Types of stablecoins explained: redemption, reserves, and what breaks under stress

Types of stablecoins explained starts with the trade each coin offers: a par-like, on-demand claim that only holds if redemptions clear under stress. The real divider is the redemption path and the reserve or stabilization mechanism behind the peg, because that structure can create run risk and even spill over into money markets.

Key Takeaways

• Stablecoin “types” are best compared by redemption rights and reserve composition, not by a marketing label like “USD stablecoin.”
• Reserve-backed designs perform liquidity transformation when demandable coins are backed by a mix of cash and less-liquid short-dated government bonds, which can create run dynamics.
• BIS research flags two stress channels: holder losses if fire sales erode issuer capital, and broader money-market spillovers if forced bond selling pushes prices down.
• Policy debates focus on liquidity requirements and capital requirements, and BIS modeling finds “usable buffers” can reduce risk while strict minimums can backfire.

How stablecoins claim price stability

A stablecoin’s stability claim starts with a reference asset and a promise about convertibility. The reference is usually a fiat currency, and the operational goal is keeping the market price near a peg, typically 1.00 of that currency. That is the core intuition behind what is a stablecoin as a category, but it is not a guarantee of how the coin behaves when redemptions accelerate.

The useful way to classify stablecoin categories is by the liability they sell to holders. Many coins are effectively offering something that looks like “redeemable at par, on demand,” even if most holders never touch the issuer. Once that promise exists, the next question is mechanical: who can actually redeem, through what rails, and at what price. If only a narrow set of counterparties can redeem directly, everyone else is relying on secondary-market liquidity to stay close to the peg.

This framing matters because stablecoins are overwhelmingly dollar-referenced. BIS Papers No 170 finds roughly 98% of stablecoins’ value is dollar-denominated, which implies stablecoins are more likely to reinforce existing currency hierarchies than to compete with them at the outset. For a user, that shifts the first-order question away from FX drift and toward the plumbing of a private dollar substitute: when the market wants out, does the redemption channel absorb it cleanly, or does the coin trade like a stressed credit instrument.

The “4 types of stablecoins” shorthand seen in many guides is not wrong, but it is incomplete unless it is tied back to the redemption promise. A fiat backed stablecoin, a crypto backed stablecoin, and an algorithmic stablecoin can all print a $1 quote on a screen. Only some designs can reliably turn that quote into cash at scale when flows reverse.

Reserve-backed stablecoins and redemption

Reserve-backed coins live or die on a simple sequence that looks boring until it is crowded. The issuer takes in dollars (or dollar equivalents), issues tokens, invests the reserve portfolio, and stands ready to meet redemptions at par for eligible redeemers. The stability mechanism is not the peg chart. It is the issuer’s ability to convert reserve assets into cash fast enough to pay out.

That is where the “1:1 backed” story gets people into trouble. BIS Working Paper No 1355 highlights that large issuers hold sizeable portfolios of short-dated government bonds. That portfolio choice creates liquidity transformation: the issuer offers demandable coins subject to volatile redemptions while holding a mix of cash and less-liquid bonds. The mismatch is the product. It is also the stress point.

Collateralization is the hinge concept here, but it needs to be read as asset quality plus asset liquidity. Cash is immediately usable for redemptions. Short-dated government bonds are high quality, but they are not cash, and they can become a fire-sale problem if many holders want out at once. Under heavy outflows, the issuer may have to sell bonds into a falling market to raise cash. If those sales happen at a discount, the reserve can take losses.

Redemption access determines how quickly that stress transmits to the market price. If only a small set of institutions can redeem directly, the broader market’s “redemption” is effectively selling the token to someone else. That can keep the peg tight in calm conditions and then gap it when balance sheets step back. The key comparison inside “fiat backed vs crypto backed” is not just the collateral type. It is whether the design has a credible, scalable par exit when everyone wants the same door.

Non-reserve stabilization designs and trade-offs

Once a coin stops relying primarily on an issuer-managed reserve portfolio, the stabilization problem shifts from asset management to mechanism design. A crypto backed stablecoin typically uses on-chain collateral and rules that aim to keep the system solvent even if the collateral price moves. The stability claim comes from overcollateralization, liquidation mechanics, and incentives that try to keep the system’s liabilities covered.

The trade-off is that the peg is now mediated by market structure and liquidation capacity. If collateral prices fall quickly and liquidations cannot clear without slippage, the system can drift away from par even if it is technically solvent on paper. The stability is only as good as the ability to convert collateral into the reference value during a rush.

