Author: @BlazingKevin_, the Researcher at Movemaker

In the cryptocurrency asset market, traders often encounter two typical problems: first, there is a significant gap between the highest bid and the lowest ask price of the target trading token; second, after submitting a large market order, the asset price experiences drastic fluctuations, causing the execution price to deviate significantly from expectations, resulting in high slippage costs. Both phenomena are caused by the same fundamental factor—insufficient market liquidity. The core market participants who systematically address this issue are market makers.

The precise definition of a market maker is a professional quantitative trading firm whose core business is to continuously and simultaneously submit dense buy (Bid) and sell (Ask) quotes around the current market price of an asset on the exchange's order book.

Their fundamental function is to provide continuous liquidity to the market. Through bilateral quoting behavior, market makers directly narrow the bid-ask spread and increase the depth of the order book. This ensures that the buying and selling intentions of other traders can be matched with orders instantly at any point in time, allowing trades to be executed efficiently and at a fair price. As compensation for this service, market makers' profits come from the small spreads captured in massive trades and the rebates paid by exchanges to incentivize liquidity provision.

The market conditions of 1011 have made the role of market makers the focus of market discussions. When prices exhibit extreme volatility, a key question arises: Are market makers passively triggering a chain liquidation, or are they actively withdrawing liquidity quotes as risks escalate?

To analyze the behavioral patterns of market makers in similar situations, it is necessary to first understand the basic principles of their operations. This article aims to systematically answer the following core questions:

1. What is the business model that market makers rely on for profit?
2. What quantitative strategies do market makers adopt to achieve their business objectives?
3. When market volatility increases and potential risks arise, what risk control mechanisms do market makers initiate?

By clarifying the above questions, we will be able to infer more clearly the behavioral logic and decision-making trajectory of market makers in the 1011 market.

Basic Profit Model of Market Makers

1.1 Core Profit Mechanism: Spread Capture and Liquidity Rebates

To understand the behavior of market makers in the market, one must first know their most fundamental source of profit. Market makers provide continuous bilateral quotes (i.e., "making a market") on the exchange's order book, and their profits mainly consist of two parts: capturing the bid-ask spread and earning liquidity rebates from the exchange.

To illustrate this mechanism, we construct a simplified contract order book analysis model.

Source: Movemaker

Assume there is an order book with the following distribution of buy and sell orders:

• Bids: densely distributed at $1000.0, $999.9, $999.8, etc.
• Asks: densely distributed at $1000.1, $1000.2, $1000.3, etc.

At the same time, we set the following market parameters:

• Taker Fee: 0.02%
• Maker Rebate: 0.01%
• Minimum Price Increment: $0.1
• Current Spread: The difference between the best bid ($1000.0) and the best ask ($1000.1) is $0.1.

1.2 Trading Process and Cost-Benefit Analysis

Now, we will break down the profit process of market makers through a complete trading cycle.

Step 1: Market Maker's Buy Order is Passively Executed (Taker Sells)

Source: Movemaker

• Event: A trader (Taker) sells a contract at market price, and this order is matched with the best limit buy order on the order book, which is the buy order placed by the market maker at $1000.0.
• Nominal Cost: From the transaction record, it appears that the market maker has established a long position in one contract at $1000.0.
• Effective Cost: However, since the market maker is the liquidity provider (Maker), this transaction incurs no fees and earns a rebate of 0.01% from the exchange. In this case, the rebate amount is $1000.0 * 0.01% = $0.1. Therefore, the actual cash outflow (effective cost) for the market maker establishing this long position is: $1000.0 (nominal cost) - $0.1 (rebate) = $999.9.

Step 2: Market Maker's Sell Order is Passively Executed (Taker Buys)

• Event: A trader (Taker) buys a contract at market price, and this order is matched with the best limit sell order on the order book, which is the sell order placed by the market maker at $1000.1. This action closes the long position established by the market maker in Step 1.
• Nominal Revenue: The transaction record shows that the market maker sells at $1000.1.
• Effective Revenue: Similarly, as a liquidity provider, the market maker again earns a rebate of 0.01% on this sell transaction, amounting to $1000.1 * 0.01% ≈ $0.1. Therefore, the actual cash inflow (effective revenue) for the market maker closing the position is: $1000.1 (nominal revenue) + $0.1 (rebate) = $1000.2.

