Editor | Wu Says Blockchain

October 11, 2025, was the day with the largest liquidation amount in cryptocurrency history, with total liquidations possibly reaching $20-30 billion. Wu Says interviewed the CEO of the quantitative trading team ZeroDivision, Hazelnut, to discuss the details of the market crash on October 11, 2025. On that day, altcoins sometimes dropped by 50% or more in a single day, with some coins even going to zero. Many market makers and trading teams had their margins breached, leading to forced liquidations and a chain reaction in the market that spiraled out of control. Some proactive market makers lost over 50% or even went bankrupt. The content of this article only represents the personal views of the interviewee and does not represent Wu Says' views.

Xiaoyuzhou:

https://www.xiaoyuzhoufm.com/episodes/692da1ecba2292550fb534f0

YouTube:

https://youtu.be/fCUcIyjlGyM

Introduction to the ZeroDivision Trading Team and Its Quantitative Trading Strategies

Cat Brother: Welcome to the Wu Says Non-Crypto Podcast. Today, we have a market maker who experienced the crash on October 11, the CEO of the ZeroDivision trading team, Hazelnut. First, could you introduce your team and the business you do?

Hazelnut: Hello everyone, I am the CEO of the ZeroDivision trading team. We focus on neutral strategies, asset management, and liquidity provision. We manage approximately $300 million in assets, with a monthly trading volume of $3-4 billion. Our main trading varieties are mainstream coins with good liquidity, and our position structure and risk management system are consistent across multiple exchanges, with the core being strict neutral hedging.

Our work involves acquiring a large amount of market data, such as trading records, orders on the order book, and market volatility. By analyzing this data, we use mathematical models for modeling and make predictions about short-term trends, followed by automated execution of trades. Unlike manual trading, our trading process is fully automated, and each trade is completed in an instant. Our profits come from capturing pricing differences in different trading markets for the same underlying asset in a neutral manner, the premiums and discounts between spot and derivatives, amplifying LST holdings in a neutral portfolio, options volatility arbitrage, predicting market trends, and providing liquidity to exchanges to earn fee rebates.

Early Stage of the Crash on October 11: Bitcoin Price Drop and Systemic Risk Warnings

Cat Brother: On October 11, when did you first notice something unusual? Can you briefly describe the process of that day?

Hazelnut: In the early hours of October 11, Bitcoin began to drop, and I noticed this trend. Around 5 AM, ADL was triggered on OKX, and we received an alert, prompting the team to quickly go online. Bitcoin fell several 5-minute bars in a row, dropping about 5-10 points, and other altcoins like USDe, BNSOL, and WBETH also plummeted.

As everyone knows, altcoins sometimes drop by 50% or more in a single day, with some coins going to zero. Many market makers and trading teams had their margins breached, leading to forced liquidations, and the market experienced a chain reaction that spiraled out of control.

Cat Brother: What measures did you take after discovering the problem?

Hazelnut: Our trading system enters a protection mode and stops opening new positions. However, the situation that day exceeded the risk control system's ability to respond; phenomena like USDe decoupling were beyond the risk control system's handling. The exchange also lacked sufficient liquidity, preventing us from closing all positions in a short time.

Cat Brother: So you were also powerless at that time?

Hazelnut: You could say that.

Cat Brother: Some community members mentioned that when some altcoins dropped to very low prices, they tried to buy the dip, but the exchange couldn't process their orders. Did you encounter this situation?

Hazelnut: We did not buy the dip, so we did not encounter exactly the same situation, but we did face some trading obstacles. The trading interface's latency significantly increased, and when Binance was conducting large-scale liquidations, accounts were also unable to place orders.

Cat Brother: You said that when Binance was liquidating, accounts couldn't place orders, right?

Hazelnut: Yes, the design of Binance's PM accounts is such that users cannot intervene during the liquidation process, and accounts cannot place orders.

Cat Brother: Since you deployed strategies across multiple exchanges, did you encounter different problems or sequences?

Hazelnut: Yes, we first triggered ADL on the FIL and DOGE trading pairs on OKX and noticed a net value drawdown in our accounts, but there was no exposure or naked position risk. The problem was the disappearance of liquidity on OKX, which caused us to hedge at unreasonable prices, resulting in losses. On Binance, the issue was that tokens like BNSOL and USDe lacked sufficient liquidity to support reasonable pricing, but Binance compensated us promptly afterward.

Detailed Explanation of Trading System Protection Mode and ADL Mechanism

Cat Brother: Back to October 11, you mentioned that you noticed market anomalies at that time. Did you not take any actions before the positions were liquidated, or were there reasons you couldn't act?