An algorithmic stablecoin pushes further. Instead of leaning on hard collateral, it leans on rules that attempt to manage supply and demand, often by minting and burning tokens or using a second token as a shock absorber. BIS work on stablecoins’ risks and on identifying “real” stability emphasizes that the robustness of the stability claim depends on what stands behind it when confidence breaks. If the mechanism requires continuous market demand for a reflexive asset, stress can turn the stabilizer into an accelerant.

These designs are often presented as clean categories, but the more accurate map is a spectrum of redemption credibility. Reserve-backed coins promise par redemption but can be runnable if reserves are not cash-like. Crypto-backed coins can be transparent and rules-based but depend on liquidation liquidity. Algorithmic designs can look stable until the market tests the backstop. The labels matter less than the loss-absorption path when the peg is challenged.

Key risks by stablecoin type

The BIS lens that matters for users is not “does it usually trade at $1.” It is how losses show up when redemptions surge. BIS Working Paper No 1355 spells out two distinct stress channels for reserve-backed issuers with cash-plus-bonds portfolios: default risk for coin-holders if bond fire sales erode capital, and spillovers to money markets if forced sales depress bond prices.

A practical comparison framework starts with redemption and then works outward:

1. Redemption rights and bottlenecks. Who can redeem directly with the issuer, and who is forced into secondary markets when the peg wobbles. 2. Reserve or collateral liquidity. Cash behaves differently from short-dated government bonds during a fast outflow, and on-chain collateral behaves differently again when liquidation liquidity is thin. 3. Loss allocation. If assets are sold below par, which balance sheet eats the loss: issuer equity, a backstop fund, a second token, or the holder via a broken peg. 4. Spillover channel. Reserve-backed designs can transmit stress into government-bond markets if the reserve is large enough and selling is concentrated, which is why stablecoin risk is not automatically contained inside crypto.

BIS modeling also adds an uncomfortable point about incentives. In the absence of regulation, issuers tend to hold little capital and favor interest-bearing but less-liquid bonds over cash. That is rational from the issuer’s perspective because bonds earn yield. For holders, it means the coin’s “stability” is partly a bet on the issuer’s liquidity management under pressure.

This is where the beginner mistake happens: treating “USD stablecoin” as dollar-like safety. With roughly 98% of value dollar-denominated, the dominant exposure is not currency. It is the structure of a private dollar claim that can face a run.

How regulation targets stablecoin stability

Regulators are not only arguing about whether reserves exist. They are arguing about how much of the reserve must be immediately liquid and how much loss-absorbing capital the issuer must fund itself with. BIS Working Paper No 1355 frames the toolkit around liquidity requirements, which set minimum liquid assets, and capital requirements, which set minimum equity capital.

The design detail that matters is whether requirements behave like hard floors or like buffers that can be used in stress. BIS modeling finds that liquidity and capital thresholds reduce default and spillover risks when designed as usable buffers. The mechanism is incentive-based: the buffer can be breached during stress, but the breach triggers dynamics that discipline issuers and encourage buffer-building in normal times. The same work finds strict minimum requirements can backfire by forcing premature bond sales, which can raise default risk.

The channels are not identical. In the BIS model, a liquidity threshold mainly pushes issuers toward holding more cash. A capital threshold pushes issuers toward holding both more capital and more cash. Both tools reduce both default and spillover risks, which makes them substitutes in one dimension. They are also complements for policymakers trying to target both holder protection and market spillovers at once.

For users, the regulatory point is simple: “more backing” is not a complete answer if the backing is not usable under stress. The stablecoin design that looks safest in calm markets is not necessarily the one that clears redemptions cleanly when everyone wants par at the same time. That is the stability question at the heart of what is a stablecoin as a payments-like instrument.

The Take

I’ve watched traders treat a tight peg as a safety signal and ignore the only question that matters when the tape gets ugly: who can actually redeem, and what has to be sold to fund it. The BIS framing is the right mental model. A reserve-backed coin that holds short-dated government bonds is running liquidity transformation, and that is a runnable balance sheet even when the assets look “high quality.”

The expensive misconception is “fully backed” equals “risk-free.” The reserve mix is where the risk is manufactured, and yield is usually the tell. If the structure is reaching for interest-bearing assets, holders are implicitly long the issuer’s liquidity management and short the market’s ability to absorb forced selling without a fire-sale discount. That is the difference between a stable quote and a stable redemption promise.

Sources

• Bank for International Settlements
• Bank for International Settlements
• Bank for International Settlements
• Bank for International Settlements