Stablecoins are not stable; they only appear stable before a collapse, which can happen in just a few hours.
Written by: YQ
Translated by: AididiaoJP, Foresight News
The first two weeks of November 2025 exposed the fundamental flaws in the decentralized finance (DeFi) sector that scholars have repeatedly warned about for years, as stablecoins faced a crisis. The collapse of Stream Finance's xUSD, followed by the chain reaction that triggered Elixir's deUSD and numerous other synthetic stablecoins, was far from an isolated mismanagement incident. These events revealed deep structural issues within the DeFi ecosystem regarding risk management, transparency, and trust.
What I observed during the Stream Finance collapse was not a clever exploitation of smart contract vulnerabilities or a traditional oracle manipulation attack. Instead, it revealed a more disturbing fact: beneath the glamorous rhetoric of "decentralization," basic financial transparency was already lacking. When an external fund manager can lose $93 million without any effective oversight, triggering a cross-protocol chain reaction of up to $285 million; when the entire "stablecoin" ecosystem loses 40% to 50% of its total locked value within a week, despite ostensibly maintaining its peg, we must acknowledge a fundamental truth about the state of decentralized finance: this industry has learned nothing.
More accurately, the current incentive structure rewards those who ignore historical lessons and punishes those who act cautiously, shifting losses to the entire market when an inevitable collapse occurs. An old saying in finance is particularly apt and harsh here: if you don't know where the returns come from, you are the source of the returns. When certain protocols promise an 18% return through opaque strategies while the yields in mature lending markets are only 3% to 5%, the high yield is likely the principal of the depositors.
The Operating Mechanism and Risk Transmission of Stream Finance
Stream Finance positions itself as a yield optimization protocol, offering up to 18% annual returns on users' USDC deposits through its interest-bearing stablecoin xUSD. Its claimed strategies include "Delta-neutral trading" and "hedged market making," but the actual operation is vague. In comparison, Aave offers an annual yield of about 4.8% for USDC deposits, while Compound is slightly above 3%. Basic financial common sense tells us to be wary of returns that are multiples of the market average; however, users still deposited hundreds of millions into the protocol. Before the collapse, the trading price of 1 xUSD reached as high as 1.23 USDC, showcasing its compound returns. At its peak, xUSD claimed to manage assets worth $382 million, but data from DeFiLlama showed its peak total locked value was only $200 million, indicating that over 60% of the so-called assets existed in unverifiable off-chain positions.
After the collapse, Yearn Finance developer Schlagonia revealed its true mechanism: it was a systematic fraud disguised as financial innovation. Stream adopted a recursive lending model, creating synthetic assets without actual value support through the following process:
Users deposit USDC.
Stream exchanges USDC for USDT via CowSwap.
Using these USDT, it mints deUSD on Elixir (choosing Elixir for its high yield incentives).
These deUSD are cross-chained to networks like Avalanche and deposited in lending markets to borrow more USDC, completing a cycle.
While this strategy appears complex and relies on cross-chain operations, it can still be viewed as a form of collateralized lending. However, Stream did not stop there. It did not use the borrowed USDC solely to expand the collateral cycle but instead minted xUSD again through its StreamVault contract, causing the supply of xUSD to far exceed the actual value of its collateral. With only $1.9 million of verifiable USDC as collateral, Stream minted as much as $14.5 million of xUSD, inflating the scale of synthetic assets by 7.6 times relative to the underlying reserves. This is akin to a fractional reserve banking system without reserves, regulation, or a lender of last resort.
Its circular dependency with Elixir further endangered the entire structure. During the inflation of xUSD supply, Stream deposited $10 million of USDT into Elixir, thereby expanding the supply of deUSD. Elixir then exchanged this USDT for USDC and deposited it in Morpho's lending market. By early November, the supply of USDC on Morpho exceeded $70 million, with borrowing exceeding $65 million, and Elixir and Stream were major participants. Stream held about 90% of the total supply of deUSD (approximately $75 million), and a significant portion of Elixir's reserve assets was loans provided to Stream through private Morpho vaults. These two stablecoins were essentially collateralized against each other, leading to mutual destruction. This "financial cycle" inevitably resulted in systemic fragility.