1.3 Conclusion: Composition of Real Profit

By completing this buy-sell cycle, the total profit for the market maker from a single transaction is:

Total Profit = Effective Revenue - Effective Cost = $1000.2 - $999.9 = $0.3

Thus, it is evident that the market maker's real profit is not merely the $0.1 nominal spread visible on the order book. The composition of their real profit is:

Real Profit = Nominal Spread + Buy Rebate + Sell Rebate

$0.3 = $0.1 + $0.1 + $0.1

This model of accumulating small profits through countless repetitions of the above process in high-frequency trading constitutes the most basic and core profit model of market maker operations.

Dynamic Strategies and Risk Exposure of Market Makers

2.1 Challenges to the Profit Model: Directional Price Movements

The effectiveness of the basic profit model mentioned above relies on the premise that market prices fluctuate within a narrow range. However, when the market experiences a clear one-sided directional move, this model faces severe challenges and exposes market makers to a core risk—adverse selection risk.

Adverse selection refers to the situation where informed traders selectively execute trades against market makers' quotes that have not yet been updated and are at "wrong" price levels, causing market makers to accumulate unfavorable positions.

2.2 Scenario Analysis: Strategic Choices in Response to Price Declines

To illustrate this, we continue with the previous analysis model and introduce a market event: the fair price of the asset rapidly drops from $1000 to $998.0.

Source: Movemaker

Assuming the market maker only holds a long contract established in a previous trade with an effective cost of $999.9. If the market maker takes no action, their buy orders placed around $1000.0 will present a risk-free profit opportunity for arbitrageurs. Therefore, once a directional price movement is detected, the market maker must respond immediately, with the primary action being to actively cancel all buy orders close to the old market price.

At this point, the market maker faces a strategic choice, primarily among the following three options:

• Option 1: Close Position Immediately, Realizing a Loss The market maker can choose to sell their long contract at market price immediately. Assuming the execution occurs at $998.0, the market maker must pay a 0.02% taker fee.

  Loss = (Effective Cost - Exit Price) + Taker Fee

  Loss = ($999.9 - $998.0) + ($998.0 × 0.02%) ≈ $1.9 + $0.2 = $2.1

  The purpose of this option is to quickly eliminate risk exposure, but it results in an immediate and certain loss.

• Option 2: Adjust Quotes, Seek Better Exit Price The market maker can lower their sell quote to a new market fair price, such as $998.1. If this sell order is executed, the market maker will receive a rebate as the maker.

  Loss = (Effective Cost - Exit Price) - Maker Rebate

  Loss = ($999.9 - $998.1) - ($998.1 × 0.01%) ≈ $1.8 - $0.1 = $1.7

  The aim of this option is to exit the position with a smaller loss.

• Option 3: Widen the Spread, Manage Existing Position The market maker can adopt an asymmetric quoting strategy: adjusting the sell quote to a relatively unattractive price (e.g., $998.8), while placing new buy orders at lower prices (e.g., $998.0 and $997.9). The goal of this strategy is to manage and reduce the average cost of the existing position through subsequent trades.

2.3 Strategy Execution and Inventory Risk Management

Assuming under a "single market maker" market structure, due to their absolute pricing power, the market maker is likely to choose Option 3 to avoid realizing an immediate loss. In this option, since the sell order price ($998.8) is significantly above the fair price ($998.0), its execution probability is low. Conversely, the buy orders closer to the fair price ($998.0) are more likely to be executed by sellers in the market.

Step 1: Lower Average Cost by Increasing Holdings

• Event: The market maker's buy order at $998.0 is executed.
• Effective Cost of New Position: $998.0 - (998.0 × 0.01%) ≈ $997.9
• Updated Total Position: The market maker now holds two long contracts, with a total effective cost of 999.9 + $997.9 = $1997.8.
• Updated Average Cost: $1997.8 / 2 = $998.9

Step 2: Adjust Quotes Based on New Cost

Through the above operations, the market maker successfully lowers the breakeven point of their long position from $999.9 to $998.9. Based on this lower cost basis, the market maker can now seek selling opportunities more aggressively. For example, they can significantly lower the sell quote from $998.8 to $998.9, achieving breakeven while significantly narrowing the spread from the previous $1.8 ($999.8 - $998.0) to $0.8 ($998.8 - $998.0) to attract buyers.

