked or lag heavily, supply adjustments can misfire. Separately, governance actions taken during a crisis can worsen panic if they appear improvised, opaque, or unfair to certain holders.
A real-world failure pattern: how a peg can collapse in days
A common pattern looks like this:
- Initial shock: a sell-off, liquidity event, or loss of confidence pushes the stablecoin slightly below the target.
- Arbitrage slows: traders hesitate because the swap mechanism or exit route feels risky.
- Redemptions surge: more holders sell or redeem, increasing pressure on the system.
- Volatility token collapses: if a second token is involved, its price falls as supply expands.
- Feedback loop: the falling second token undermines confidence further, accelerating the run.
The most famous example of this dynamic wiped out tens of billions of dollars in value within about a week and triggered broader market stress. The details vary by project, but the failure rhythm is often similar.
How UK regulation frames stablecoins, and where algorithmic models sit
In the UK, the regulatory focus has largely been on stablecoins that are backed by assets and used for payments, alongside wider cryptoasset regulation that is being rolled out in stages. That matters because many algorithmic stablecoins are designed specifically to avoid holding full reserves, which makes them harder to fit into frameworks built around redemption, backing assets, custody standards, and consumer protections.
Separately, UK regulators have also discussed limits and safeguards for stablecoins used at scale in payments, reflecting concerns about financial stability if large sums move quickly between bank deposits and stablecoin arrangements. Even when a stablecoin is designed for everyday use, the policy direction has tended to favour models with clear backing, predictable redemption, and strong risk management.
Common misconception: “If it’s algorithmic, the peg is automatic”
A frequent misunderstanding is that an algorithmic stablecoin is stable in the same way a thermostat keeps a room at a set temperature. In reality, the algorithm usually does not control the price directly. It changes incentives and supply rules and hopes the market responds in the intended way.
If traders don’t trust the system, if liquidity dries up, or if the stabilising trade becomes too risky, the algorithm cannot force the peg back. It can only keep adjusting rules, which may be ineffective or even fuel panic.
Practical checks before you trust a “stable” price
If you’re assessing a stablecoin that claims algorithmic stability, focus on the exit mechanics and stress behaviour, not just the chart during normal times:
- Redemption path: Can you redeem for something with dependable value, or only swap into another volatile token?
- Quality of backing: Is there any collateral, and if so, is it liquid and transparently managed?
- Liquidity depth: How much can realistically be sold or redeemed without severe slippage?
- Incentive credibility: If the system issues bonds or future claims, why would buyers want them during a panic?
- Concentration risk: Would a few large holders or market makers have an outsized ability to break the peg?
- Oracle robustness: What price feeds drive the mechanism, and what happens if feeds lag or fail?
In short, stability claims should be judged by worst-case behaviour: what happens when everyone wants out at the same time.
Conclusion
Algorithmic stablecoins attempt to hold a peg using programmed incentives rather than straightforward, high-quality reserves. The main designs rely on supply adjustments, arbitrage, and often a second token that absorbs volatility. These systems can appear stable in favourable conditions but are vulnerable to runs, liquidity collapse, oracle issues, and reflexive feedback loops that can drive rapid failure.
If you’re evaluating any stablecoin described as “algorithmic”, the key question is not how it behaves in calm markets, but whether its redemption and liquidity structure can survive stress without relying on confidence alone.