Original Author: A1 Research

Original Translation: AididiaoJP, Foreight News

From AMM to CLOB: Bringing Nasdaq into the $70 Trillion On-Chain Race According to a report by the Bank for International Settlements in 2022, global exchange markets process over $7.5 trillion in trading volume daily. Cryptocurrencies account for less than 2% of this, with the average daily trading volume of cryptocurrencies in the first quarter of 2025 plummeting to $14.6 billion, and spot DEXs processing only about $5 billion daily, representing a negligible error in global finance.

If the financial world is destined to move on-chain, the primary question is not when, but whether the infrastructure is sufficiently developed. Consider the scale: the U.S. stock market trades about $300 billion daily, while the U.S. Treasury market sees nearly $900 billion in daily trading volume. For decades, professional traders, market makers, and institutions have built their algorithms, risk models, and entire operational stacks around a standard: the Central Limit Order Book (CLOB).

Now compare this with DeFi. DeFi requires them to abandon this framework and embrace Automated Market Makers (AMM): trading with mathematical curves instead of order books. From the perspective of institutions managing billions of dollars, this is inefficient and unfamiliar.

So what is the result? Most of the capital entering the crypto space remains on centralized exchanges like Binance and Coinbase, which have the infrastructure familiar to traditional financial institutions. The transparency and self-custody markets promised by DeFi are appealing, but their current operation is completely at odds with Wall Street.

The good news is that this situation is changing. The emergence of on-chain CLOBs represents a moment of maturity for DeFi, where blockchain infrastructure finally matches the complexity of traditional markets.

When Citadel Securities processes about 35% of U.S. listed retail trading volume through its platform, and when Jane Street generates $20.5 billion in net trading revenue in 2024, they are not using AMM but CLOB. And now, with platforms like @HyperliquidX processing 200,000 orders per second, and with Ethereum L2 and Solana networks achieving millisecond latency, the infrastructure is becoming sufficiently developed to support $7.5 trillion in daily trading volume.

This is not about replacing AMM; AMM will continue to play a crucial role in on-chain price discovery, especially for long-tail digital assets. It is about building a bridge that brings Wall Street on-chain, enabling BlackRock to trade mainstream stocks and bonds directly on DeFi rails, making "decentralized finance" not just a vision for retail traders, and ultimately unleashing the full potential of programmable, composable on-chain DeFi primitives.

The path from AMM to on-chain CLOB is not just a technological evolution but a story of DeFi's growth. When blockchain first enabled trading, limited block space and slow transactions made traditional order books unfeasible. AMM provided an elegant solution: trading with mathematical curves instead of waiting for counterparties. They made DeFi possible. But now, as infrastructure matures and institutional demand becomes critical, the market is returning to the way it has always operated at scale: order books.

This article explores the technical mechanisms of the two systems, their fundamental trade-offs, and why the most complex trading applications are leading the trend back to CLOB, which is not a denial of DeFi innovation but a natural evolution towards institutional readiness.

Automated Market Makers: DeFi's Innovation from Zero to One

Ethereum has significant limitations: low throughput (about 15 transactions per second) and high, volatile gas fees. Replicating a traditional high-frequency CLOB requires constant order submissions, cancellations, and modifications, which is computationally and economically unfeasible. AMM is a genius solution to this problem.

AMM does not match individual buyers and sellers but allows users to trade with a reserve of asset pools, known as liquidity pools. Prices are not determined by an order book but by a deterministic algorithm.

Constant Product Market Maker (CPMM)

The most basic and popular type of AMM is the Constant Product Market Maker, implemented by Uniswap. Its mechanism is controlled by a simple yet powerful formula:

x × y = k

Where:

• x is the amount of asset A in the liquidity pool.
• y is the amount of asset B in the liquidity pool.
• k is the constant product. This value must remain unchanged during a trade (ignoring fees).

In CPMM, the price of an asset is simply the ratio of reserves, with the price of asset A = y / x.