Hazelnut: Our trading is fully automated, and we usually do not intervene manually. When the market experiences a significant drop, the trading system automatically enters protection mode, and we do not open new positions. Since our strategy is neutral, normally, a downward market trend would not affect us. Existing positions would be influenced by funding rates and price differences, and we would choose the right time to close them.

Cat Brother: You mentioned the liquidation triggered by the ADL mechanism; can you explain this mechanism? Ordinary users might not be very familiar with this concept.

Hazelnut: ADL (Automatic De-leveraging) is designed to address situations where the market experiences a breach of margin and the insurance fund is insufficient. When a user is liquidated and the insurance fund cannot cover the losses, the exchange automatically matches the positions of profitable parties to compensate for the losses. Theoretically, this only triggers in extreme market conditions or when the insurance fund is exhausted.

Cat Brother: Did you encounter ADL liquidations on different exchanges? Before the liquidation, did you notice the liquidity situation on the order book?

Hazelnut: Under normal circumstances, exchanges take measures to protect liquidity. When extreme market conditions arise, we will reduce positions as long as the price difference allows. Some exchanges, like Binance, will implement price protection. If the order book has very poor liquidity, orders may not match with the order book.

Cat Brother: Doesn't this protection exacerbate liquidity issues? If there are originally few orders, and now trades cannot be matched, wouldn't liquidity become even scarcer?

Hazelnut: From my personal perspective, this protection is beneficial. It is based on the market index price, and most exchanges will take prices from other exchanges to form an average price. If a certain exchange's quote deviates too much from the market index price, it may indicate that there is a problem with the order book or that it is being manipulated. Therefore, exchanges not allowing these orders to be matched is a form of protection. In extreme situations, this protection is actually effective.

Cat Brother: In other words, on October 11, your trading behavior was the same as that of ordinary users, except that you executed trades automatically while ordinary users operated manually. You all faced the same issues without any preferential strategies?

Hazelnut: Everyone indeed faced the same liquidity shortage issue, but market makers usually have dedicated desks and channels to continue placing and executing orders, while ordinary retail investors might be unable to place orders due to front-end page or desk outages. Therefore, in this situation, the circumstances of market makers and retail investors may differ.

Overview of Losses on October 11 and Analysis of Exchange Differences

Cat Brother: What were your losses that day? How did the losses differ across platforms?

Hazelnut: Our loss percentages across platforms were roughly similar, around 20% to 30%, referring to the initial occurrence of the event.

Cat Brother: Were there many market makers facing similar issues? Since you have a neutral strategy, were there aggressive or directional market makers that went bankrupt directly in this market?

Hazelnut: Yes, indeed, some market makers went bankrupt in this situation.

Cat Brother: You just mentioned a specific trade, which was the FIL/USDT trade, right?

Hazelnut: Yes, it was the FIL coin-margined perpetual contract on OKX.

Cat Brother: Did this liquidity disappearance happen instantaneously, or was it a gradual process?

Hazelnut: We usually collect order book data for major coins around the clock. Afterward, we checked the order book snapshot on OKX, and liquidity disappeared instantaneously, lasting about 30 seconds. After the ADL execution ended, liquidity instantly recovered, which was a very strange phenomenon.

Cat Brother: Was it the same on other exchanges, like Binance or Bybit?

Hazelnut: There was no such situation on Binance and Bybit.

Cat Brother: Then you mentioned that there were also 20% to 30% losses on other platforms; what was the reason for that?

Hazelnut: On Binance, the decoupling of trading pairs like BNSOL and WBETH led to liquidity issues for some smaller coins. However, the liquidity for larger coins was fine. Bybit was similar, mainly affecting some smaller coins.

Cat Brother: Are these issues similar to the FIL coin-margined situation? Did the ADL mechanism trigger forced liquidations?

Hazelnut: The fundamental reason for the losses was the forced execution of the ADL mechanism, and on OKX, due to the lack of trading price protection, we had to cover positions at prices with almost no liquidity, resulting in huge losses. On Binance, the inability to trade during forced liquidations led to timing mismatches, and when positions were liquidated and the accounts were taken over again, market price fluctuations also caused losses. Similar situations occurred on Binance and Bybit; although larger coins had good liquidity, some smaller coins indeed had poor liquidity.

Analysis of the Reasons for Liquidity Disappearance and Market Makers' Responses

Cat Brother: Did you analyze the reasons for this sudden disappearance of liquidity afterward? I see that the mainstream conclusion in the community is that many people or market makers used USDe as collateral, which decoupled first, leading to a chain reaction of liquidations among some market makers, thereby affecting the liquidity of other trading pairs. What do you think about this situation? Do you think this logic is correct?

