Decentralized Finance 基金管理人：百亿美元市场中的匿名赌徒

Written by: YQ

Compiled by: AididiaoJP, Foresight News

The rise of DeFi fund managers

In the past year and a half, a new type of financial intermediary has emerged in the DeFi space. These entities call themselves “risk managers,” “vault managers,” or “strategy operators.” They manage billions of dollars in user deposits on protocols like Morpho (approximately $7.3 billion) and Euler (approximately $1.1 billion), responsible for setting risk parameters, selecting collateral types, and deploying yield strategies. They take a performance fee of 5% to 15% from the generated yields.

However, these entities operate without licenses, are unregulated, do not require mandatory disclosure of qualifications or performance records, and often hide their true identities.

The Stream Finance crash event in November 2025

The collapse of Stream Finance in November 2025 fully exposed the fatal flaws of this architecture under pressure. This incident triggered a chain loss of up to $285 million throughout the ecosystem.

Several fund managers, including TelosC (USD 123.64 million), Elixir (USD 68 million), MEV Capital (USD 25.42 million), and Re7 Labs (two vaults totaling USD 27.4 million), have highly concentrated user deposits with a single counterparty. This counterparty operates with a leverage of up to 7.6 times with only USD 1.9 million in real collateral.

Warning signals have long been present and are very specific. Crypto KOLCBB publicly disclosed its leverage ratio on October 28. Yearn Finance directly warned the Stream team 172 days before the collapse. However, these warnings were ignored because the existing incentive structure precisely encouraged such neglect.

Comparison with traditional financial intermediaries

The DeFi fund manager model follows the methods of traditional finance but abandons the accountability mechanisms that were established after centuries of painful lessons.

Traditional banks or brokers face capital requirements, registration obligations, fiduciary responsibilities, and regulatory scrutiny when managing client funds. In contrast, DeFi fund managers are driven solely by market incentives when managing client funds. Market incentives promote asset accumulation and yield maximization rather than risk management.

The agreements that support these managers claim to be “neutral infrastructure,” earning fees from activities while completely avoiding responsibility for the risks generated by those activities. This position is fundamentally untenable. Traditional finance abandoned this notion decades ago after experiencing multiple disasters, as the harsh lessons indicate: intermediaries that earn fees cannot be completely exempt from liability.

The duality of a permissionless architecture

Morpho and Euler operate as permissionless lending infrastructures. Anyone can create vaults, set risk parameters, choose collateral, and start attracting deposits. The protocol provides the foundation for smart contracts and earns fees from it.

This architecture has its advantages:

Promoting Innovation: Eliminated the potential vicious competition that may hinder new funding operational methods due to unfamiliarity or competitive relationships.

Enhancing Inclusivity: Providing services to participants excluded by traditional systems.

Enhanced transparency: An auditable record of all transactions has been created on the chain.

However, this architecture also brings fundamental problems, which were exposed during the incident in November 2025:

No admission review: Unable to guarantee the quality of the manager.

No registration required: Managers cannot be held accountable in case of failure.

No identity disclosure: Managers can accumulate losses under one name and then change names to start over.

No capital requirements: The manager has no actual interests involved other than reputation, and reputation can be easily discarded.

As pointed out succinctly by Ernesto Boado, founder of BGD Labs: the manager is “selling your brand to gamblers for free.” The protocol generates revenue, the manager earns fees, while the users bear all the losses in the inevitable failures.

Typical failure mode: Bad money drives out good money

Stream Finance perfectly highlights the specific failure modes arising from permissionless architectures. Since anyone can create a vault, managers can only compete for deposits by offering higher yields. Higher yields come either from genuine alpha returns (rare and hard to sustain) or from higher risks (common and catastrophic once they occur).

Users see “18% annualized return” and stop questioning, assuming that the so-called “risk manager” experts have done their due diligence. Managers see the opportunity for fee income and accept risks that prudent risk management should reject. The protocol sees the growth in total locked value and fee income and chooses not to intervene, because the “permissionless” system should not have restrictions.

This competition has led to a vicious cycle: conservative curators have low returns and few deposits; aggressive curators have high returns, many deposits, and make a fortune in fees, until disaster strikes. The market cannot distinguish between sustainable profits and unsustainable risks before failures occur. By then, the losses are borne collectively by all participants, while the managers are almost unaffected, aside from a reputation that can be easily discarded.

Conflict of interest and incentive failure

The manager model embeds fundamental conflicts of interest, making failures similar to Stream Finance almost inevitable.

Objective divergence: Users pursue security and reasonable returns, while managers pursue fee income.

Risk mismatch: This type of target discrepancy is most dangerous when the potential for returns requires taking on risks that users would normally refuse.

