What is the best token to use as collateral to mint a decentralized stablecoin that is not pegged to the USD?

If you mean “what collateral asset should I use to build a decentralized stablecoin whose value is not targeted to $1?”, my first choice would be ETH.

The closest proven example is Reflexer’s RAI: it is backed exclusively by ETH and uses a managed-float monetary policy rather than maintaining a fixed USD peg. Its price is allowed to move with supply and demand, while the protocol adjusts incentives to dampen volatility.

Why ETH is probably the best collateral

• Deep liquidity: ETH has one of the deepest on-chain markets, making liquidations and arbitrage substantially easier.
• Decentralized: You don't need a centralized issuer, bank deposits, or tokenized Treasuries.
• Native to Ethereum: No bridge dependency if the stablecoin lives on Ethereum.
• Long track record: ETH-backed systems such as RAI and Liquity have demonstrated the basic collateralized-debt model.
• Yield potential: Staked ETH introduces an additional economic component if you use wstETH or another liquid-staking derivative.

I'd rank potential collateral roughly:

1. ETH — best overall for decentralization + liquidity + simplicity.
2. wstETH — potentially better economically because the collateral earns staking yield, but adds LST-specific smart-contract/depeg risk.
3. BTC (wrapped/native L2 version) — excellent monetary asset, but adds implementation/bridging/oracle complexity depending on the chain.
4. A diversified ETH/LST basket — potentially more robust, but considerably more complicated to design safely.
5. Governance tokens / volatile altcoins — generally poor choices because collateral crashes can coincide with falling liquidation liquidity.

The more interesting question: what should it be stable against?

If you don't want a USD peg, I wouldn't necessarily try to make it "stable" in the conventional sense. You could instead target something like:

• ETH-denominated stability — e.g. 1 token ≈ a fixed amount of ETH.
• A CPI/inflation-adjusted unit — stability in terms of purchasing power.
• A crypto basket — e.g. an index of ETH/BTC/etc.
• A floating unit of account — similar conceptually to RAI, where the protocol manages volatility without promising a fixed external price.
• A commodity-linked unit — e.g. purchasing power corresponding to a quantity of gold, although this introduces oracle/market-design issues.

For a genuinely non-USD decentralized money, I'd be particularly interested in ETH collateral + a floating unit of account, rather than ETH collateral + an artificial USD target. RAI is the clearest historical precedent for this design.

One important distinction: BOLD is not what you're looking for. Liquity V2 uses ETH/wstETH/rETH as collateral, but BOLD explicitly targets $1, so it remains a USD stablecoin despite being crypto-native and decentralized.

If you're thinking about actually designing a new stablecoin, I can also lay out the optimal architecture—collateral, target price mechanism, liquidation ratio, oracle, interest-rate controller, and how to prevent death spirals—and compare an ETH-backed RAI-style design against an ETH/BTC basket.

When designing or choosing collateral for a decentralized stablecoin that is not pegged to the USD (for instance, a stablecoin pegged to the Euro, a basket of commodities, a local consumer price index, or a neutral unit of account), there is no single "best" token. Instead, the ideal collateral depends entirely on what the stablecoin is trying to track.

Using a mismatched collateral type introduces dangerous "wrong-way risk" or severe de-pegging pressure during market stress. Here is how the top collateral choices stack up based on the target peg:

1. For a Euro-Pegged (or Other Fiat-Pegged) Stablecoin

• Top Collateral Choice: Wrapped native fiat-backed stablecoins of that specific currency (e.g., regulated Euro stablecoins like EURC) combined with decentralized short-term European Real World Assets (RWAs).
• Why it works: If your stablecoin tracks the Euro (€), backing it with native Euro-denominated assets eliminates foreign exchange (/€) volatility. If a user redeems the stablecoin, the underlying collateral moves in lockstep with the target unit.
• The Catch: Pure crypto like ETH or BTC introduces massive FX risk. If ETH drops 20% against the Euro, your Euro-stablecoin risks undercollateralization even if the Euro itself remained stable against the dollar.