Trade Example

Let’s understand how trades are executed in CPMM through a specific example and why slippage becomes a key consideration.

Initial Liquidity Pool State

Consider an ETH/USDC pool with the following reserves:

• x = 1,000 ETH
• y = 4,500,000 USDC
• k = 1,000 × 4,500,000 = 4,500,000,000

The spot price before the trade is the ratio of reserves:

Price = y / x = 4,500,000 / 1,000 = 4,500 USDC per ETH

Executing the Trade

Scenario: A trader wants to exchange 10 ETH for USDC.

When the trader withdraws 10 ETH from the pool, the constant product formula dictates:

New ETH balance: x' = 1,000 - 10 = 990 ETH

The USDC balance must adjust so that: 990 × y' = 4,500,000,000

Thus: y' = 4,500,000,000 ÷ 990 = 4,545,454.55 USDC

The trader must deposit:

Required USDC = 4,545,454.55 - 4,500,000 = 45,454.55 USDC

Effective payment price = 45,454.55 ÷ 10 = 4,545.45 USDC per ETH

Note that the trader paid 4,545.45 USDC for each ETH instead of the initial spot price of 4,500. This difference is known as price impact.

Understanding Slippage

Slippage represents the percentage difference between the expected price (spot price) and the actual execution price. In our example:

Slippage = (4,545.45 - 4,500) ÷ 4,500 × 100% = 1.01%

This 1.01% slippage may seem acceptable, but as the trade size increases, the constant product formula leads to exponentially worse prices:

For a 50 ETH trade:

• New ETH balance: 950
• New USDC balance: 4,500,000,000 ÷ 950 = 4,736,842.11
• Required USDC: 236,842.11
• Price per ETH: 4,736.84
• Slippage: 5.26%

For a 100 ETH trade:

• New ETH balance: 900
• New USDC balance: 4,500,000,000 ÷ 900 = 5,000,000
• Required USDC: 500,000
• Price per ETH: 5,000
• Slippage: 11.11%

Price Impact Curve

The relationship between trade size and price impact follows a hyperbolic curve. When you trade close to a larger percentage of the pool's liquidity:

• 1% of pool liquidity → approximately 1% slippage
• 5% of pool liquidity → approximately 5.3% slippage
• 10% of pool liquidity → approximately 11.1% slippage

Key AMM Concepts and Challenges

Liquidity Providers: Anyone can provide assets to the pool (e.g., depositing 1 ETH and 2,000 USDC) to become a liquidity provider. In return, they earn a portion of the trading fees generated by that pool.

Impermanent Loss: The most misunderstood risk for LPs is that AMM pools are isolated markets. Prices within them are not set by external information sources but determined by the constant product formula. Whenever the market price of an asset changes—like ETH doubling on Coinbase—arbitrageurs will intervene and trade with the pool until its price matches the global market. This rebalancing process extracts value from LPs: they end up holding more depreciated assets and fewer appreciated ones. This loss is termed "impermanent" because it can disappear if prices return to their original ratios, but in a volatile market, it often materializes as a very real opportunity cost compared to simply holding.

Capital Inefficiency: In a standard CPMM model, liquidity is distributed across the entire price curve from zero to infinity. This means that at any given time, the vast majority of capital in the pool is idle, as trades only occur at the current market price. For stablecoin pairs like USDC/DAI that fluctuate narrowly around $1.00, providing liquidity at $0.10 or $10.00 is extremely inefficient.

Evolution: Concentrated Liquidity (Uniswap v3)

To address capital inefficiency, Uniswap v3 introduced concentrated liquidity. LPs no longer need to provide liquidity across the entire price range but can choose to provide liquidity within specific price ranges.

For example, an LP can provide liquidity only for the ETH/USDC pair in the range of $4,400 - $4,800. This concentrates their capital where most trades actually occur, allowing them to earn significantly more fees with the same amount of capital. Functionally, this creates a series of deeper liquidity positions, beginning to resemble "limit orders" similar to an order book, marking the first major conceptual bridge between AMM and CLOB.