The case of RE7 Labs is highly educational. During their due diligence before integrating xUSD, they correctly identified “centralized counterparty risk” as a hidden danger. Stream centralized risk in an anonymous external fund manager with a completely opaque position and strategy. RE7 Labs is aware of the risks but still pushed for integration citing “strong user and network demand.” The temptation of fee income overshadowed the risks to user funds. When funds were lost, RE7 Labs only suffered reputational damage, while users bore 100% of the financial loss.

This incentive structure not only mismatches but also actively punishes prudent behavior:

Managers who refuse high-risk, high-reward opportunities will see deposits flow to competitors who accept the risks.

The prudent manager has low fees, which appear to perform poorly.

The reckless hanging of your fees is high, with more deposits until exposed. The huge fees earned during this period can still be retained.

Many managers invest user funds in xUSD positions without adequate disclosure, exposing depositors to up to 7.6 times leverage and off-chain opacity risks without their knowledge.

Asymmetric fee structure

Managers typically take a performance fee of 5%-15% from the profits. This may seem reasonable, but it is actually highly asymmetric:

Share profits: The manager shares the upward profits.

No loss exposure: There is no corresponding risk exposure to downside losses.

Example: A vault with a deposit of 100 million dollars that generates a 10% return earns the manager (with a 10% performance fee) 1 million dollars. If the manager takes on double the risk to achieve a 20% return, they can earn 2 million dollars. If the risk materializes, the user loses 50% (50 million dollars) of their principal, while the manager only loses future income from the vault, having already pocketed the earned fees.

Conflict of interest in the agreement

The protocol has its own conflicts of interest when dealing with manager failures. Morpho and Euler earn fees from treasury activities and have the incentive to maximize activity levels, which means allowing high-yield (high-risk) treasuries that can attract deposits to exist. They claim to be “neutral” and assert that permissionless systems should not have restrictions. However, they are not neutral; they profit from the activities they facilitate.

Traditional financial regulators recognized centuries ago that entities profiting from intermediary activities cannot be completely exempt from the risks generated by those intermediaries. Brokers who earn commissions have obligations to their clients, and DeFi protocols have not yet accepted this principle.

Accountability vacuum

Traditional finance: Loss of customer funds may trigger regulatory investigations, license revocation, civil liability, and even criminal prosecution. This has deterred reckless behavior in advance.

DeFi fund managers: They face only reputational damage for losing client funds, and can often rebrand to start anew. There is no regulatory jurisdiction, no fiduciary duty (legal status unclear), and no civil liability (identity unknown + service terms disclaimer).

March 2024 Morpho Incident: Approximately $33,000 loss due to oracle price deviation. When users sought accountability, the protocol, administrators, and oracle providers passed the buck to each other, with no one taking responsibility or offering compensation. Though this matter seems minor, it established a precedent of “loss occurs, no one is responsible.”

This accountability vacuum is deliberately designed, rather than a result of negligence. The agreement avoids responsibility by exempting service terms, emphasizing “uncontrolled behavior without permission,” and placing governance in loosely regulated foundations/DAOs. Legally, this is beneficial for the agreement, but it creates an environment of moral hazard where managing billions of dollars of user funds can be done without accountability: profits are privatized, and losses are socialized.

Anonymity and Accountability

Many managers operate anonymously or under pseudonyms, and while they claim to prioritize safety and privacy, they directly undermine accountability:

Cannot be held legally liable.

Cannot be prohibited from operation due to failure records.

Unable to impose professional or reputational sanctions that follow the true identity.

In traditional finance, even without regulation, those who destroy client funds still face civil liability and reputational tracking, while DeFi fund managers have neither.

Black Box Strategy and Blindly Following Authority

The manager boasts of being a risk management expert, but in November 2025, it showed that many lack the necessary infrastructure, expertise, or even willingness.

Traditional institutions: 1-5% of employees are dedicated to risk management, with independent committees, oversight teams, stress testing, and scenario analysis requirements.

DeFi fund managers: often small teams or individuals, focusing on yield generation and asset accumulation.

The details of the strategy are rarely disclosed meaningfully. Terms like “delta-neutral trading” and “hedging market making” sound sophisticated but do not reveal actual positions, leverage, counterparty risks, or risk parameters. The opacity under the guise of a “protection strategy” is, in fact, a breeding ground for fraud and recklessness until it is exposed.

The opacity of Stream Finance has reached a disaster level: it claims a total locked value of $500 million, but only $200 million is verifiable on-chain, with the remaining $300 million held by “external fund managers” whose identities, qualifications, strategies, and risk controls are all undisclosed. The actual positions and leverage hidden under the terminology are unknown. Post-analysis reveals that it created 7.6 times synthetic expansion through recursive lending with $1.9 million of real collateral, leaving depositors completely in the dark, unaware that its “stablecoin” is supported by infinitely recursive borrowed assets rather than real reserves.

The danger of blindly following authority lies in causing users to give up independent judgment. The RE7 Labs case shows that even when due diligence identifies risks, commercial incentives can overpower the correct conclusions. This is worse than incompetence; it is the ability to identify but choosing to ignore due to incentives.