2.4 Limitations of the Strategy and Exposure to Risks

However, this strategy of averaging down through increasing holdings has obvious limitations. If the price continues to decline, for example, from $1000 to $900, the market maker will be forced to keep increasing their holdings while incurring continuous losses, and their inventory risk will escalate sharply. At that point, continuing to widen the spread will lead to a complete halt in trading, creating a vicious cycle, ultimately forcing them to liquidate at a significant loss.

This raises a deeper question: How do market makers define and quantify risk? What core factors are associated with varying degrees of risk? Answering these questions is key to understanding their behavior in extreme markets.

Core Risk Factors and Dynamic Strategy Formulation

The profit model of market makers essentially involves taking on specific risks in exchange for returns. Their losses primarily stem from significant adverse deviations in asset prices in the short term that are unfavorable to their inventory positions. Therefore, understanding their risk management framework is crucial for analyzing their behavioral logic.

3.1 Identification and Quantification of Core Risks

The risks faced by market makers can be summarized into two interrelated core factors:

1. Market Volatility: This is the primary risk factor. Increased volatility means that the likelihood and magnitude of price deviations from the current mean are both rising, directly threatening the value of the market maker's inventory.
2. Speed of Mean Reversion: This is the second key factor. After a price deviation occurs, whether it can return to equilibrium levels in a short time determines whether the market maker can ultimately profit by averaging down costs or will fall into continuous losses.

A key observable indicator for assessing the likelihood of mean reversion is trading volume. In the author's article published on April 22 of this year, titled "Review of Intensified Market Divergence: Does the Rebound Turn into a Reversal or a Second Distribution in a Downtrend?", the theory of marbles in the order book was mentioned. Different price levels of limit orders form layers of varying thickness based on order volume, and a fluctuating market behaves like a marble. We can view the limit orders at different price levels on the order book as "liquidity absorption layers" with different thicknesses.

Short-term price fluctuations in the market can be seen as the force of an impacting marble. In a low trading volume environment, the impact force is weak, and prices are usually confined to narrow movements between the densest liquidity layers. In a high trading volume environment, the impact force increases, capable of penetrating multiple layers of liquidity. The consumed liquidity layers are difficult to replenish instantly, especially in a one-sided market, which can lead to prices continuously moving in one direction, reducing the probability of mean reversion. Therefore, the trading volume per unit time is an effective proxy indicator for measuring the strength of this impact force.

3.2 Dynamic Strategy Parameterization Based on Market Conditions

Source: Movemaker

Based on the performance of volatility over different time scales (intraday vs. daily), market makers dynamically adjust their strategy parameters to adapt to different market environments. Their basic strategies can be summarized into the following typical states:

• In a Stable Market, when both intraday and daily price fluctuations are low, market makers' strategies are highly aggressive. They will use large orders with extremely narrow spreads, aiming to maximize trading frequency and market share to capture as much volume as possible in a low-risk environment.

• In a Range-Bound Market, when prices exhibit high intraday volatility but low daily volatility, market makers have high confidence in short-term mean reversion. Therefore, they will widen the spread to capture higher profits per trade while maintaining a large order size, ensuring they have enough "ammunition" to average down costs during price fluctuations.

• In a Trending Market, when intraday price fluctuations are stable but daily prices show a clear one-sided trend, market makers' risk exposure increases sharply. At this point, strategies shift to defense. They will use extremely narrow spreads and small orders, aiming for quick transactions to capture liquidity and exit quickly to avoid losses when trends are unfavorable to their inventory.

• In an Extremely Volatile Market (Crisis State), when both intraday and daily volatility increase significantly, risk management takes precedence for market makers. Strategies become extremely conservative, and they will significantly widen spreads and use small orders to manage inventory risk cautiously. In such high-risk environments, many competitors may exit, leaving potential opportunities for market makers capable of managing risks.

3.3 Core of Strategy Execution: Fair Price Discovery and Spread Setting

Regardless of the market state, the execution of market maker strategies revolves around two core tasks: determining the fair price and setting the optimal spread.

