Author: YQ

Translator: Block unicorn

Preface

The first two weeks of November 2025 exposed the fundamental flaws of decentralized finance (DeFi), which academia has been warning about for years. The collapse of Stream Finance's xUSD, followed by Elixir's deUSD and numerous other synthetic stablecoins, is not merely a series of isolated incidents of mismanagement. These events reveal structural issues within the DeFi ecosystem regarding risk, transparency, and trust.

What I observed during the Stream Finance collapse was not traditional smart contract exploits or oracle manipulation attacks, but rather a more concerning issue: a lack of fundamental financial transparency wrapped in decentralized language. When an external fund manager loses $93 million without effective oversight, triggering a $285 million cross-protocol chain reaction; when the entire "stablecoin" ecosystem loses 40-50% of its total locked value within a week while maintaining its peg, we must acknowledge a fundamental truth about the current state of decentralized finance: the industry has made no progress.

More precisely, the incentive mechanisms reward those who ignore lessons and punish those who act conservatively, shifting losses onto everyone when inevitable failures occur. There is an old saying in finance that is particularly apt here: if you don't know where the returns come from, you are the return itself. When certain protocols promise an 18% return through undisclosed strategies, while mature lending markets offer only 3-5%, the source of these returns is the principal of the depositors.

Stream Finance Mechanism and Contagion Pathway

Stream Finance positions itself as a yield optimization protocol, offering an annualized return of 18% on USDC deposits through its yield-stablecoin xUSD. Its claimed strategies include "delta-neutral trading" and "hedged market making," terms that sound complex but are rarely mentioned in practical operations. In contrast, at that time, mature protocols like Aave offered an annualized return of 4.8% on USDC deposits, while Compound was slightly above 3%. Despite widespread skepticism about returns three times the market rate, users deposited hundreds of millions of dollars. Before the collapse, the trading price of 1 xUSD was 1.23 USDC, reflecting its claimed compound returns. The asset management scale of xUSD peaked at $382 million, but data from DeFiLlama shows its peak total locked value was only $200 million, indicating that over 60% of the assets were in unverifiable off-chain holdings.

Yearn Finance developer Schlagonia revealed the true operational mechanism of Stream after its collapse, exposing a systematic fraud disguised as financial engineering. Stream implemented a recursive lending mechanism to create uncollateralized synthetic assets through the following process: users deposit USDC, and Stream exchanges it for USDT via CowSwap. USDT is then used to mint deUSD on Elixir, chosen for its high-yield incentive mechanism. These deUSD are bridged to blockchains like Avalanche and deposited in lending markets to borrow USDC, completing a cycle. So far, this strategy resembles standard collateralized lending, although its complexity and cross-chain dependencies are concerning. But Stream did not stop there. Stream did not merely use the borrowed USDC for additional collateral cycles; instead, it re-minted xUSD through its StreamVault contract, causing the supply of xUSD to far exceed any actual collateral backing. With only $1.9 million in verifiable USDC collateral, Stream minted $14.5 million in xUSD, resulting in a synthetic asset expansion ratio of up to 7.6 times relative to the underlying reserves. This is a fractional reserve banking system without reserves, without regulation, and without a lender of last resort.

The circular dependency with Elixir made the structure even more unstable. During the inflation of xUSD supply, Stream deposited $10 million in USDT into Elixir, thereby expanding the supply of deUSD. Elixir converted these USDT into USDC and deposited them in Morpho's lending market. By early November, the USDC supply on Morpho exceeded $70 million, with borrowings over $65 million, with Elixir and Stream being the two dominant participants. Stream held about 90% of the total supply of deUSD (approximately $75 million), while Elixir's support primarily came from loans provided to Stream by Morpho. These stablecoins were mutually collateralized, destined to collapse simultaneously. This is a form of "incest" in the financial realm, creating systemic fragility.

Industry analyst CBB publicly pointed out these issues on October 28, writing: "There are only about $170 million in funds supporting xUSD on-chain. They borrowed about $530 million from lending protocols. That’s a leverage ratio of 4.1 times. Moreover, many positions have poor liquidity. This is not yield farming; this is gambling." Schlagonia had warned the Stream team 172 days before the collapse that a five-minute check of their positions would reveal an inevitable failure. These warnings were public, specific, and accurate. However, users chasing yields, custodians seeking fee income, and protocols supporting the entire structure turned a blind eye.

