More Markets $9.3M Reserve Theft: The E-mode Assumption That Broke a Lending Protocol

The loot was $9.3 million in WFLOW. The mechanism was not a novel zero-day exploit, but a predictable failure of two now-standard DeFi primitives: liquid staking derivatives and E-mode. For those of us who spent 2017 auditing ICO whitepapers for structural flaws, the pattern feels disturbingly familiar. A small protocol, running on a niche chain, configures aggressive risk parameters to attract liquidity in a bull market. Then the market tests those parameters with surgical precision. This is not the story of a single hack. It is a case study in how the pursuit of capital efficiency in a bull market creates systemic fragility. Liquidity is the only truth in a volatile market. The rest is configuration, and configuration is where trust dies. The attack on More Markets, a lending protocol on the Flow blockchain, is a forensics lesson in how E-mode, a mechanism designed to maximize capital efficiency between correlated assets, becomes a weapon when the correlation assumption is held too loosely. More Markets is not a household name like Aave or Compound. It operates within the Flow ecosystem, the blockchain originally built for NFTs, which has struggled to establish a competitive DeFi landscape beyond its native assets. To bootstrap liquidity and differentiate itself, More Markets adopted an Aave V3-inspired architecture, including E-mode, or Efficiency Mode. This feature allows borrowers to access higher loan-to-value ratios when their collateral and borrowed asset are deemed to be highly correlated. The logic is sound: if I deposit ETH and borrow stETH, both assets essentially represent the same underlying value, so the liquidation risk is minimal. The efficiency gain is a feature. The problem arises when the protocol's risk oracle or its governance parameters classify assets as correlated when they are not truly so. In this case, the collateral was an Ankr liquid staking derivative, likely ankrFLOW, and the borrowed asset was WFLOW, the wrapped version of the Flow native token. The assumption, presumably, was that these two assets would move in tandem. The attack vector suggests that this assumption was fatally flawed, allowing an attacker to either manipulate the price of the collateral or exploit a pricing deviation to borrow more than the collateral actually warranted. The mechanics of the exploit, as inferred from industry-standard attack patterns, point to a classic oracle manipulation or pricing inefficiency. The attacker deposits the Ankr derivative as collateral. Due to E-mode, this deposit qualifies for a high LTV. The attacker then manipulates the price of the Ankr derivative, possibly through a liquidity pool with thin depth, inflating its value artificially. With the inflated collateral value, the attacker borrows the maximum amount of WFLOW, up to $9.3 million, and exits the protocol. The price of the collateral then normalizes, leaving the protocol with a bad debt position. The WFLOW is sold or bridged, and the protocol's reserve is depleted. This is not a novel technical breakthrough. It is the same playbook we have seen for years: an assumption about price stability is made, liquidity is thin, and the protocol lacks a robust price discovery mechanism. My 2022 report on the Terra Luna collapse highlighted similar concerns about algorithmic stablecoin pegs and the cascading effect of a single point of failure. The same principle applies here with LSDs. The attack is a direct consequence of a risk parameter being too loose for the actual liquidity profile of the asset. The core insight here is not that More Markets was hacked. It is that this attack was entirely predictable, and likely preventable, through rigorous pre-mortem analysis. Let us measure the risk dimensions. First, the oracle dependency. Did More Markets rely on a decentralized oracle network like Chainlink, or did it derive its price of the Ankr derivative from a DEX liquidity pool? If the latter, the medium to high confidence inference is that the protocol was vulnerable. A TWAP price with a sufficient time window could have mitigated the manipulation. A Chainlink feed, which aggregates from multiple high-liquidity sources, would have been more robust. The fact that the attack succeeded suggests that the price source was manipulable, a core failure in market microstructure. Second, the E-mode parameter configuration. Aave V3 is careful about which assets can be paired in E-mode, typically requiring a strong peg correlation, such as between USDC and DAI. Grouping an LSD with its underlying asset can be risky if the LSD trades at a significant premium or discount due to staking yields or liquidity issues. The protocol may have overlooked the fact that the staking derivative's market price can deviate substantially from its fair value, especially during times of high volatility or low liquidity. The E-mode allowed the attacker to leverage this deviation. The full scope of the vulnerability is best understood by examining the potential attack sequence step-by-step. First, the attacker sources a substantial amount of the Ankr liquid staking token. Given reduced liquidity on the Flow chain, this could be done via a flash loan from a cross-chain bridge or a decentralized exchange. Second, the attacker deposits this token into More Markets as collateral. Third, the attacker uses a separate pool or exchange to execute a large buy order, or series of orders, on the Ankr derivative, driving its market price up by a significant margin, say 10-20%. Fourth, because the protocol's oracle reads this inflated price, the collateral is now worth far more in the protocol's eyes. Fifth, the attacker borrows the maximum LTV for WFLOW, which, due to E-mode, might be as high as 90-97%. Sixth, the attacker swaps the borrowed WFLOW to a stablecoin or ETH and exits the chain. Seventh, the price of the Ankr derivative naturally falls back to its mean, and the protocol is left with collateral that is worth only a fraction of the debt. The loan is now underwater, and the bad debt is socialized across the protocol's other lenders. The $9.3 million represents real user deposits