Evermore is a new Layer 1 with an enshrined order book and a novel market microstructure, based on frequent batch auctions, at its core. The flagship native application of Evermore is a decentralized perpetual futures exchange with unrivaled fairness and market efficiency.

Evermore processes 100,000+ onchain orders per second and is further capable of supporting a vast ecosystem of DeFi applications including spot exchange, lending, money markets, margin trading, perpetual futures, options, structured products and more.

Evermore aligns with the principles of decentralization by enshrining the order book to be fully onchain, with self-custody and permissionless trading capabilities. The core mechanisms of Evermore are specifically designed to establish a higher standard of transparency and fairness for complex DeFi infrastructure.

To understand Evermore’s unique approach, we first need to understand the continuous-time market design underpinning nearly all of finance.

Traditional markets and nearly all centralized exchanges utilize a market design called the continuous limit order book (CLOB). CLOBs are a specific type of limit order book where orders are submitted and executed in real-time. All orders are processed serially, and matching engines process each order strictly according to price-time priority. Under this market design, it is possible to buy or sell at any instant, reflecting the real-time dynamics of demand and supply.

The basic building block is the limit order. As a refresher, a limit order specifies a price, a quantity, and a direction to buy or sell. Traders can submit these orders to the exchange, and can also submit messages to cancel or modify their existing orders. The set of all outstanding limit orders at any given time is known as the limit order book.

High-frequency traders and market makers prefer to use limit orders due to their ability to directly express trading intent at precise price levels and quantities. Limit order books offer fast and accurate price discovery and granular control over trade execution.

However, despite the advantages of limit order books, there are significant challenges to building them in a decentralized environment. These challenges and constraints have motivated other notable exchange designs.

There are several alternative models for decentralized exchanges. Given the availability of other market designs, why focus on building an order book?

While automated market makers (AMMs), concentrated liquidity pools, oracle-based liquidity pools, intents, and RFQs can offer certain benefits, each faces significant limitations:

• Automated market makers (AMMs) offer simplicity and passive liquidity but introduce impermanent loss and inefficient pricing formulas.

• Concentrated liquidity pools partially address efficiency concerns but still suffer from impermanent loss and scalability issues in volatile markets.

• Oracle-based liquidity pools heavily depend on external data providers, creating centralization risks, potential for oracle manipulation, and price latency.

• Intent-based systems and RFQs offer flexibility but encounter key structural issues, reduce transparency, and rely on complex intermediaries, compromising decentralization.

Order books, when properly implemented, offer precise and transparent price discovery, deeper liquidity, and fairer market conditions for all participants. This makes them a strong choice compared to alternative models. Unfortunately, as we’ll see, the continuous-time assumption underlying the CLOB market design itself has significant issues.

In continuous-time markets, there will always be wasteful and unnecessary competition.

Every time asset prices move, there are high-frequency traders looking to “snipe” stale quotes before they can be updated. As a result, liquidity providers and market makers are always trying to update their outstanding quotes before they get sniped. Since CLOBs process messages serially, in order of arrival, liquidity providers usually lose this race: any request to update a stale quote has to be the first to reach the exchange, ahead of every single attempt to pick off the order.

Even the smallest difference in timestamps decides whether an order is filled at a favorable price or not, because the first order to hit the book always gets execution priority. This dynamic creates a huge premium on speed: high-frequency traders invest heavily in low-latency infrastructure and faster trading systems, all in an effort to beat other traders sniping liquidity providers’ quotes.

Building and maintaining all of this low-latency trading infrastructure is expensive, so these costs are ultimately passed along to all other market participants, in the form of thinner liquidity and wider spreads.

Even worse, all this expenditure on tiny speed gains doesn't yield any fundamental improvements. Essentially, the CLOB is a zero-sum game to be first to seize an opportunity that shouldn't exist in a well-designed market.

Continuous-time markets also assume that infrastructure can process transactions instantly and serially, an assumption that is entirely incompatible with blockchain environments that operate with significant constraints on computing and discrete intervals dictated by block times.

While some decentralized exchanges attempt to these problems of adverse selection and “toxic flow” via transaction reordering (e.g., prioritizing new cancellations above order placements), this is a stop-gap solution for an issue that requires fundamental reworking.

Evermore prioritizes fairness in its market design.

To accomplish this, Evermore’s enshrined order book implements a novel market microstructure mechanism known as frequent batch auctions (FBAs). In Evermore, markets are structured as uniform-price double auctions conducted in short, discrete-time intervals.

FBAs are actually similar to CLOBs, except for two key differences:

• Time is treated as discrete, not continuous. This approach aligns with blockchain constraints, eliminating unrealistic assumptions about instantaneous transaction processing.

• Orders are processed in batch, not serially. Orders are processed simultaneously within each batch interval, rather than continuously as they arrive.

Traders submit regular limit orders, just like in a CLOB. However, any orders arriving within the same batch interval are treated as having arrived at exactly the same time. The exchange periodically batches all outstanding orders together and determines a single, reference clearing price. Since all trades in a batch share the same price, no one receives a worse price due to being a few milliseconds slower. In fact, liquidity providers frequently see price improvement.

Evermore transforms the nature of competition itself. Traders no longer gain an advantage by milliseconds. Instead, they must compete for fills by providing better prices. This approach shifts market dynamics from speed-based strategies to price-based competition, thereby promoting fairer and more effective trading.

By enshrining the order book at the protocol level, Evermore ensures consistent, reliable, and unbiased order execution for all traders, enhances liquidity and encourages market stability, and facilitates more accurate price discovery while eliminating front-running and stale quote sniping.