On November 4, after Stream announced that an external fund manager had lost approximately $93 million in fund assets, the platform immediately suspended all withdrawals. With no redemption mechanism in place, panic quickly spread. Holders rushed to sell xUSD in the illiquid secondary market. Within just a few hours, xUSD plummeted 77%, dropping to about $0.23. This stablecoin, which once promised stability and high yields, evaporated three-quarters of its value in a single trading day.

Contagion Effect Data

According to data from the DeFi research institution Yields and More (YAM), the direct debt exposure related to Stream across the entire ecosystem reached $285 million. These exposures include: TelosC holding $123.64 million in loans collateralized by Stream assets (the largest single custodian exposure); Elixir Network lending $68 million through Morpho's private vault (accounting for 65% of deUSD's supporting funds); MEV Capital holding $25.42 million, including about $650,000 in bad debt due to oracles freezing the xUSD price at $1.26 while the actual market price plummeted to $0.23; Varlamore holding $19.17 million; Re7 Labs holding $14.65 million in one vault and $12.75 million in another; Enclabs, Mithras, TiD, and Invariant Group holding smaller exposures. Euler faced 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 may be more stablecoins/vaults affected," as the full scale of interconnections remains unclear weeks after the initial collapse.

Elixir's deUSD concentrated 65% of its reserves in loans to Stream through Morpho's private vault, leading to a 98% price drop within 48 hours, from $1 to $0.015, marking the fastest large stablecoin collapse since Terra's UST in 2022. Elixir processed redemptions for about 80% of deUSD holders other than Stream, allowing them to exchange for USDC at $1, thereby protecting most of its community members. However, this protective measure came at a significant cost, ultimately borne by Euler, Morpho, and Compound. Subsequently, Elixir announced it would completely cease all stablecoin products, acknowledging that trust had been utterly destroyed.

The broader market response indicated a systemic loss of confidence. According to data from Stablewatch, although most yield-bearing stablecoins maintained their peg to the dollar, their total locked value (TVL) still lost 40-50% within a week after the Stream collapse. This means approximately $1 billion in funds flowed out of protocols that did not collapse and had no technical issues. Users could not distinguish between legitimate projects and fraudulent ones, leading to a mass exodus. By early November, the total locked value (TVL) in DeFi had decreased by $20 billion. The market reflected a pervasive contagion risk rather than a reaction to the failure of specific protocols.

October 2025: A $60 Million Chain Liquidation Trigger

Less than a month before the collapse of Stream Finance, the crypto market experienced an on-chain forensic analysis revealing that it was not an ordinary collapse, but a precise attack exploiting known vulnerabilities, comparable in scale to institutional-level trading. From October 10 to 11, a meticulously planned $60 million market sell-off triggered oracle failures, leading to large-scale liquidations across the entire DeFi ecosystem. This was not excessive leverage but a design flaw in institutional-level oracles, whose attack patterns have been documented and public since February 2020.

The attack began at 5:43 AM UTC on October 10, when $60 million worth of USDe was sold into the spot market of a single exchange. In a well-designed oracle system, the impact of such behavior should be minimal, as multiple independent pricing sources would absorb these fluctuations through time-weighted averages, preventing manipulation. However, this oracle system adjusted the value of collateral (wBETH, BNSOL, and USDe) in real-time based on the manipulated exchange's spot price. Large-scale liquidations were immediately triggered. Millions of simultaneous liquidation requests overwhelmed the system's capacity, causing infrastructure overload. With API interfaces down, withdrawal requests piled up, and market makers could not quote prices in time. Liquidity vanished. This chain reaction continued to self-reinforce.

Attack Methods and Precedents

The oracle faithfully reported the manipulated price of a certain trading venue while prices in all other markets remained stable. Major exchanges showed that USDe was priced at $0.6567, and wBETH was at $430. The price deviations in other venues were less than 30 basis points from normal prices. The impact on on-chain liquidity pools was minimal. As Guy Young, founder of Ethena, pointed out during the entire event, "over $9 billion in on-demand stablecoin collateral was available for immediate redemption," proving that the underlying assets were not impaired. However, the oracle reported the manipulated price, and the system liquidated based on these prices, leading to the destruction of positions based on valuations that did not exist anywhere else in the market.