Remaining challenges faced by concentrated liquidity:

Amplified Impermanent Loss

When prices move outside their range, concentrated positions experience amplified impermanent loss. LPs face a cruel dilemma: narrower ranges earn more fees but suffer greater losses when prices drift. A position concentrated within a 1% range could lose 100% of one asset if the price moves just 1% in either direction.

Active Management Burden

Unlike the "set and forget" approach of V2, V3 requires continuous monitoring and rebalancing. When ETH moves from $4,500 to $4,600, a position centered at $4,500 becomes inactive, earning zero fees until manually adjusted. This creates operational overhead comparable to traditional market making.

Gas Cost Complexity

Managing concentrated positions requires frequent trading for rebalancing, position adjustments, and fee collection. During periods of high volatility, gas costs can exceed fee income, especially for smaller positions. This poses a barrier to entry for retail LPs.

MEV Vulnerability Persists

Instant liquidity attacks become more complex. MEV bots can surgically target concentrated positions, extracting value just before large trades and immediately removing liquidity afterward, leaving losses for regular LPs.

Price Discovery Still Fails

The x*y=k formula, even in a concentrated context, does not reflect true market dynamics. There is no market sentiment, order flow, or price-time priority concept. Every trade moves the price, regardless of size or intent, creating artificial volatility.

Liquidity Fragmentation

LPs choosing different ranges create a fragmented liquidity landscape. Traders may face good liquidity at $4,500 but terrible slippage at $4,550, leading to unpredictable execution quality across price levels.

No Native Limit Orders

While concentrated positions resemble limit orders, they are not true limit orders. They continue to provide liquidity in both directions and may be partially filled multiple times, without guaranteeing execution at a specific price.

Success of Spot and Issues with Perpetual Contracts

Thus, while AMMs have revolutionized spot trading (with Uniswap alone facilitating over $2 trillion in cumulative trading volume), their success does not translate to the perpetual futures market. This divergence reveals a fundamental truth about market structure: different tools require different infrastructures.

What does this mean? The spot market is forgiving. Traders accept slippage as the cost of immediate execution when exchanging ETH for USDC. Trades settle instantly, with no ongoing obligations. AMMs excel here because their simplicity aligns with the direct nature of spot trading.

On the other hand, perpetual futures require precise entry and exit prices, ongoing funding rate calculations, real-time clearing engines, and leverage management. Platforms like @GMX_IO and other AMM-based perpetuals struggle with these demands. Their reliance on oracle pricing creates toxic order flow opportunities, allowing traders to exploit price discrepancies between oracle feeds and actual market conditions. The lack of true price discovery means positions are often mispriced, exposing liquidity providers to asymmetric risks. AMM-based perpetual platforms have implemented stop-loss and limit order systems, but these lack the finesse, reliability, and price discovery advantages of a true order book market.

The result is predictable: professional traders remain on centralized exchanges. While Uniswap has captured significant spot market share from Coinbase, GMX and its peers have barely made a dent in Binance's dominance in the perpetual contract space. Perpetual futures trading volume is 3-5 times that of the spot market and remains firmly in the hands of CeFi.

This is not a failure of execution but a mismatch of architecture. Perpetual futures evolved from traditional futures markets, which have always relied on order books for price discovery and risk management. Trying to force them into an AMM model is like asking Formula 1 cars to run on square wheels—technically possible, but fundamentally inefficient.

The market is ready for solutions, and Hyperliquid and the new generation of on-chain CLOBs are now providing these solutions, recognizing a simple truth: to capture institutional perpetual contract flow, you need institutional-grade infrastructure. Not approximations, not workarounds, but the real thing: an on-chain order book with performance that rivals centralized venues.

Central Limit Order Book (CLOB): Precision and Efficiency

CLOB is the cornerstone of traditional finance, powering everything from the New York Stock Exchange to Coinbase. It is a transparent and efficient system for matching buyers and sellers.