that have been effectively drained. The aftermath of this event is where the macro analysis becomes critical. In a bull market, events like this are often absorbed quickly. The psychological impact is real, but the market context is forgiving. However, for the Flow chain and for Ankr, the implications are broader. Ankr is a well-known liquid staking provider. This attack will cause users to question the reliability of its Flow-related derivatives. While the integrity of the underlying staked assets may not be compromised, the trading price and liquidity of the derivative will suffer. The market confidence in using these derivatives as DeFi collateral, particularly on smaller chains, will be shaken. For the Flow ecosystem, the attack is a significant setback. DeFi relies entirely on trust and security, and a localized event on a primary lending protocol can deter future builders and users. The ecosystem was already struggling to gain traction against the dominance of Ethereum, Solana, and other chains. This hack provides a negative signal that Flow's infrastructure is not yet hardened for serious financial applications. It is worth noting that this event may have a contagion effect on other lending protocols that are not on Flow. The risk is not specific to More Markets, but to the mechanism itself. Any protocol using E-mode with liquid staking derivatives as collateral is now under suspicion. The industry has already seen the fall of Terra, the collapse of FTX, and countless bridge hacks. The narrative of DeFi as an unsafe Wild West is reinforced, even if the actual damage is contained to $9.3 million, which is a modest sum compared to some of the $1 billion plus hacks in previous cycles. Despite the bearish narrative, there are structural lessons that market participants should internalize. The DeFi sector is often criticized for prioritizing growth over security. Attacks are not random; they are a direct consequence of misaligned incentives. When an unscrupulous actor can profit, and the cost is borne by passive liquidity providers, the incentive to find the weakest protocol is high. E-mode is a powerful tool, but it requires deep market analysis to configure correctly. The risk is not avoided; it is priced and hedged. The attack highlights the need for a more intelligent approach to risk management. For example, a protocol could implement a circuit breaker that pauses borrowing if the price of a collateral asset moves more than a certain percentage in a short period, or it could use a time-weighted average price to smooth out volatility. More importantly, audits and security reviews must go beyond simply checking for reentrancy and integer overflow. They must include what are called economic audits, seeking to manipulate the protocol parameters to understand the possible flash loan attacks and economic leverage vulnerabilities. My experience in 2020 with DeFi yield logic verification involved exactly this kind of modeling, identifying potential liquidity fragmentation risks if stablecoin pegs deviated from their target. The same methodology should be used to validate E-mode correlation assumptions. The contrarian angle in this event is that the $9.3 million incident might do more damage to the broader LSD narrative than to lending protocols themselves. Liquid staking derivatives are the backbone of a multi-billion dollar ecosystem, with Lido being the largest player. The promise of LSDs is that you can earn staking yield while remain composable with DeFi. This attack reveals a fatal caveat: you cannot blindly rely on the exchange price of an LSD as a proxy for its fundamental value. The premium or discount on these derivatives is a function of liquidity, and on a long-tail chain like Flow, that liquidity can be disastrously thin. The path forward is not to abandon E-mode or LSDs, but to recognize the limits of the assumptions that underpin them. The market will continue to move toward stricter risk parameters and more conservative asset listing policies. We are likely to see a flight to quality, with funds moving from small-chain DeFi back to the more liquid and battle-tested markets on Ethereum. The impact on Ankr is particularly concerning because the incident calls into question the verification of their reserve and the stability of its derivative tokens. This may accelerate the demand for real-time proof-of-reserves and robust oracle solutions for liquid staking derivatives. The use of a decentralized oracle, or a custom price feed that takes into account the total value locked in the liquidity pool, could have prevented this attack. In a bid to understand the market structure, let us evaluate the timing of the event within the broader liquidity cycle. Currently, in the 2026 bull market, there is significant liquidity flowing into risk assets, including crypto. However, this liquidity is concentrated in mainstream assets. The capital that flows into long-tail assets like WFLOW is often speculative and highly volatile. The attack on More Markets is unlikely to trigger a mass liquidation of the broader market, but it will likely lead to a repricing of risk for assets on smaller chains. The market will begin to hedge against the possibility of similar attacks by demanding higher yields for providing liquidity to these protocols, or by reducing the LTV for volatile collateral. This is the invisible cost of such events: it no longer about a one-time loss of $9.3 million, but about a permanent increase in the cost of capital for the affected ecosystem. The risk premium is higher. It is a tax on certainty, and certainty is hard to come by in decentralized finance. The legal and regulatory implications are also worth considering, albeit with lower confidence. DeFi protocols are often decentralized to a degree, but the founders and the development teams are not always fully protected. If the United States SEC determines that the protocol or its token constitutes a security, there could be potential lawsuits. However, an attack event itself is not technically a regulatory violation. The regulatory risk is more about how the protocol responds. If More Markets chooses to socialize the losses by minting new