2. For a Commodity-Pegged Stablecoin (e.g., Gold, Energy)

• Top Collateral Choice: Tokenized physical commodities (e.g., PAXG or XAU₮ for gold) or decentralized liquid representations of that specific commodity.
• Why it works: Perfect correlation. If your stablecoin aims to represent a gram of gold or a gigawatt-hour of decentralized compute/energy, using the commodity itself (or a deeply liquid, bankruptcy-remote tokenized receipt of it) as collateral ensures the value proposition remains seamless.

3. For a Consumer Price Index (CPI) or Purchasing Power Peg

• Top Collateral Choice: A diversified basket of inflation-hedged assets—primarily Ether (ETH) for network yield/growth, Wrapped Bitcoin (wBTC) as digital gold, and tokenized short-term sovereign debt bonds across multiple jurisdictions.
• Why it works: If you aren't tethered to a single fiat currency, you want a collateral basket that neutralizes single-country monetary policy failures. ETH acts as a productive asset that generates native staking yield (which can subsidize stability fees), while BTC acts as a hard macro hedge.

Summary: What is the overall "Gold Standard" baseline?

If you are forced to pick a single, permissionless, highly liquid crypto asset to act as the foundational collateral engine for a non-USD stablecoin, Ether (ETH) remains the best technical choice.

• High Liquidity: Deepest secondary markets across decentralized exchanges and lending protocols, minimizing liquidation slippage.
• Productive Yield: Staked ETH (stETH) generates an organic yield that can dynamically stabilize the system or reward holders, something non-yielding assets like Bitcoin cannot do natively.
• Decentralization: Immune to single-issuer freeze risks or jurisdiction-specific banking collapses (unlike USDC/USDT).

However, remember the golden rule of stablecoin design: Your collateral must correlate with your peg. If your peg isn't USD, relying solely on volatile native crypto like ETH requires aggressive overcollateralization and robust liquidation mechanisms to prevent a fatal death spiral during a market downturn.

To help narrow down the ideal architecture, what asset or index is your stablecoin trying to peg to, and are you prioritizing fully decentralized (crypto-only) or hybrid (RWA/fiat-backed) collateral?

When minting a decentralized stablecoin that is not pegged to the USD (e.g., pegged to a national index, a commodity like gold, a basket of goods, or a non-dollar fiat currency like the Euro or Swiss Franc), the "best" collateral depends entirely on your target unit of account.

However, looking at it through the lens of decentralization, liquidity, and economic alignment , a few top-tier tokens stand out:

• Ether (ETH): The gold standard for crypto-backed, decentralized systems. It has deep liquidity, widespread integration across all major DeFi lending markets, and sufficient volatility history to calculate precise overcollateralization ratios (CR > 150%). If your non-USD stablecoin is pegged to a global basket or a floating computational unit, ETH remains the most robust native Web3 collateral.
• Wrapped Bitcoin (wBTC or native BTC via trustless bridges/bitcoins): Ideal if your non-USD stablecoin aims to be completely sovereign from platform-specific risks or single-ecosystem monetary policy. It introduces less correlation to smart-contract platform shifts, though liquidity fragmentation across chains must be managed.
• Liquid Staking Tokens (LSTs like stETH) or Liquid Restaking Tokens (LRTs): Using yield-bearing collateral allows the stablecoin system to capture native staking yield, which can subsidize stability fees or backstop the peg. However, it adds smart contract and slashing risks on top of the base asset's volatility.
• Real-World Assets (RWAs) or Foreign Fiat-backed Tokens (EURC, etc.): If your stablecoin is pegged specifically to the Euro, using euro-denominated tokenized assets or EURC as collateral eliminates currency mismatch risk (the dreaded delta risk between USD/crypto collateral and a non-USD peg), though it compromises on pure censorship resistance if the underlying RWA is centralized.

Key Trade-offs to Consider

• The Currency Mismatch Problem: If your collateral (ETH or BTC) fluctuates heavily against your target non-USD peg (e.g., Euro or a consumer price index), your liquidation engine must reprice and react much faster to regional forex fluctuations.
• Liquidity Depth: Obscure or hyper-localized collateral will cause your liquidation auctions to fail during black swan market events, breaking the peg regardless of the math on paper.

To help narrow down the optimal choice, tell me:

• What is the specific asset or index your stablecoin is pegged to?
• Are you prioritizing maximum decentralization (censorship resistance) or capital efficiency (low collateral requirements)?