Industry analyst CBB publicly pointed out the problem as early as October 28: "xUSD has about $170 million of on-chain assets backing it, but they borrowed about $530 million from lending protocols. This is a 4.1x leverage built on many illiquid positions. This is not yield farming; it's gambler's risk." Schlagonia also warned the Stream team 172 days before the collapse that a five-minute check of their positions would reveal that a collapse was inevitable. These warnings were public, specific, and accurate, but were ignored by users chasing high yields, curators eager for fee income, and various protocols supporting the structure.
When Stream announced on November 4 that an external fund manager had lost about $93 million in fund assets, the platform immediately suspended all withdrawals. Due to the lack of an effective redemption mechanism, panic spread instantly. Holders rushed to sell xUSD in a liquidity-starved secondary market. Within hours, the price of xUSD plummeted 77% to about $0.23. This stablecoin, which had promised stability and high returns, evaporated three-quarters of its value in a single trading day.
Specific Data on Risk Transmission
According to a report from the DeFi research institution Yields and More (YAM), the direct debt exposure related to Stream reached as high as $285 million across the ecosystem. The main parties involved include:
TelosC: Facing a loan risk of $123.64 million due to accepting Stream assets as collateral (the largest single risk exposure).
Elixir Network: Lending $68 million through private Morpho vaults (accounting for 65% of deUSD reserves).
MEV Capital: Risk exposure of $25.42 million, with about $650,000 becoming bad debt (due to the oracle freezing the price at $1.26 when the market price of xUSD dropped to $0.23).
Others: Varlamore ($19.17 million), Re7 Labs (one vault $14.65 million, another vault $12.75 million), and institutions like Enclabs, Mithras, TiD, and Invariant Group holding smaller positions.
Euler: Facing about $137 million in bad debt.
Over $160 million in funds were frozen across various protocols.
Researchers noted that this list is not exhaustive and warned that "there are likely more stablecoins or vaults affected," as the full picture of risk transmission remains unclear weeks after the collapse.
Elixir's deUSD, with 65% of its reserves concentrated in loans to Stream through private Morpho vaults, plummeted 98% from $1.00 to $0.015 within 48 hours, becoming the fastest major stablecoin failure since the collapse of Terra UST in 2022. Elixir redeemed for about 80% (excluding Stream itself) of deUSD holders, allowing them to exchange 1:1 for USDC, thereby protecting most community users, but this protective measure came at a significant cost, with losses being passed on to protocols like Euler, Morpho, and Compound. Subsequently, Elixir announced it would completely shut down all stablecoin products, acknowledging that market trust had been irreparably destroyed.
The broader market response reflected a systemic loss of confidence. According to data from Stablewatch, within a week after the Stream collapse, interest-bearing stablecoins lost 40% to 50% of their total locked value, although most still maintained their peg to the dollar. This means about $1 billion in funds flowed out of protocols that had not failed and had no technical issues. Users could not distinguish between quality projects and fraudulent ones, so they chose to flee all similar products. The total locked value in DeFi dropped by about $20 billion in early November. The market is pricing in the risk of a widespread chain reaction rather than just the failure of specific protocols.
October 2025: A chain liquidation triggered by $60 million
Less than a month before the collapse of Stream Finance, the cryptocurrency market experienced another catastrophic crash. On-chain forensic analysis indicates that this was not an ordinary market downturn, but a precision attack leveraging known vulnerabilities on an institutional scale. From October 10 to 11, 2025, a carefully orchestrated market sell-off worth $60 million triggered oracle failures, leading to a chain reaction of large-scale liquidations across the entire DeFi ecosystem. This was not due to over-leveraged liquidations caused by positions being under-collateralized, but rather a replay of the design flaws in the oracle system at an institutional level, repeating attack patterns that have been recorded and disclosed multiple times since February 2020.
The attack began at 5:43 UTC on October 10, with $60 million worth of USDe being concentratedly sold into the spot market of a single exchange. In a well-designed oracle system, the impact of such localized selling pressure would be absorbed by multiple independent price sources with time-weighted average price mechanisms, thereby minimizing manipulation risks. However, the actual oracle system adjusted the value of the related collateral (wBETH, BNSOL, and USDe) in real-time based on the manipulated spot prices from the trading venue. Large-scale liquidations were triggered immediately. The system infrastructure was overwhelmed, and millions of simultaneous liquidation requests exhausted the system's processing capacity. Market makers were unable to quickly submit buy orders due to API data stream interruptions and a backlog of withdrawal requests. Market liquidity evaporated instantly, creating a vicious cycle of cascading failures.
