What are decentralized stablecoins?
A decentralized stablecoin is a coin that tries to hold a stable value, but without a single firm issuing it or buying back dollars. Everything goes onchain. There is no central company that decides who can mint or redeem.
Stablecoins are already a cornerstone of the DeFi ecosystem. Fiat is not a natural part of the blockchain, which is why stablecoins serve as a “working currency” between protocols. They are used to transfer value, as collateral, and to park capital while waiting for the next trade.
Regulators have also noticed this. Stablecoins are essential to the functioning of DeFi, as they enable transfers, deposits, and collateralization. Without them, most protocols practically cannot work.
That is why it is interesting that Vitalik Buterin warned at the beginning of 2026 that crypto still does not have good enough decentralized stablecoins. In a post dated January 11, 2026, he listed three issues that have not yet been resolved.
The first is dependence on the USD benchmark. Most stablecoins follow the dollar, and thus the entire system remains tied to a single fiat currency.
Another problem is oracles. If they can be controlled by players with deep pockets, then the “decentralized” model loses its meaning.
The third is staking yields. If staking yields a solid yield, stablecoin models that attempt to be decentralized must offer a competitive design to maintain liquidity.
It is also interesting that at the beginning of 2026, the total supply of stablecoins is around $300 billion, depending on the tracker and the day of measurement. Most of this liquidity is still centralised. This is exactly what Vitalik warns about.
Source: cointelegraph
Vitalik's idea
In a post dated January 11, 2026 on X, Vitalik Buterin wrote that DeFi still does not have “stable money” that is truly independent of individual issuers and a single benchmark.
According to him, there are three main problems.
The first is the index. Most stablecoins follow the USD price. Vitalik believes that a better benchmark index than the dollar itself should be designed. In the long run, the idea is that the stablecoin doesn’t just depend on a fiat price ticker. If the USD loses purchasing power in 20 years or enters stronger inflation, the system should not be tied exclusively to that.
Another problem is oracles. They should be decentralized and not be “capturable” by a large pool of money. If someone with enough capital can influence the oracle feed, then the entire model becomes vulnerable. This directly affects collateral, liquidations, and protocol security.
The third problem is staking yield. Staking is a competition to stablecoin design. If staking yields a good return, users will prefer to hold a staked coin rather than a stablecoin that has no yield or has a weaker reward model.
Vitalik lists several possible directions of solution. One is to reduce the staking yield, say to 0 to 2 percent, to reduce competition with stablecoins. The second is to create a new type of staking with a yield close to regular staking, but without the same slashing risk. The third is to try to align the slashing mechanism with the fact that the staked coin can still serve as collateral, without users taking on additional risk that they don’t understand.
He also emphasizes that “slashing risk” is not just a technical detail. On the one hand, it protects the network from bad validator behavior. On the other hand, there is a risk of censorship or inaction that can have serious consequences.
One important point he points out is that a stablecoin cannot be safe if it is 100 percent secured only by ETH collateral. In an extreme scenario, such as a large price drop or a change in staking yield, the model would need to have a clear plan on how to react.
The point of his idea is simple. If we want a truly decentralized stablecoin, we need to solve the problem of benchmark, secure oracles, and staking yield competition. Without this, most of the liquidity will still remain in centralized stablecoins.
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo.
Let's stop looking at "$1" as the only definition of stability
The first thing that Vitalik opens up is the benchmark itself. In a post dated January 11, 2026, he says that tracking the USD is okay in the short term. But if we talk about the resilience of the system over decades, then a stablecoin should not depend on only one price and one country.
This is a direct criticism of today’s DeFi model. Even the most well-known decentralized projects target the USD soft peg. An example is DAI , which clearly targets a value of $1 in the documentation.
Vitalik does not offer a ready-made plan of what exactly would replace the dollar. But he mentions the idea of a broader price index or purchasing power measure instead of a pure USD peg.
Conceptually, this could mean something similar to the CPI model. It is a basket of goods and services whose price is tracked over time to measure inflation. Another option is composite currency baskets such as Special Drawing Rights used by the International Monetary Fund. SDR is based on a weighted combination of several large fiat currencies, so it doesn’t depend on just one.
But as soon as you switch such a model onchain, you open up new questions. How to measure this index? Who determines the composition of the basket? How to update the data? This is where we immediately come to the oracle problem.