If your goal is a genuinely decentralized, crypto-collateralized stable asset that is not pegged to the U.S. dollar, my first choice for collateral is ETH.

The strongest precedent is Reflexer Finance's RAI: it is overcollateralized exclusively with ETH and deliberately has a managed float rather than a USD peg. Its redemption price adjusts in response to supply and demand rather than attempting to stay at $1.

Why ETH is probably the best collateral

• Decentralization: ETH itself doesn't depend on a centralized issuer, unlike USDC, USDT, or tokenized T-bills.
• Deep liquidity: ETH has one of the deepest crypto markets, which is important for liquidations.
• Native to DeFi: It has extensive on-chain liquidity, lending infrastructure, derivatives, and oracle coverage.
• No additional issuer risk: Using something like WBTC introduces another layer of custody/bridge risk.
• Strong precedent: Both RAI and Liquity independently chose ETH as their primary collateral for decentralized borrowing systems.

But there's an important distinction

If you're designing a new stablecoin, I wouldn't simply say "ETH-backed = stable."

ETH is volatile. The interesting design question is what you want the unit of account to represent.

For example:

If your objective is "a decentralized money that isn't ultimately a synthetic dollar," I'd investigate an ETH-only system with a floating target, rather than a basket of collateral.

RAI is particularly relevant because the protocol explicitly describes itself as an ETH-backed stable asset that is not designed to return to a fixed initial value or peg.

One caveat: RAI itself is now quite small—the latest DeFiLlama data I found puts circulating supply around 531,000 RAI / ~$1.2M market cap—so I would treat it as an important design precedent, not necessarily evidence of current economic scale.

If you're actually designing the protocol, I can go one level deeper and compare ETH vs BTC vs staked ETH vs SOL vs a diversified crypto basket as collateral, including liquidation risk, oracle risk, capital efficiency, and what mechanism I'd use to create a non-USD unit of account.

If by “not pegged to USD” you mean a decentralized stable asset whose unit of account is independent of the dollar, my first choice for collateral is ETH.

Why ETH is the strongest collateral

1. It is relatively decentralized.
ETH has no issuer, redemption counterparty, or centralized reserve manager. That matters enormously for a censorship-resistant stablecoin.

2. It has deep liquidity.
A collateralized stablecoin needs to be able to liquidate collateral during a crisis. ETH has much deeper on-chain liquidity than most alternative crypto assets.

3. It is native to the dominant DeFi settlement layer.
If the stablecoin lives on Ethereum, using ETH avoids adding another protocol or token dependency.

4. It has no direct USD liability.
USDC, USDT, tokenized Treasuries, etc. are inappropriate if your goal is to create a genuinely non-USD monetary unit because their value and/or redemption ultimately depends on the dollar system.

5. There is already a strong proof of concept: RAI.
Reflexer's RAI is explicitly designed as a non-pegged, ETH-backed stable asset. Its price is allowed to float rather than targeting $1. Reflexer describes RAI as ETH-backed with a managed-float monetary policy.

My ranking

There is an important distinction between ETH and staked ETH. If you're designing the protocol from scratch, I'd probably use ETH as the primary collateral, with wstETH/rETH as optional secondary collateral only after the core system has demonstrated robust liquidation behavior. Liquity's current design is instructive here: it supports WETH, wstETH and rETH, while explicitly separating collateral markets because each asset has different liquidity and liquidation risks.

But there's a bigger design question

If you're trying to build something better than RAI, I wouldn't simply ask:

“What token should back it?”

I'd ask:

“What should the stablecoin be stable against?”

RAI deliberately doesn't maintain a permanent $1 peg. Its target/reference price can move according to the protocol's monetary policy, while market supply and demand determine the actual exchange rate.

You could therefore build an ETH-collateralized stable asset with, for example:

• Floating purchasing power — target a basket of commodities.
• CPI purchasing power — target an inflation-adjusted unit rather than USD.
• ETH-denominated stability — e.g., 1 token represents a relatively stable claim on a defined amount of ETH.
• Global consumption basket — target a basket of currencies/commodities rather than USD.
• No external peg whatsoever — algorithmically manage volatility around a floating reference price.