1. Determining the Fair Price: This is a complex issue with no single correct answer. If the model is incorrect, market makers' quotes will be "eaten" by more informed traders, leading to systematic accumulation of losing positions. Common foundational methods include using aggregated index prices from multiple exchanges or taking the mid-price of the current best bid and ask. Ultimately, regardless of the model used, market makers must ensure their quotes are competitive enough to effectively clear inventory. Holding large amounts of one-sided positions for extended periods is the primary cause of significant losses.

2. Setting the Optimal Spread: The difficulty of setting spreads is even greater than that of discovering fair prices, as it is a dynamic, multi-party game. Overly aggressive narrowing of spreads can lead to a "competitive equilibrium trap": while it may capture the optimal quoting position, profit margins are compressed, and once prices move, it is easy to be executed first by arbitrageurs. This requires market makers to build a more intelligent quantitative framework.

3.4 A Simplified Quantitative Framework for Optimal Spread

To clarify its internal logic, we reference a simplified model constructed by author David Holt on Medium, deriving the optimal spread under highly idealized assumptions.

• A. Core Assumptions and Volatility Calculation: Assume that market prices follow a normal distribution in the short term, with a sampling period of 1 second, examining the past 60 seconds of sample data. It is calculated that the standard deviation (σ) of the marked prices relative to the average mid-price within this sample is $0.4. This means that for about 68% of the time, the price in the next second will fall within the range of [mean - $0.4, mean + $0.4].

Source: Idrees

• B. Correlating Spread, Probability, and Expected Revenue: Based on this, we can deduce the probabilities of different spreads being executed and calculate their expected revenues. For example, if a $0.8 spread is set (i.e., placing orders $0.4 on either side of the mean), the price must fluctuate at least one standard deviation to hit the orders, with a probability of about 32%. Assuming each execution captures half a spread ($0.4), the expected revenue per time period is approximately $0.128 (32% × $0.4).

Source: Zhihu

• C. Finding the Optimal Solution: Through iterative calculations of different spreads, we find that: a spread of $0.2 has an expected revenue of about $0.08; a spread of $0.4 has an expected revenue of about $0.122; a spread of $0.6 has an expected revenue of about $0.135; and a spread of $0.8 has an expected revenue of about $0.128. The conclusion is that under this model, the optimal spread is $0.6, meaning placing orders at a distance of $0.3 (approximately 0.75σ) from the average price maximizes expected revenue.

Source: Movemaker

3.5 From Static Models to Dynamic Realities: Multi-Timeframe Risk Management

The fatal flaw of the above model is that it assumes a constant mean. In real markets, price means drift over time. Therefore, professional market makers must adopt a multi-timeframe layered strategy to manage risks.

The core of the strategy lies in using the micro level (second-level) to apply quantitative models to set optimal spreads while monitoring the drift of price means and changes in volatility structure at the meso level (minute-level) and macro level (hourly/daily level). When the mean shifts, the system will dynamically recalibrate the central axis of the entire quoting range** and adjust inventory positions accordingly.

This layered model ultimately leads to a set of dynamic risk control rules:

1. When second-level volatility increases, automatically widen the spread.
2. When mid-term volatility increases, reduce the size of individual orders but increase the number of order levels, dispersing inventory across a wider price range.
3. When long-term trends oppose the direction of inventory positions, actively intervene, such as further reducing order sizes or even pausing strategies to prevent systemic risks.

Risk Response Mechanisms and Advanced Strategies

4.1 Inventory Risk Management in High-Frequency Market Making

The dynamic strategy model discussed earlier falls under the category of high-frequency market making. The core objective of such strategies is to maximize expected profits by precisely managing inventory risk through algorithmically setting optimal buy and sell quotes.

Inventory risk is defined as the risk that market makers are exposed to adverse price fluctuations due to holding net long or net short positions. When market makers hold long inventory, they face the risk of losses from price declines; conversely, when holding short inventory, they face the risk of losses from price increases. Effectively managing this risk is key to the long-term survival of market makers.