Core Mechanism

CLOB essentially consists of two order lists for specific asset pairs:

• Buyers: a list of buy orders sorted by price from high to low.
• Sellers: a list of sell orders sorted by price from low to high.

The difference between the highest bid (the highest price someone is willing to pay) and the lowest ask (the lowest price someone is willing to accept) is called the bid-ask spread. In addition to the spread, the depth at each price level also affects execution quality. For example, a CLOB with 100 ETH available at $4,500 provides better execution for large trades than one with only 10 ETH at that level, as deeper liquidity reduces slippage.

Order Types and Matching Engine

Users can interact with the CLOB using various order types:

Limit Order: An order to buy or sell at a specific price or better. A limit order to buy ETH at $4,495 will only execute if the sell price reaches $4,495 or lower. If it cannot be filled immediately, it remains on the order book, adding depth to the market. This is how market makers provide liquidity.

Market Order: An order to buy or sell immediately at the best available market price. A market buy order will "sweep the order book," consuming the lowest sell orders until the entire order is filled. This provides certainty of execution but does not guarantee price.

Stop-Loss Order: Activated only when a specified trigger price is reached. For example, a stop-loss sell order at $4,400 will execute once ETH drops to that level, helping traders manage downside risk.

The matching engine is the core algorithm that executes these rules, typically following a price-time priority principle. Orders at better price levels are matched first. If multiple orders exist at the same price, the first one placed is matched first. This FIFO method at each price level ensures fairness and prevents front-running, unlike AMMs where larger trades can extract more value.

The Order Book Engine: Professional Market Makers

An order book is just a list of intentions until liquidity is present. Unlike AMMs, where liquidity is passively provided by a diverse pool of LPs, CLOB relies on a class of specialized participants to operate effectively: market makers. These are complex entities, often professional trading firms or specialized liquidity funds, whose primary business is to provide liquidity.

What Do Market Makers Actually Do?

The core function of market makers is to be ready to buy and sell an asset at any given time. They achieve this by placing both a buy order and a sell order on the order book simultaneously. This action serves two key objectives:

• Ensuring Liquidity: Market makers ensure that there are always orders available for retail traders to trade against. Traders wanting to sell an asset can immediately hit the market maker's buy price, while those wanting to buy can immediately get their sell price.
• Narrowing the Spread: Competition among multiple market makers forces the difference between the highest bid and the lowest ask to be as small as possible. A narrow spread is a hallmark of a healthy, liquid market and provides traders with better prices.

Market makers' primary profit comes from capturing the bid-ask spread. For example, if they have a buy order for ETH at $1,999.50 and a sell order at $2,000.00, their goal is to buy from sellers at the lower price and sell to buyers at the higher price, earning a $0.50 spread on each round trip. Their total profit is essentially (spread) x (volume).

This is not a risk-free activity. Market makers face significant inventory risk.

If the overall market price of ETH suddenly drops, the market maker's buy orders will be filled, and the accumulated ETH inventory will now be worth less than what they paid. If a market maker accumulates 100 ETH at $4,500 and the price drops to $4,400, they face an unrealized inventory loss of $10,000.

Conversely, if the price of ETH rises sharply, their sell orders will be filled, selling their inventory at a price lower than the new higher market price.

To manage this, market makers use complex algorithms to continuously adjust their quotes based on market volatility, trading volume, and their current inventory levels. Professional market makers often use perpetual futures or options on centralized exchanges to hedge their inventory risk, maintaining delta-neutral positions. This is a highly active and data-driven process, contrasting sharply with the "deposit and forget" nature of passive LPs in standard AMMs.

On-Chain Liquidity Fund Landscape

The transition to on-chain CLOBs has attracted professional liquidity funds and trading firms that have honed their skills in traditional finance and centralized crypto markets. Companies like @wintermutet, @jump, and @GSR_io are now major players in DeFi, providing deep liquidity for on-chain order books.