The last approach is closest philosophically to RAI.

What I would build

If the objective is maximum decentralization + non-USD monetary policy, my starting architecture would be:

ETH collateral → overcollateralized CDPs → floating stable asset → PID/controller monetary policy → decentralized ETH/USD + potentially non-USD reference oracle → liquidation auctions

with ETH-only collateral initially.

I would not start with a multi-collateral system. More collateral sounds safer, but it introduces additional oracle, liquidity, liquidation and governance dependencies. Liquity's documentation explicitly notes that adding collateral branches can reduce redemption efficiency and that collateral liquidity/volatility are critical considerations.

One caveat: “stablecoin” is potentially misleading here. If it isn't pegged to USD, you need to define precisely what “stable” means. RAI is probably the most relevant existing model to study rather than BOLD, because BOLD is explicitly designed around a $1 USD value despite being crypto-collateralized.

If you're considering actually designing a new protocol, I can also compare ETH vs BTC vs a diversified ETH/BTC basket quantitatively across collateral volatility, liquidation risk, oracle requirements, minimum collateral ratio, and potential monetary-policy designs.

When designing or minting a decentralized stablecoin that is not pegged to the U.S. dollar, the choice of the "best" collateral depends entirely on what target asset or currency your stablecoin is tracking (e.g., Euro, Gold, or a basket of goods/currencies) and how much volatility you can tolerate in the system.

The primary options and their trade-offs depend on your target unit of account:

• If pegging to Gold / Commodities: Gold-backed tokens like PAXG (Paxos Gold) or XAUT (Tether Gold) are the gold standard. They eliminate currency-inflation risks and provide a hard-asset backing that behaves independently of fiat central bank policies.
• If pegging to the Euro (or another major foreign fiat): Decentralized, high-liquidity fiat-backed stablecoins denominated in that specific currency, such as EURC (Euro Coin), are ideal. Using an existing, highly liquid Euro stablecoin as collateral minimizes the foreign exchange (FX) tracking error and basis risk within smart contracts.
• If aiming for a truly sovereign, non-fiat, non-pegged index (Purchasing Power/Basket): A diversified basket of native crypto-assets (like ETH and BTC ) or decentralized real-world asset (RWA) tokens weighted through an on-chain index works best. This prevents systemic reliance on any single national currency or centralized corporate issuer.

Key Trade-Offs to Consider

• Volatility vs. Capital Efficiency: Using native, highly volatile crypto like ETH or BTC as collateral requires severe over-collateralization (often 150% to 200%+) and aggressive automated liquidations to protect the system.
• Oracle and Liquidity Risk: Non-USD collateral pools often suffer from thinner secondary market liquidity, making the protocol more susceptible to liquidation slippage during extreme market cascades.

To help narrow down the ideal architecture, could you tell me:

• What specific asset or currency is your stablecoin trying to track?
• Do you want the collateral to be fully on-chain crypto (like ETH) or real-world/fiat-backed tokens (like EURC or PAXG)?

If your goal is a decentralized stablecoin that is not pegged to USD, my first choice for collateral would be ETH—and the strongest existing design example is Reflexer’s RAI.

Why ETH is probably the best collateral

ETH has several properties that are unusually valuable for a non-USD stable asset:

• Highly decentralized: ETH doesn't depend on a company, bank, or custodian.
• Deep liquidity: It has one of the deepest on-chain markets, which matters enormously when liquidating collateral.
• Native to DeFi: No bridge or external asset issuer is required if the system lives on Ethereum.
• No direct fiat dependency: You can construct the monetary policy around ETH rather than trying to maintain a dollar price.
• Yield-bearing variants exist: wstETH/rETH can potentially provide staking yield, although they introduce additional smart-contract/protocol risks.

The important distinction is that the collateral and the unit of account don't have to be the same thing. You can use ETH as the reserve asset while making the stablecoin's monetary policy target something other than $1.

The model I'd look at: RAI

RAI is specifically designed as a non-pegged, ETH-backed stable asset. Its price is allowed to float according to supply and demand, while a controller adjusts the system's redemption price and incentives to dampen volatility. Reflexer explicitly describes RAI as not being pegged to anything.