Professional quantitative models, such as the classic Stoikov Model, provide a mathematical framework for understanding their risk management logic. This model aims to actively manage inventory risk by calculating a dynamically adjusted "reference price." The market maker's bilateral quotes will revolve around this new reference price rather than a static market mid-price. Its core formula is as follows:

Where the meanings of the various parameters are as follows:

• r(s,q,t): the dynamically adjusted reference price, which serves as the baseline axis for the market maker's quotes.
• s: the current market mid-price.
• q: the current inventory level. Positive if long, negative if short.
• γ: risk aversion parameter. This is a key variable set by the market maker that reflects their current risk preference.
• σ: asset volatility.
• (T−t): the remaining time until the trading period ends.

The core idea of this model is that when the market maker's inventory (q) deviates from its target (usually zero), the model systematically adjusts the quoting axis to incentivize market transactions that can bring their inventory back to equilibrium. For example, when holding long inventory (q > 0), the calculated r(s,q,t) will be below the market mid-price s, meaning the market maker will lower their buy and sell quotes overall, making sell orders more attractive and buy orders less attractive, thereby increasing the probability of liquidating long inventory.

4.2 Risk Aversion Parameter (γ) and Final Strategy Selection

The risk aversion parameter γ is the "regulator" of the entire risk management system. Market makers dynamically adjust the value of γ based on a comprehensive assessment of market conditions (such as volatility expectations, macro events, etc.). In stable markets, γ may be low, and strategies lean towards aggressively capturing spreads; when market risks escalate, γ is increased, making strategies extremely conservative, with quotes significantly deviating from the mid-price to quickly reduce risk exposure.

In extreme cases, when the market signals the highest level of risk (e.g., liquidity exhaustion, severe price dislocation), the value of γ becomes extremely large. At this point, the optimal strategy calculated by the model may generate quotes that are extremely far from the market, almost impossible to execute. In practice, this equates to a rational decision—temporarily and completely withdrawing liquidity to avoid catastrophic losses due to uncontrollable inventory risks.

4.3 Complex Strategies in Reality

Finally, it must be emphasized that the models discussed in this article are merely a simplified explanation of the core logic of market makers under idealized assumptions. In real, highly competitive market environments, top market makers employ far more complex, multi-layered strategy combinations to maximize profits and manage risks.

These advanced strategies include but are not limited to:

• Hedging Strategies: Market makers typically do not leave their spot inventory exposed to risks but instead establish opposite positions in derivative markets such as perpetual contracts, futures, or options to achieve Delta neutrality or more complex risk exposure management, transforming their risks from price directional risks to other controllable risk factors.

• Special Execution: In certain specific scenarios, the role of market makers transcends passive liquidity provision. For example, after a project's TGE, they may use strategies such as TWAP (Time-Weighted Average Price) or VWAP (Volume-Weighted Average Price) to sell large amounts of tokens over a certain period, constituting an important source of profit.

Review of 1011: Risk Triggers and Market Makers' Inevitable Choices

Based on the analytical framework established earlier, we can now review the market upheaval of 1011. When prices exhibit drastic one-sided movements, the internal risk management systems of market makers are inevitably triggered. The triggering of this system may be due to multiple factors: the average loss over a certain timeframe exceeds a preset threshold; net inventory positions are "filled" by counterparties in the market within a very short time; or after reaching maximum inventory limits, they cannot effectively liquidate positions, leading the system to automatically execute position contraction procedures.

5.1 Data Analysis: Structural Collapse of the Order Book

To understand the true state of the market at that time, we must analyze the microstructure of the order book in depth. The following chart, sourced from an order book visualization tool, provides us with evidence:

Source: @LisaLewis469193

(Note: For the sake of analytical rigor, please consider this chart as a typical representation of the market conditions at that time)

This chart visually demonstrates the changes in order book depth over time:

• Gray Area: Represents sell-side liquidity, i.e., the total of limit orders waiting to sell above the current price.
• Blue/Black Area: Represents buy-side liquidity, i.e., the total of limit orders waiting to buy below the current price.