Hazelnut: This logic may hold true for Binance, but it doesn't apply as much to OKX and Bybit, as there were no particularly severe decouplings on those exchanges. More often, it was due to the strong unilateral movement of market prices, prompting many market makers to choose to hedge. Continuing to quote could lead to significant losses.

Cat Brother: Can it be understood that some market makers, like you, saw the anomalies in Bitcoin prices and chose to intervene manually, directly stopping the liquidity on the order book, while you did not intervene manually, and after they stopped, the prices became thin on liquidity, and then you "ran late"? If you had intervened manually like them, would you have avoided these losses?

Hazelnut: This analysis is very interesting, but there are indeed some differences. As you mentioned earlier, those market makers who exited early typically have no obligation to continue providing liquidity, or they are not required to provide liquidity until the last moment. It is understandable that market makers choose to hedge in such situations because they need to manage their inventory. In contrast, exchanges like Binance and Bybit must bear the consequences if market makers do not take on the responsibility of providing liquidity; they must cover the losses and take responsibility afterward. Unlike these market makers, we provide liquidity in a neutral manner. We do not need to manage inventory like those market makers; for instance, if we hold Bitcoin in spot, we would have a short position in Bitcoin perpetual contracts. This way, theoretically, price fluctuations do not affect our account net worth. Therefore, during a market downturn, if the exchange's risk control mechanisms operate normally, we would not incur losses. For example, we have experienced several historical crashes like March 12 and May 19, during which our strategy did not incur losses.

Market Makers' Liquidity Obligations and Hyperliquid Strategies

Cat Brother: You mentioned that some market makers have an obligation to provide liquidity. Does that mean they must provide liquidity at certain moments?

Hazelnut: Yes, especially for top market makers, the obligation to provide liquidity is stronger. For example, they must meet a certain order volume and order duration within specific price levels.

Cat Brother: What are the consequences if these obligations are not met? Will the exchange punish them? Do you have a cooperation agreement?

Hazelnut: Yes, especially for major market makers, they usually sign agreements that clearly outline the reward and punishment system.

Cat Brother: You mentioned the ADL mechanism and liquidation order. Are these details disclosed in the cooperation agreement in advance?

Hazelnut: From what I understand, most exchanges do not achieve this level of transparency. When ADL is executed, information such as the exchange's insurance fund and the scale of breaches is usually not disclosed.

Cat Brother: I heard that the liquidation process on Hyperliquid is relatively transparent. Will you deploy strategies on Hyperliquid?

Hazelnut: We do have some positions on Hyperliquid, but they are small, so we did not encounter such risks that day and cannot provide data analysis, and we quickly closed those positions without incurring losses.

Post-Event Reflection and Risk Control Strategy Adjustments

Cat Brother: You mentioned that after liquidity was restored, what measures did you take? Did you revert to a neutral strategy or make other adjustments?

Hazelnut: After the event, we immediately rechecked the exposures of all accounts, assessed the overall net worth losses, and backed up trading logs, fault logs, liquidity snapshots, and specific trading records. Then, we submitted the organized materials to various exchanges for claims. Other exchanges have handled it properly, but OKX is still processing it.

defioasis: After experiencing the events of October 11, did you adjust your risk assessment and strategies for the exchanges? Do you have new insights compared to before?

Hazelnut: Yes, this event made us deeply aware that systemic risks can occur. We did not anticipate that tokens like USDe, BNSOL, and WBETH would decouple, and we did not expect such significant liquidity risks for mainstream coins like FIL and DOGE. After the event, we reassessed systemic risks and made more conservative adjustments to our strategies, including limiting leverage usage. Additionally, although we previously thought that systemic risks like USDT decoupling were unlikely to happen, we have now implemented controls for the decoupling risks of USDT and USDC.

Cat Brother: Your losses seem to be related to the low liquidity of altcoins. Why not choose to trade Bitcoin or Ethereum, which have better liquidity?

Hazelnut: On Binance and OKX, we primarily trade larger coins, such as Bitcoin, Ethereum, and Solana, and FIL and DOGE are also part of our trading. It's just that the losses from FIL and DOGE were larger, while Bitcoin, Ethereum, and Solana did not incur losses.

Cat Brother: So, to recap, the main reason for the losses was the sudden disappearance of liquidity, which triggered ADL liquidations, especially for low liquidity altcoins that triggered liquidations earlier, resulting in larger losses, right?