That's quite different from something like Liquity's BOLD, which is also decentralized and ETH-backed but explicitly targets $1 USD.

So if you're asking:

"What collateral should I choose if I want to build the next RAI, but potentially with a better monetary policy?"

I'd rank the candidates approximately:

One subtle point: I would favor ETH over an ETH-staking derivative if minimizing systemic dependencies is the priority. wstETH and rETH add another layer of smart-contract, oracle, liquidity, and staking-provider risk. Liquity, for example, deliberately limits its collateral to WETH, wstETH and rETH and discusses liquidity and collateral-volatility risks explicitly.

But there's an even more interesting question

If you're designing a new stablecoin, I wouldn't necessarily copy RAI's approach of simply trying to make the coin's USD price stable.

I'd consider defining the unit in terms of a real economic objective, such as:

1 unit = a stable amount of purchasing power

or

1 unit = a target basket of commodities/assets

or even

1 unit = a fixed quantity of ETH's long-term economic output

Then ETH becomes the collateral, rather than the thing you're trying to peg the coin to.

That distinction could make the protocol substantially more interesting: ETH-backed, floating-value, volatility-managed money rather than another dollar clone.

If you're considering actually designing such a protocol, I can also compare ETH vs BTC vs gold vs a diversified crypto basket as collateral, including liquidation risk, oracle risk, target mechanism, optimal collateral ratio, and how I'd design the monetary policy.

When designing or choosing a decentralized stablecoin that is not pegged to the U.S. dollar (such as a stablecoin pegged to a basket of goods, a foreign fiat currency like the Euro, or a floating purchasing-power index like Reflexer’s RAI), the "best" collateral token depends entirely on your primary goal: maximum decentralization/censorship resistance versus price stability correlation.

Top Collateral Options Ranked by Use Case

• Ether (ETH) — The Gold Standard for Pure Decentralization
  • Why it works: ETH is native to the largest smart contract ecosystems, possesses deep and unfragmented liquidity, and relies on no centralized issuer or real-world bank account. Protocols like RAI use ETH exclusively as collateral because it eliminates counterparty and regulatory choke points.
  • Trade-off: ETH is volatile against almost every metric. If your non-USD stablecoin tracks a European consumer price index or a floating redemption rate, ETH's high volatility requires strict overcollateralization ratios and robust automated liquidation modules.
• Wrapped Bitcoin (WBTC or native BTC) — Maximum Store of Value
  • Why it works: Like ETH, Bitcoin is untethered from sovereign fiat debt systems. For protocols aiming to escape the broader U.S. monetary orbit entirely, utilizing a decentralized, Bitcoin-backed framework offers a neutral, globally recognized reserve asset.
  • Trade-off: Fully native BTC bridges or wrapped variations introduce varying degrees of bridge or custodial risk depending on the implementation.
• Decentralized Commodity Tokens (e.g., PAXG or XAUT for Gold)
  • Why it works: If your non-USD stablecoin is explicitly pegged to a real-world commodity (like gold or a multi-commodity resource basket), using tokenized physical assets as direct backing minimizes tracking error against inflation or hard assets.
  • Trade-off: Introduces centralized off-chain custody risk (trusting the vault that holds the physical gold), which compromises pure on-chain decentralization.
• A Diversified Crypto Basket (ETH + BTC + Liquid Staking Tokens)
  • Why it works: Diversification reduces the idiosyncratic risk of a single asset crashing. Using a basket that includes yield-bearing liquid staking tokens (like stETH) can simultaneously generate protocol revenue or offset stability fees.
  • Trade-off: Increases smart contract composition risk and complexity for oracles feeding multi-asset pricing data into the system.

Summary Recommendation

If the core objective is uncompromised decentralization and censorship resistance, ETH remains the superior collateral choice. It provides the most resilient, deep-liquidity foundation to back a floating or alternative-pegged asset without importing the regulatory and de-pegging vulnerabilities tied to centralized fiat stablecoins (like USDC or USDT).

If you have a specific target for your stablecoin—such as what it is trying to track or peg to (e.g., a specific foreign currency like the Euro, a commodity, or a floating index)—I can help you determine the ideal collateral ratio and stability mechanism for that model.