Interestingly, the CPI basket measures inflation through the prices of everyday products and services, while the SDR is a synthetic reserve asset based on multiple currencies with the aim of reducing dependence on one country. Vitalik wants to see similar thinking with decentralized stablecoins.
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo.
Oracles that are not easy to download
Another problem is oracles (third parties that connect blockchains to external data). If a stablecoin depends on external data, then it is only as strong as its oracle design. This is the main point that Vitalik emphasizes.
The target should be an oracle that is really decentralized and that cannot be easily “bought” with large capital. If someone with enough money can manipulate the price, index, or collateral valuation, then they can provoke bad mints, bad liquidations, or even lead the system to insolvency.
This is not a theory. This is a well-known DeFi risk. When a stablecoin is used as collateral and a settlement asset in multiple protocols, a single oracle fail can spill over through the entire ecosystem. Liquidations start in a chain, forced selling and hashrate occur, staking or the price of coins are no longer the only problem.
The MakerDAO documentation shows how complex this is even with mature systems. The median value from the whitelisted data feeds is used. There is governance control and permissioning. There are also parameters such as the minimum quorum for data updates.
A minimum quorum means that a sufficient number of sources or participants must participate for the update to be valid. This is an attempt to prevent a small number of actors from changing key data.
In the end, stablecoin decentralization often comes down to oracle governance, constant system maintenance, and a clear plan for what happens in the event of an error. Without it, the whole model may look decentralized, but in practice it remains vulnerable.
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo.
Staking yield is a competition to stable collateral
The third issue that Vitalik points out is the staking yield on Ethereum. According to him, this is an underrated source of tension for decentralized stablecoins.
If staking on Ethereum with a yield on ETH is on an ETH basis, then stablecoin models have a problem. Either they have to offer a similar yield through incentives that may not survive the stress test of the market, or they have to accept that the capital will go where the yield is more attractive in the long run.
In other words, if you can stake ETH and get a solid return without the added complexity, why would you hold a stablecoin with no yield or a worse risk reward ratio.
Vitalik does not give a single solution, but presents several directions as a thought experiment.
One option is to lower the staking yield to around 0.2 percent, something he describes as a “hobby level.” Thus, staking would no longer be a strong competition to stablecoin design.
Another option is to create a new staking category with a yield closer to classic staking, but without the typical slashing risk.
The third option is to design a mechanism that clearly aligns slashable staking and the use of that same asset as collateral. This means that the model must count on the risk of slash in advance and have a clear way to deal with it.
The conclusion is simple. The resilience of stablecoins must be tested in an environment where incentives change and where there may be a sharp drop in the price of the coin. If the model only works as long as the market is calm and yield is stable, then it is not sustainable in the long run.
Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo.
What this means for protocol design
If you analyze a decentralized stablecoin or the DeFi protocol that depends on it, these questions directly affect the possible failure modes that Vitalik warns about.
The first question is simple. What exactly is a stablecoin tied to? The strict 1 USD peg is clear and easy to understand, but in the long run it imports USD benchmark risk. If a project claims to track a basket, index, or purchasing power, it’s crucial to know who defines that benchmark and how it’s updated.
The second is run scenarios. What happens in a quick sell off? Does the model rely solely on the trust of users, or is there a clear, mechanical way to restore collateral without a death spiral? The history of DeFi has shown that this is precisely a common breaking point.
The third is the integrity of the oracle. What data do you have to trust? What is the official policy if feeds fail, disagree, or are manipulated? Research by the Bank for International Settlements cites oracles as one of the main risk surfaces in the DeFi system. In practice, we have already seen liquidations and losses due to oracle manipulation.
The fourth is the realism of collateral and liquidation. Is there actual onchain liquidity during volatility, or does the model assume “normal” market conditions? In a panic, the spread expands, liquidity disappears and the assumptions from the whitepaper often no longer apply.
The fifth is incentives versus resilience. If stability depends on yield or subsidies, what happens when the staking yield rises or when the incentives expire? Capital is opportunistic. It goes where the risk reward is better.
Ultimately, the design of a stablecoin is not just a matter of freckles. It is about how the model reacts under stress, how incentives are set and whether the system can survive without relying on constant market growth or perfect conditions.