Hazelnut: Yes, the overall logic is like that. When ADL occurs, the disappearance of liquidity leads to unfavorable prices when we hedge, with a significant buy-sell spread. This was the case on OKX, while on Binance, we could not timely cover our positions during forced liquidations, and after the account was taken over again, the prices had already moved, leading to losses. The losses caused by liquidity decoupling on Binance were particularly severe.

ADL Liquidation Mechanism and Responses to Neutral Strategies

Cat Brother: On Binance, you do not have the authority and must execute according to their trading logic. It seems that OKX also executes this way. What are the differences between the two?

Hazelnut: The difference is that on OKX, when ADL is executed, the account is still under control, and we can monitor the forced liquidation of short positions and almost immediately reopen those positions. However, on Binance, the account cannot hedge simultaneously during liquidation.

Cat Brother: From this logic, isn't OKX better? After liquidation, you can buy back, theoretically incurring no losses. Isn't your neutral strategy supposed to maintain neutrality?

Hazelnut: Taking OKX as an example, when ADL is executed, the selling price of FIL is 0.8, while the buying price is only 0.2. The exchange buys our short position at 0.8, and to maintain neutrality, we must sell the liquidated position again, but we can only sell at 0.2, resulting in a fourfold price difference.

Cat Brother: Didn't your automated strategy recognize this extreme difference? Does it strictly execute the neutral strategy?

Hazelnut: Under normal circumstances, our strategy is to buy spot and sell perpetual contracts. ADL only occurs on contracts; spot does not get affected by ADL. The difference of 0.8 and 0.2 is on the contract side, with the sell side at 0.8 and the buy side at 0.2. Our strategy sells contract positions to hedge and maintain neutrality.

Cat Brother: If the trading strategy could recognize such extreme situations, it would not execute the hedge. If other market makers' strategies could recognize this situation, they could avoid this problem, right?

Hazelnut: You mentioned two issues. First, if we identify insufficient liquidity, should we hedge? This is complex; looking back, we might judge that prices will continue to fall, and waiting to hedge until conditions worsen would only increase losses. Therefore, from a loss control perspective, we can only maintain neutrality. Secondly, other market makers may use CTA-type directional strategies, which differ from our trading model, facing different issues, making direct comparisons impossible.

Market Makers' Losses and Changes in Market Liquidity

Cat Brother: You said that the price difference of 0.8 and 0.2 should not occur under normal circumstances. From my understanding, such gaps should be arbitraged and smoothed out, and exchanges do not allow such situations to persist. The occurrence of a fourfold price difference indicates a withdrawal of liquidity, triggering ADL, combined with your neutral strategy, leading to losses. Am I correct in my understanding?

Hazelnut: The execution of ADL is not directly related to liquidity itself. This ADL happened to occur during a liquidity shortage. If there were no ADL, we would not incur losses. No matter how large the buy-sell spread is, we do not participate in trading. If there is no ADL execution, the exchange would not forcibly buy positions, and we would not forcibly sell. If the exchange has sufficient liquidity, even if ADL occurs, the losses would be minimal.

Cat Brother: So, you will judge strategies based on short-term market trends. If you misjudge in extreme market conditions, wouldn't that lead to greater losses?

Hazelnut: Yes, if the directional judgment is wrong, it can indeed lead to significant losses, especially with CTA strategies; a misjudgment in direction can lead to substantial drawdowns.

Cat Brother: How much do your market maker friends around you generally lose? What is the specific scale of losses?

Hazelnut: Some conservative market makers lost about 10% to 20%. Some aggressive market makers lost 40% to 50%, and some even went bankrupt.

Cat Brother: So, market makers generally lose 10% to 20%, and about 10% of market makers completely go bankrupt, is that right?

Hazelnut: Yes, especially for market makers with non-neutral strategies.

Cat Brother: After October 11, will market makers like you be more conservative in providing liquidity, leading to even thinner market liquidity?

Hazelnut: The current liquidity is indeed much worse than before October 11, and the overall market's open interest has also decreased. Most market makers have suffered significant losses, so they have become more conservative. Additionally, some market makers have gone bankrupt, leading to an inability to continue providing liquidity.

defioasis: After experiencing major risk events like March 12, is the extreme event of October 11 more severe than March 12?

Hazelnut: The severity of this event lies in the change in market dynamics. Liquidity is now extremely dispersed; previously, at a table of 10 people, perhaps 8 were trading the same coins, but now each person might be trading different coins. This makes it difficult for market makers to concentrate funds to provide liquidity for certain coins. Additionally, the influx of market capital has surged, and the scale of products is unprecedented. This time, there was also a decoupling of LST tokens, which occurred on Binance, a situation that many did not anticipate before.

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。