tokens or pausing withdrawals, this could be seen as an act that violates the terms of service for the users. But if it chooses to cover the losses from its own treasury, that is a different story. The distinction between a protocol, a DAO, and a company becomes crucial. Given the lack of information, we cannot fully assess the legal status. The attack will likely attract the attention of regulators, not for the code itself, but for the outcome of users losing assets. It provides another data point for regulators advocating for stricter control over open-source DeFi code, which is a dangerous precedent. We must separate the actor from the tool. The code was not the culprit. The configuration and the economic incentives were. The technical architecture of the Flow chain may also be a contributing factor to the success of the attack. Flow is designed for scalability, but its blockchain is not the most straightforward to integrate with typical Ethereum Virtual Machine (EVM)-based DeFi. This often leads to a fragmented DeFi ecosystem with lower liquidity and a more limited selection of tools. The interoperability issue makes it easier for an attacker to understand the local market dynamics and exploit the inefficiencies. If More Markets had access to the same composability and liquidity depth as on the main Ethereum network, the manipulation would be more difficult. The issue of chain-specific liquidity is a recurring theme. This is why the "omnichain app" narrative propagated by venture capital firms is often misguided. Users are not going to use a lending protocol on ten different chains if the liquidity is fragmented. They want the deepest liquidity, the best rates, and the highest security. This attack confirms that the most important thing is not how the network is structured, but how the risk is managed. The responses from the wider DeFi community will be critical in the next few days. Will we see an emergency proposal in Aave or Compound to re-evaluate the E-mode parameters for LSDs? Will we see decentralized oracles like Chainlink being integrated more deeply? The market tends to forget quickly, but the technical debt remains. If there is a lesson to be extracted from this event, it is that DeFi is not a purely software engineering challenge; it is a risk management challenge. I have built my framework over 18 years of observing markets, from the ICO boom of 2017 where 70% of tokens had no revenue model, to the DeFi summer of 2020, and the black swan events of 2022. The patterns are constant: leverage amplifies glitches. The $9.3 million withdrawn from More Markets is a reminder that the baseline of trust in decentralized systems is not measured by the number of audits, but by the robustness of the constraints on capital efficiency. We are building on the assumption that the market will always be rational, but this event proves that rational actor exploitation is the true irrationality. The next step for this sector is not to build more complex financial products, but to build a more exhaustive and predictive risk framework that includes the political and economic factors of each chain. The Flow chain community now faces a decision. Will it rally behind More Markets, contributing to a recovery plan? Or will it let the protocol fail, which would likely set the chain back years? The response of the team will determine whether this is a temporary setback or a terminal blow. In the short term, we should expect to see significant outflows from the protocol as users move their funds to safer venues. They will pay the security premium. This is why the risk not being avoided but being priced and hedged is a useful framework. The yield you earn from a small-chain lending protocol needs to compensate you for the tail risk of a total collapse. For WFLOW, the price will be volatile. The attacker will likely attempt to sell the stolen WFLOW, and given the relatively thin order books on smaller exchanges, it could see significant slippage, depressing the price for other holders. For Ankr, the impact on its token price and liquid staking derivatives supply will be negative. In the medium term, the security industry should see a boost. Every hack, no matter how small, pushes more protocols to spend on security audits and monitoring services. The question is whether they will spend the money on technical audits or economic audits. The immediate takeaway is a bleak one: capital efficiency without security is just a slower path to zero. The architecture of the network and the sophistication of the smart contract code are not enough. The market must adopt a new standard for what it means to be a "secure" DeFi protocol. That standard must include robust oracle design, conservative E-mode parameterization, and an audit process that actively attempts to break the protocol through economic manipulation. Risk is not avoided; it is priced and hedged. The $9.3 million stolen is not a tragedy because of the amount. It is a tragedy because it was entirely avoidable. The attacker did not even need a deep technical skill set; they only needed to understand the gap between the paper risks and the actual liquidity. That gap is where vulnerabilities live. We will see more of these events as long as the market rewards aggressive risk-taking over prudent risk management. However, we must not respond by over-regulating the code. The problem is not cryptography; the problem is the blind trust in parameters that have not been stress-tested. This is the lesson of More Markets. It is a signal for the entire industry. The biggest risk is not that we lose funds; it is that we ignore the warning and build the same flawed systems on a larger scale. Will the next bull run see a $100 million version of this event? It is not a question of if, but when, unless the foundation of trust is rebuilt with a more rigorous and skeptical methodology. The smart contracts execute, and they do not negotiate. The market, however, does negotiate, and it prices in the risk accordingly. The next time you look at a lending APY, remember the E-mode assumption that broke the protocol. And ask yourself: what is the assumed correlation, and what is the actual liquidity?