Proof of Reserve: Technologically mature but rarely implemented

Verifiable reserve proof technologies (such as Merkle trees and zero-knowledge proofs) have matured for decades, being efficient and capable of protecting privacy. Stream Finance has not implemented any reserve proof technology, which is a deliberate choice for opacity, allowing fraud to persist for months despite multiple public warnings. The protocol should require custodians managing large deposits to provide reserve proofs. The absence of reserve proof should be treated as a bank refusing external audits.

Event evidence of November 2025

The collapse of Stream Finance is a complete case of the failure of the manager model, highlighting all the issues: insufficient due diligence, conflicts of interest, ignoring warnings, lack of transparency, and no accountability.

Failure Timeline

172 Days Before the Crash: Schlagonia Analyzes and Directly Warns that the Stream Structure is Doomed. A 5-Minute Analysis Reveals a Fatal Problem: 170 Million On-Chain Collateral Supports 530 Million in Loans (4.1 Times Leverage), the Strategy Involves Recursive Lending Creating Circular Dependencies, and Another 330 Million in Total Locked Value is Completely Off-Chain and Opaque.

October 28, 2025: CBB publicly issues specific warnings, outlining leverage and liquidity risks, directly calling it “degenerated gambling.” Other analysts follow up.

Warning Ignored: Managers like TelosC, MEV Capital, and Re7 Labs still hold large positions and accumulate. Acting on the warning means reducing positions and fee income, which can make them appear to perform poorly in competition.

November 4, 2025: Stream announces that external fund managers have incurred losses of approximately $93 million. Withdrawals are suspended, xUSD plummets by 77%, and Elixir's deUSD (65% of reserves lent to Stream) crashes by 98%. The total contagion risk reaches $285 million, with Euler's bad debts at about $137 million, and over $160 million in funds frozen.

DeFi Fund Managers vs. Traditional Brokers

The comparison aims to reveal the accountability mechanisms missing in the curator model, and does not imply that traditional finance is perfect or that its regulations should be directly copied. Traditional finance has its flaws, but its accountability mechanisms developed through costly lessons have been explicitly discarded by the curator model.

Technical advice

The manager model does have its advantages: it enhances capital efficiency by setting parameters through professionals; allows experimentation to promote innovation; and lowers barriers to improve inclusivity. These benefits can be retained while addressing accountability issues. Recommendations are based on five years of failed DeFi experience:

Mandatory identity disclosure: Managers of large deposits (such as those over 10 million) are required to disclose their true identity to the protocol or an independent registration agency. Detailed privacy is not publicly disclosed, but it must ensure accountability in the event of fraud or gross negligence. Anonymous operations are incompatible with the large-scale management of others' funds.

Capital Requirements: Managers must hold risk capital and will incur losses when the treasury losses exceed the threshold (e.g., 5% of deposits). This can align their interests with those of users, such as providing collateral or holding a subordinate share of their proprietary treasury to absorb first losses. The existing risk-free capital structure creates moral hazard.

Mandatory Disclosure: Managers are required to disclose strategies, leverage, counterparty risks, and risk parameters in a standardized format. “Protecting proprietary strategies” is often an excuse; most strategies are known yield farming variants. Real-time disclosure of leverage and concentration without loss of alpha helps users understand the risks.

Reserve Proof: The protocol should require the curator managing large deposits to provide a reserve proof. Mature cryptographic technology can verify solvency and reserve ratios without disclosing strategies. Those without reserve proof should be disqualified from management. This requirement can prevent Stream from operating with 300 million unverifiable off-chain heads.

Concentration Limits: The protocol should enforce concentration caps at the smart contract level (e.g., single counterparty risk exposure of 10-20%) to prevent overconcentration. The lesson learned is that Elixir lending 65% of its reserves to Stream means it must be affected.

Accountability of Agreements: Agreements that earn fees from the activities of managers should bear some responsibility. For example, setting aside an insurance fund from the agreement fees to compensate users for losses, or excluding individuals with poor records or insufficient disclosures from the list of officials. The current model where agreements profit while being completely exempt from liability is economically unreasonable.

Conclusion

The current implemented management model is an accountability vacuum, with billions of dollars in user funds managed by entities that are unbound by substantial constraints and face no significant consequences for failure. This does not deny the value of the model itself, as capital efficiency and professional management do have merit. Rather, it emphasizes that this model needs to incorporate the accountability mechanisms that traditional finance has developed through painful lessons.

DeFi can develop mechanisms suitable for its characteristics, but it cannot simply abandon accountability and expect results to be better than those of traditional finance before accountability mechanisms were established. The existing structure is doomed to repeat failure. Failure will continue until the industry accepts that intermediaries who earn fees cannot be completely exempt from the risks they create.