This is precisely the pattern that led to Compound's collapse in November 2020: at that time, the price of DAI on Coinbase Pro soared to $1.30 while all other trading platforms maintained a price of $1.00 for an hour, ultimately resulting in $89 million in liquidations. Although the trading platforms have changed, the vulnerabilities remain. The attack methods are similar to those that caused bZx to suffer an attack in February 2020 (manipulating the Uniswap oracle to steal $980,000), Harvest Finance in October 2020 (manipulating Curve to steal $24 million and triggering a $570 million run), and Mango Markets in October 2022 (manipulating across multiple platforms to steal $117 million). From 2020 to 2022, there were 41 recorded oracle manipulation attacks that stole $403.2 million. The industry's response has been slow and fragmented. Most platforms still use inadequately redundant, spot-trading-dominated oracles.

This amplification effect indicates that as market sizes grow, these lessons become increasingly important. In 2022, Mango Markets experienced a $5 million manipulation that led to a $117 million theft, amplifying by 23 times. In October 2025, a $60 million manipulation triggered a chain reaction, amplifying even more amounts. The attack patterns have not become more complex; the underlying system's scale has increased, but the fundamental vulnerabilities remain.

Historical Patterns: Failure Cases from 2020 to 2025

The collapse of Stream Finance is not a new phenomenon, nor is it unprecedented. The DeFi ecosystem has repeatedly faced stablecoin failures, each time exposing similar structural vulnerabilities. However, the industry continues to repeat the same mistakes, and the scale is growing larger. The recorded failure cases over the past five years exhibit a consistent pattern. Algorithmic stablecoins or partially collateralized stablecoins offer unsustainable yields to attract deposits. These yields do not come from actual income but are subsidized through token issuance or new deposits. The operation of the protocols is characterized by excessive leverage, opaque disclosures of actual collateral ratios, and circular dependencies—Protocol A supports Protocol B, which in turn supports Protocol A. Once any shock reveals potential insolvency issues or subsidies become unsustainable, a run is triggered. Users rush to exit, collateral values plummet, and liquidations surge, causing the entire system to collapse within days or even hours. This crisis can also spread to protocols that accepted the failed stablecoins as collateral or held positions within the ecosystem.

May 2022: Terra (UST/LUNA)

Loss: $45 billion in market capitalization evaporated within three days. UST is an algorithmic stablecoin pegged to LUNA through a mint-and-burn mechanism. The Anchor Protocol offered unsustainable yields of up to 19.5% on UST deposits, with about 75% of UST deposited on the platform to earn rewards. The system relied on continuous inflows of funds to maintain its peg to LUNA.

Trigger: On May 7, $375 million in withdrawals occurred on the Anchor platform, followed by a massive sell-off of UST, causing the token to decouple from LUNA. As users exchanged UST for LUNA to exit, the supply of LUNA surged from 346 million to over 6.5 trillion within three days, triggering a death spiral that ultimately drove both tokens' prices close to zero. This collapse left individual investors with nothing and led to the failure of several major cryptocurrency lending platforms, including Celsius, Three Arrows Capital, and Voyager Digital. Terra's founder, Do Kwon, was arrested in March 2023 and faced multiple fraud charges.

June 2021: Iron Finance (IRON/TITAN)

Loss: $2 billion in total value locked (TVL) plummeted to near zero within 24 hours. IRON was partially collateralized with 75% USDC and 25% TITAN. Unsustainable mining incentives offered annual interest rates of up to 1700% to attract deposits. When large holders began exchanging IRON for USDC, the sell-off pressure on TITAN created a self-reinforcing cycle. The price of TITAN collapsed from $64 to $0.00000006, destroying the collateral for IRON.

Lesson: Under pressure, partial collateralization is insufficient. When the collateral token itself enters a death spiral, the arbitrage mechanism fails under extreme stress.

March 2023: USDC

Decoupling: When $3.3 billion in reserves were trapped in the collapsed Silicon Valley Bank, the price of DAI fell to $0.87 (a 13% drop). This situation should not have occurred for a stablecoin that had regular audits and was "fully backed by fiat." DAI's price only recovered after the Federal Deposit Insurance Corporation (FDIC) invoked systemic risk exception clauses to guarantee deposits at Silicon Valley Bank.

Contagion Effect: Over 50% of DAI's collateral was USDC, and this event triggered DAI's decoupling, leading to over 3,400 automatic liquidations on the Aave platform, totaling a loss of $24 million. This indicates that even well-intentioned and regulated stablecoins face concentration risks and dependence on the stability of traditional banking systems.