Attack Methods and Historical Precedents
The oracle faithfully reported the manipulated prices from a single trading venue, while prices in all other markets remained stable. Major exchange data showed USDe dropping to $0.6567 and wBETH falling to $430. Meanwhile, prices in other venues deviated by less than 0.3% from normal levels. The liquidity pools on decentralized exchanges were minimally affected. As pointed out by Ethena founder Guy Young, "there were over $9 billion in stablecoin collateral available for redemption throughout the event," proving that the underlying assets themselves were not impaired. However, the oracle reported the manipulated prices, and the liquidation system executed based on these prices, ultimately destroying a large number of positions based on valuations that did not exist in the market.
This mirrors the pattern that severely impacted Compound in November 2020: at that time, DAI's trading price on Coinbase Pro spiked to $1.30 for a brief hour while other markets remained at $1.00, resulting in $89 million in liquidations. The trading venue changed, but the vulnerabilities remained. This attack method is identical to the one that destroyed bZx in February 2020 (stealing $980,000 by manipulating the Uniswap oracle) and the one that severely impacted Harvest Finance in October 2020 (stealing $24 million by manipulating Curve prices, leading to a $570 million bank run), as well as the multi-exchange manipulation attack in October 2022 that caused Mango Markets to lose $117 million.
Statistics show that between 2020 and 2022, there were 41 oracle manipulation attacks, resulting in losses of up to $403.2 million. The industry's response has been slow and fragmented, with most platforms continuing to use oracle solutions that heavily rely on spot prices and lack sufficient redundancy. The amplification effect of these attacks underscores the importance of learning these lessons as market sizes grow. In the 2022 Mango Markets incident, $5 million in manipulation led to $117 million in losses, a 23-fold amplification. In the October 2025 incident, $60 million in manipulation triggered a chain reaction with a significant amplification effect. The attack methods have not become more sophisticated; rather, the underlying systems have grown larger while retaining the same fundamental vulnerabilities.
Historical Patterns: Failure Cases from 2020 to 2025
The collapse of Stream Finance is neither a first nor an isolated incident. The DeFi ecosystem has experienced multiple stablecoin failures, each exposing similar structural vulnerabilities. Yet the industry continues to repeat the same mistakes, with each instance growing larger in scale. The recorded failure cases over the past five years exhibit a highly consistent pattern:
Unsustainable High Yields: Algorithmic or partially collateralized stablecoins offer yields far above market levels to attract deposits.
Questionable Sources of Yield: High yields are often subsidized by token issuance or new inflows of deposits, rather than genuine business income.
Excessive Leverage and Opacity: Protocol operations imply high leverage, with unclear actual collateralization rates.
Circular Dependencies: Assets of Protocol A support Protocol B, while assets of Protocol B in turn support Protocol A, creating a fragile closed loop.
Triggering of Collapse: Once any shock occurs, exposing their under-collateralized nature, or when subsidies become unsustainable, a bank run will occur.
Death Spiral: Users panic and exit, collateral values plummet, cascading liquidations occur, and the entire structure collapses within days or even hours.
Risk Transmission: The collapse spreads to other protocols that accept the stablecoin as collateral or are associated with its ecosystem.
Typical Case Reviews:
May 2022: Terra (UST/LUNA)
Losses: Approximately $45 billion in market value evaporated within three days.
Mechanism: UST is an algorithmic stablecoin that maintains its peg through a mint-and-burn mechanism with LUNA. The Anchor Protocol offered an unsustainable 19.5% yield on UST deposits.
Trigger: Large-scale UST redemptions and sell-offs led to a decoupling, triggering LUNA overissuance and a death spiral.
Impact: Led to the bankruptcy of several major crypto lending platforms, including Celsius, Three Arrows Capital, and Voyager Digital. Founder Do Kwon was arrested and faced fraud charges.
June 2021: Iron Finance (IRON/TITAN)
Losses: Approximately $2 billion in total locked value went to zero within 24 hours.
Mechanism: IRON was partially collateralized by 75% USDC and 25% its native token TITAN, offering yield incentives of up to 1700% annually.
Trigger: Large redemptions led to the sale of TITAN, causing its price to collapse, which in turn destroyed the collateral base of IRON.
Lesson: Partial collateralization and reliance on volatile tokens as collateral are extremely dangerous under pressure.