These companies do not trade manually. They connect to DEX protocols via APIs and run high-frequency, automated strategies. To attract these key participants, on-chain CLOBs have developed robust incentive structures:

Maker Rebates: Many order books offer a "maker-taker" fee model. The "taker" pays a fee, while the "maker" receives a small rebate. For high-volume market makers, these rebates can become a significant source of income.

Liquidity Mining Programs: Protocols often reward market makers directly with their native governance tokens. These programs typically require market makers to meet specific key performance indicators, such as maintaining a certain order depth on designated trading pairs, maximum spreads, and over 90% uptime. This is an efficient strategy for protocols to guide liquidity into new markets.

Operating on the blockchain presents unique challenges not found in traditional finance:

• Gas Costs: Every order placement, cancellation, and update is an on-chain transaction that consumes gas. This creates ongoing operational costs that market makers must factor into their profit models. Low fees on L2 and high throughput on L1 are crucial for achieving this.
• Delays and MEV: The block time of the blockchain introduces delays. For example, with Ethereum's approximately 12-second block time, a market maker's orders may be "in transit" and unmodifiable for up to 12 seconds, while traditional finance can achieve microsecond updates.

During this time, the market may turn against them. Complicating matters, orders are visible in the public memory pool before confirmation, exposing them to maximum extractable value strategies, such as front-running. To mitigate this, market makers employ techniques such as order splitting, routing through private memory pools, or utilizing off-chain execution.

Why CLOB is Making a Comeback: Technological Drivers

The initial barriers faced by on-chain CLOBs were computational

High Throughput L1: Chains like @solana, @SeiNetwork, @monad, @Aptos, and @SuiNetwork are built for high throughput and low latency, making on-chain order books feasible. These are general-purpose L1s designed to host many applications. In contrast, dedicated L1s like Hyperliquid's HyperCore are tailored for trading, with matching engines optimized for speed and performance.

Rollups: High-performance rollups like @megaethlabs, @fuelnetwork, and @risechain aim to achieve real-time, low-latency trading on Ethereum by leveraging parallel transaction processing. In addition to general rollups, we are also seeing specialized L2 application chains. For example, @hibachixyz based on @celestia or Solana network expansions like @bulletxyz_ are specifically built to host on-chain matching engines.

Crucially, these designs rely on scalable data availability layers, such as @eigen_da and Celestia, which achieve the throughput required for order book-style exchanges. Meanwhile, advancements in ZK infrastructure have made it possible to run verifiable off-chain CLOBs, combining performance with Ethereum-level security.

Projects like Hyperliquid, Bullet, and @dYdX are excellent examples of DEXs based on the CLOB model.

Part Three: Direct Comparison: AMM vs. CLOB

What This Means for Users

The transition from AMM to CLOB is not just a technological upgrade; it directly reshapes the user experience:

Retail Traders: Gain better prices and lower slippage, with a trading interface familiar to anyone used to centralized exchanges.

Institutions: Now have access to professional-grade tools, advanced order types, risk management, and deep liquidity on a transparent, decentralized platform.

DeFi Protocols: Unlock more composable liquidity, with capital effectively allocated and seamlessly integrated across the ecosystem.

As blockchain approaches traditional financial levels of performance, the gap between centralized and decentralized trading experiences will narrow, making on-chain markets not just an alternative but a competitive arena in global finance.

Conclusion: The Maturation of DeFi Trading

AMMs were an innovation that took DeFi from zero to one, solving the cold start problem that made on-chain trading possible when blockchains were slow and expensive. They democratized market making and provided a simple, unstoppable way to trade on-chain.

However, as DeFi matures from a niche market of early adopters into a parallel financial system seeking to attract institutional capital and professional traders, its infrastructure must also mature. Central limit order books provide market makers with unparalleled capital efficiency, precise price control, and the fine-tuned control required for complex trading strategies.

While AMMs will always have a place in long-tail assets and simple exchanges, the future of high-volume, professional-grade decentralized trading belongs to CLOBs. The movement of CLOBs is not to replace AMMs but to build the next layer of complex financial infrastructure on-chain.

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。