When designing a decentralized stablecoin that is not pegged to the USD (for instance, a stablecoin pegged to the Euro, Gold, a basket of consumer goods, or another fiat currency), the "best" collateral token depends entirely on your target peg and risk tolerance.

Because an asset-backed or crypto-backed stablecoin relies on the economic correlation between the collateral and the target unit of account, using a mismatched collateral (like using raw ETH to back a Euro stablecoin) introduces foreign exchange (FX) risk.

The best options based on your target non-USD design include:

• Regulated Non-USD Fiat Stablecoins (e.g., EURC or XSGD) : If you are minting a decentralized stablecoin pegged to a specific fiat currency like the Euro or Singapore Dollar, using a decentralized wrapper or liquidity pool of an established fiat-backed token (like Circle's EURC ) as part of the collateral mix is ideal. It eliminates FX tracking error, though it introduces the counterparty/regulatory risk of the issuer.
• Tokenized Commodities (e.g., PAX Gold / PAXG or Tether Gold / XAUT) : If your decentralized stablecoin is designed to track a hard asset or inflation-hedged unit (like grams of gold), utilizing gold-backed tokens as primary collateral provides a direct 1:1 economic correlation, shielding the protocol from crypto market crashes.
• Ether (ETH) or Wrapped Bitcoin (WBTC) : If you want a fully trust-minimized, censorship-resistant decentralized collateral that relies on no external traditional finance (TradFi) or centralized issuers, ETH remains the gold standard. However, because ETH is volatile and priced globally (often natively against the USD), the protocol must implement aggressive overcollateralization ratios, dynamic debt-ceiling adjustments, and robust oracle feeds to handle FX drift between the local target peg and crypto market movements.
• A Diversified Crypto Basket : Combining volatile blue-chip assets (ETH, BTC) with native protocol governance/utility tokens or decentralized liquid staking tokens (stETH) helps decentralize risk, but adds complexity to liquidation logic during market downturns.

To help narrow down the ideal architecture, tell me:

• What asset or currency is your stablecoin pegged to (e.g., Euro, Gold, a local emerging-market currency)?
• Do you prioritize maximum decentralization (accepting crypto volatility) or low price deviation (accepting some centralized stablecoin collateral)?

If your goal is to create a decentralized stablecoin whose unit of account is not the U.S. dollar, I would choose ETH as the primary collateral.

My ranking

The strongest precedent is ETH-backed debt systems. Liquity, for example, deliberately restricts its current collateral to ETH and liquid-staking derivatives (WETH, wstETH and rETH), emphasizing decentralized, crypto-native collateral.

But the more interesting question is: what should the stablecoin be stable against?

If you don't want a USD peg, I'd separate the two design decisions:

Collateral: ETH
Unit of account: something independent of USD

For example:

• CPI-indexed unit → stable purchasing power rather than stable dollar value
• Gold-indexed unit → 1 token targets a fixed quantity of gold
• ETH-denominated unit → targets a fixed amount of ETH
• Commodity basket → targets a diversified real-world purchasing-power index
• Floating crypto-native unit → stability is defined algorithmically rather than against fiat

This distinction matters enormously. A stablecoin can be ETH-collateralized while being pegged to gold, CPI, EUR, or another reference asset. The collateral doesn't have to determine the thing you're stabilizing against.

If I were designing it

I'd probably start with:

ETH / wstETH collateral → overcollateralized CDPs → non-USD unit of account → floating redemption/liquidation mechanism.

I'd make ETH the canonical collateral and potentially allow wstETH later. ETH gives you the cleanest combination of liquidity, decentralization and absence of an external issuer.

One important caveat: don't use ETH simply because it's the largest crypto asset. During a market crash, ETH collateral can fall at the same time that demand for the stablecoin rises. Research on crypto-backed stablecoins specifically identifies collateral-price shocks as a core failure mode, so the liquidation mechanism and collateralization ratio can be more important than the choice between ETH and another major token.

If you're thinking about actually designing such a stablecoin, I can also lay out what I think is the best architecture for a non-USD ETH-backed stablecoin—including the target index, collateral ratio, liquidation mechanism, oracle design, and how to prevent a death spiral.