At the precise moment marked by the red vertical line at 5:13 AM, we can observe two unusual phenomena occurring simultaneously:

1. Instantaneous Evaporation of Buy Support: The blue area at the bottom of the chart shows a massive, nearly vertical "cliff." This shape is entirely different from the situation where buy orders are consumed through large transactions— the latter should present a gradual erosion of liquidity. The only reasonable explanation for this uniform vertical disappearance is that a large number of limit buy orders were actively, simultaneously, and en masse canceled.
2. Simultaneous Disappearance of Sell Resistance: The gray area at the top of the chart also shows an almost identical "cliff." A large number of limit sell orders were actively withdrawn at the same moment.

This series of actions is referred to in trading terms as "liquidity withdrawal." It signifies that the primary liquidity providers in the market (mainly market makers) have almost simultaneously withdrawn their bilateral quotes within a very short time, instantly transforming a seemingly liquid market into an extremely fragile "liquidity vacuum."

5.2 Two Phases of the Event: From Active Withdrawal to Vacuum Formation

Thus, the process of the crash on 1011 can be clearly divided into two logically progressive phases:

Phase One: Active, Systematic Risk Aversion Execution

Before 5:13 AM, the market may still have appeared to be in a stable state. However, at that moment, a key risk signal was triggered—this could have been a sudden macro news event or an alert from a core protocol's (e.g., USDe/LSTs) on-chain risk model.

Upon receiving the signal, the algorithmic trading systems of top market makers immediately executed the pre-set "emergency risk aversion protocol." The sole goal of this protocol was to minimize their market risk exposure in the shortest time possible, prioritizing this over any profit objectives.

• Why cancel buy orders? This is the most critical defensive operation. The market maker's system anticipates that an unprecedented wave of selling pressure is about to arrive. If they do not immediately withdraw their buy orders, these orders will become the "first line of defense" in the market, forced to absorb a large amount of assets that are about to plummet, resulting in catastrophic inventory losses.

• Why simultaneously cancel sell orders? This is also based on strict risk control principles. In an environment where volatility is about to increase sharply, retaining sell orders also carries risks (for example, prices may experience a brief upward "false breakout" before plummeting, causing sell orders to be executed at unfavorable prices). Within an institutional-level risk management framework, the safest and most rational choice is to "clear all quotes and enter observation mode" until the market re-establishes predictability, then redeploy strategies based on the new market conditions.

Phase Two: Formation of Liquidity Vacuum and Free Fall of Prices

After 5:13 AM, with the formation of the "cliff" in the order book, the market structure underwent a fundamental qualitative change, entering the "liquidity vacuum" state we described.

Before the active withdrawal, to cause a 1% drop in market prices, a large number of sell orders would have been needed to consume the stacked buy orders. However, after the withdrawal, with the support structure below no longer in place, it may only take a few sell orders to cause an equal or even more severe price impact.

Conclusion

The epic market crash of 1011 was catalyzed and amplified directly by a large-scale, synchronized active liquidity withdrawal executed by top market makers, as revealed by the charts. They were neither the "culprits" nor the initiators of the crash, but they were the most efficient "executors" and "amplifiers" of the collapse. Through rational, self-preserving collective actions, they created an extremely fragile "liquidity vacuum," providing perfect conditions for subsequent panic selling, protocol decoupling pressures, and ultimately a chain liquidation of centralized exchanges.

About Movemaker

Movemaker is the first official community organization authorized by the Aptos Foundation, jointly initiated by Ankaa and BlockBooster, focusing on promoting the construction and development of the Aptos ecosystem in the Chinese-speaking region. As the official representative of Aptos in the Chinese-speaking area, Movemaker is committed to building a diverse, open, and prosperous Aptos ecosystem by connecting developers, users, capital, and numerous ecological partners.

Disclaimer:

This article/blog is for reference only and represents the author's personal views, not the position of Movemaker. This article does not intend to provide: (i) investment advice or recommendations; (ii) offers or solicitations to buy, sell, or hold digital assets; or (iii) financial, accounting, legal, or tax advice. Holding digital assets, including stablecoins and NFTs, carries high risks, significant price volatility, and may even become worthless. You should carefully consider whether trading or holding digital assets is suitable for you based on your financial situation. For specific issues, please consult your legal, tax, or investment advisor. The information provided in this article (including market data and statistics, if any) is for general reference only. Reasonable care has been taken in compiling this data and charts, but no responsibility is accepted for any factual errors or omissions expressed therein.