November 2025: Stream Finance (xUSD)

Loss: Direct loss of $93 million, with total risk exposure across the ecosystem reaching $285 million.

Mechanism: Recursive lending created uncollateralized synthetic assets (actual collateral scaled up by 7.6 times). 70% of the funds were managed by an anonymous external manager through opaque off-chain strategies. No reserve proof.

Current Status: xUSD trading price is $0.07-$0.14 (87-93% below the pegged price), with almost zero liquidity. Withdrawals are indefinitely frozen. Multiple lawsuits have been filed. Elixir has completely ceased operations. The entire industry is selling yield-bearing stablecoins.

All cases exhibit some common failure patterns.

Unsustainable Yields: The return rates of Terra (19.5%), Iron (1700% annual yield), and Stream (18%) are disconnected from actual income.

Circular Dependencies: UST-LUNA, IRON-TITAN, and xUSD-deUSD all exhibit mutually reinforcing failure patterns, where the collapse of one inevitably leads to the collapse of another.

Opacity: Terra concealed the costs of its peg subsidies, Stream hid 70% of its operational activities off-chain, and Tether's reserve composition has also faced scrutiny.

Partial Collateralization or Self-Issued Tokens: Relying on volatile tokens or self-issued tokens can create a vicious cycle under pressure, as the value of collateral tends to drop precisely when it is needed for support.

Oracle Manipulation: Freezing or manipulating price information can prevent normal liquidations, turning price discovery into trust discovery, and continuously accumulating bad debts until the system collapses.

The information is clear: so-called stablecoins are not truly stable. They may appear stable, but once a shift occurs, it often takes just a few hours.

Oracle Failures and Infrastructure Collapse

At the onset of the Stream collapse, oracle issues were immediately exposed. As the actual market price of xUSD fell to $0.23, many lending protocols hard-coded oracle prices at $1.00 or higher to prevent chain liquidations. While this decision aimed to maintain market stability, it created a fundamental disconnect between market realities and protocol behaviors. This hard coding was not a technical failure but an intentional policy. Many protocols adopted manual updates to oracle prices to avoid triggering liquidations during short-term market fluctuations. However, when the price drop reflects actual insolvency rather than temporary market pressure, this method becomes completely ineffective.

Protocols face a dilemma.

Using Real-Time Prices: The risk is that manipulation and chain liquidations may occur during market volatility, as the painful lessons of October 2025 have shown.

Using Delayed Prices or Time-Weighted Average Prices (TWAP): This cannot address true insolvency and will accumulate bad debts, as demonstrated in the case of Stream Finance, where the oracle showed a price of $1.26 while the actual price was only $0.23, resulting in MEV Capital alone incurring $650,000 in bad debts.

Using Manual Updates: This introduces centralization, discretionary intervention, and the possibility of covering up insolvency by freezing oracles. All three methods have resulted in losses of hundreds of millions or even billions of dollars.

Infrastructure Capacity Under Pressure

In October 2020, Harvest Finance experienced an infrastructure collapse, leading users to flee after suffering a $24 million network attack, causing its total value locked (TVL) to plummet from $1 billion to $599 million. The lessons from this incident are clear: oracle systems must consider infrastructure capacity during stress events; liquidation mechanisms must implement rate limits and circuit breakers; exchanges must maintain additional capacity equivalent to ten times normal load. However, the events of October 2025 indicate that institutional users still have not learned this lesson. When millions of accounts faced simultaneous liquidations, and billions of dollars in positions were liquidated in less than an hour, the order book was left blank as all bids were exhausted and the system overloaded, preventing new bids from being submitted. The extent of the infrastructure collapse was as thorough as that of the oracle system. Although technical solutions exist, they have yet to be implemented because they would reduce efficiency under normal circumstances and result in losses of funds that could otherwise be converted into profits.

If you cannot determine the source of returns, you cannot achieve profits; instead, you will become the cost of others' gains. This principle is not complex. However, billions of dollars have been poured into those "black box" strategies because people prefer to accept comforting lies rather than confront uncomfortable truths. The next Stream Finance model is already in operation.

Stablecoins are not stable. Decentralized finance is neither decentralized nor secure. Unidentified sources of returns are not profits but thefts with a countdown. These are not personal opinions but verifiable facts validated by substantial costs. The only question is whether we will ultimately act based on known facts or spend another $20 billion to repeat the same mistakes. History suggests that the latter may still be inevitable.

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。