March 2023: USDC
Decoupling: USDC briefly dropped to $0.87 due to $3.3 billion in reserves being trapped in the collapsed Silicon Valley Bank.
Impact: Undermined market confidence in "fully reserved" fiat stablecoins. Triggered a decoupling of DAI (over 50% of its collateral was USDC), leading to over 3,400 liquidations worth a total of $24 million on Aave.
Lesson: Even the most trusted stablecoins face concentration risks and dependencies on the traditional financial system.
November 2025: Stream Finance (xUSD)
Losses: Direct losses of $93 million, with a total risk exposure of $285 million in the ecosystem.
Mechanism: Recursive lending created synthetic assets without actual support (leverage of 7.6 times). 70% of funds were managed by an anonymous external manager through opaque off-chain strategies.
Current Status: xUSD trading price ranges between $0.07 and $0.14, liquidity has dried up, withdrawals are frozen, and multiple lawsuits are pending. Elixir has shut down.
Summary of Common Failure Patterns:
Unsustainable Yields: Terra (19.5%), Iron (1700% APR), Stream (18%).
Circular Dependencies: UST-LUNA, IRON-TITAN, xUSD-deUSD.
Lack of Transparency: Concealing subsidy costs, hiding off-chain operations, questionable reserve assets.
Fragile Collateral Structures: Partial collateralization or reliance on volatile tokens, easily leading to death spirals under pressure.
Oracle Manipulation: Distorted price inputs leading to erroneous liquidations and accumulation of bad debt.
The conclusion is clear: stablecoins are not stable; they only appear stable before a collapse, which can happen in just a few hours.
Oracle Failures and Infrastructure Collapse
At the onset of the Stream collapse, the issues with the oracle became glaringly apparent. When the market price of xUSD had dropped to $0.23, many lending protocols had hard-coded its oracle price at $1.00 or higher, aiming to prevent chain liquidations. This well-intentioned decision led to a severe disconnect between protocol behavior and market reality. This hard-coding was a proactive policy choice, not a technical failure. Many protocols adopted manual updates for oracle prices to avoid triggering liquidations during temporary market fluctuations. However, when the price drop reflects the project's substantive under-collateralization, this approach can have catastrophic consequences.
Protocols faced a difficult choice:
Use real-time prices: This may expose them to market manipulation and the risk of triggering chain liquidations during volatility (as demonstrated by the costly October 2025 incident).
Use delayed prices or time-weighted average prices: This fails to timely reflect the true under-collateralization situation, leading to the accumulation of bad debt (as in the Stream incident, where the oracle showed $1.26 while the actual price was $0.23, resulting in $650,000 in bad debt for MEV Capital).
Use manual updates: This introduces risks of centralized decision-making and human intervention, potentially covering up the truth of under-collateralization by freezing prices.
All three methods have previously led to losses in the hundreds of millions or even billions of dollars.
Infrastructure Capacity Under Pressure
As early as October 2020, Harvest Finance faced a $24 million attack that triggered a bank run, causing its total locked value to plummet from $1 billion to $599 million. The lessons from its infrastructure collapse were clear: the oracle system must consider the infrastructure capacity under extreme stress events; the liquidation mechanism must have rate limits and circuit breakers; exchanges must reserve redundant capacity for traffic ten times the normal level. However, the events of October 2025 proved that this lesson has still not been heeded at the institutional level. When thousands of accounts faced liquidation simultaneously, when billions of dollars in positions were liquidated within an hour, and when the order book became blank due to all buy orders being exhausted and the system overloaded and unable to accept new orders, the failures of the infrastructure and the oracle were equally catastrophic. Technical solutions to these problems exist, but they are often not implemented because they would reduce efficiency during normal operations and increase costs (affecting profits).
Core Warnings
If you cannot identify the source of the yield, then you are not the recipient of the yield; you are, in fact, the cost of someone else's profit. This principle is not complicated; however, billions of dollars continue to flow into "black box" strategies because people tend to believe comforting lies rather than confront unsettling truths. The next Stream Finance may very well be operating at this moment.
Stablecoins are not stable.
Decentralized finance is often neither fully decentralized nor secure.
Unidentified sources of yield are not profits; they are assets in transfer with a countdown.
These are not personal opinions but verifiable facts repeatedly validated at a great cost. The only question is: will we ultimately act based on these known lessons, or will we prefer to pay another $20 billion in tuition to relearn them? History seems to lean toward the latter.
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