The innovations introduced by Evermore, like highly-parallelized batch processing and an enshrined order book architecture, require deep integration at the blockchain level. Implementing these features on existing high-performance Layer 1s would result in inefficiencies, higher operational costs, and other complexities, such as cranking and additional layers of economic abstraction, that would compromise scalability and performance.

A dedicated Layer 1 allows Evermore to optimize components for maximum efficiency. This approach ensures parallelization of batch order processing and exceptional throughput with minimal latency, which are critical for achieving global scale. Moreover, Evermore achieves true composability and interoperability and prioritizes the specific needs of complex DeFi applications.

By adapting the latest research in market design into the lowest levels of a new Layer 1, Evermore enhances market quality and sets a new standard for decentralized trading and exchanges.

In subsequent posts, we will delve deeper into Evermore’s mechanism design, architecture, and the technical foundations of a discrete-time approach to markets.

Evermore is a new Layer 1 with an enshrined order book and a novel market microstructure, based on frequent batch auctions, at its core. The flagship native application of Evermore is a decentralized perpetual futures exchange with unrivaled fairness and market efficiency.

Evermore processes 100,000+ onchain orders per second and is further capable of supporting a vast ecosystem of DeFi applications including spot exchange, lending, money markets, margin trading, perpetual futures, options, structured products and more.

Evermore aligns with the principles of decentralization by enshrining the order book to be fully onchain, with self-custody and permissionless trading capabilities. The core mechanisms of Evermore are specifically designed to establish a higher standard of transparency and fairness for complex DeFi infrastructure.

To understand Evermore’s unique approach, we first need to understand the continuous-time market design underpinning nearly all of finance.

Traditional markets and nearly all centralized exchanges utilize a market design called the continuous limit order book (CLOB). CLOBs are a specific type of limit order book where orders are submitted and executed in real-time. All orders are processed serially, and matching engines process each order strictly according to price-time priority. Under this market design, it is possible to buy or sell at any instant, reflecting the real-time dynamics of demand and supply.

The basic building block is the limit order. As a refresher, a limit order specifies a price, a quantity, and a direction to buy or sell. Traders can submit these orders to the exchange, and can also submit messages to cancel or modify their existing orders. The set of all outstanding limit orders at any given time is known as the limit order book.

High-frequency traders and market makers prefer to use limit orders due to their ability to directly express trading intent at precise price levels and quantities. Limit order books offer fast and accurate price discovery and granular control over trade execution.

However, despite the advantages of limit order books, there are significant challenges to building them in a decentralized environment. These challenges and constraints have motivated other notable exchange designs.

There are several alternative models for decentralized exchanges. Given the availability of other market designs, why focus on building an order book?

While automated market makers (AMMs), concentrated liquidity pools, oracle-based liquidity pools, intents, and RFQs can offer certain benefits, each faces significant limitations:

• Automated market makers (AMMs) offer simplicity and passive liquidity but introduce impermanent loss and inefficient pricing formulas.

• Concentrated liquidity pools partially address efficiency concerns but still suffer from impermanent loss and scalability issues in volatile markets.

• Oracle-based liquidity pools heavily depend on external data providers, creating centralization risks, potential for oracle manipulation, and price latency.

• Intent-based systems and RFQs offer flexibility but encounter key structural issues, reduce transparency, and rely on complex intermediaries, compromising decentralization.

Order books, when properly implemented, offer precise and transparent price discovery, deeper liquidity, and fairer market conditions for all participants. This makes them a strong choice compared to alternative models. Unfortunately, as we’ll see, the continuous-time assumption underlying the CLOB market design itself has significant issues.

In continuous-time markets, there will always be wasteful and unnecessary competition.

Every time asset prices move, there are high-frequency traders looking to “snipe” stale quotes before they can be updated. As a result, liquidity providers and market makers are always trying to update their outstanding quotes before they get sniped. Since CLOBs process messages serially, in order of arrival, liquidity providers usually lose this race: any request to update a stale quote has to be the first to reach the exchange, ahead of every single attempt to pick off the order.

Even the smallest difference in timestamps decides whether an order is filled at a favorable price or not, because the first order to hit the book always gets execution priority. This dynamic creates a huge premium on speed: high-frequency traders invest heavily in low-latency infrastructure and faster trading systems, all in an effort to beat other traders sniping liquidity providers’ quotes.

Building and maintaining all of this low-latency trading infrastructure is expensive, so these costs are ultimately passed along to all other market participants, in the form of thinner liquidity and wider spreads.

Even worse, all this expenditure on tiny speed gains doesn't yield any fundamental improvements. Essentially, the CLOB is a zero-sum game to be first to seize an opportunity that shouldn't exist in a well-designed market.

Continuous-time markets also assume that infrastructure can process transactions instantly and serially, an assumption that is entirely incompatible with blockchain environments that operate with significant constraints on computing and discrete intervals dictated by block times.

While some decentralized exchanges attempt to these problems of adverse selection and “toxic flow” via transaction reordering (e.g., prioritizing new cancellations above order placements), this is a stop-gap solution for an issue that requires fundamental reworking.

Evermore prioritizes fairness in its market design.

To accomplish this, Evermore’s enshrined order book implements a novel market microstructure mechanism known as frequent batch auctions (FBAs). In Evermore, markets are structured as uniform-price double auctions conducted in short, discrete-time intervals.

FBAs are actually similar to CLOBs, except for two key differences:

• Time is treated as discrete, not continuous. This approach aligns with blockchain constraints, eliminating unrealistic assumptions about instantaneous transaction processing.

• Orders are processed in batch, not serially. Orders are processed simultaneously within each batch interval, rather than continuously as they arrive.

Traders submit regular limit orders, just like in a CLOB. However, any orders arriving within the same batch interval are treated as having arrived at exactly the same time. The exchange periodically batches all outstanding orders together and determines a single, reference clearing price. Since all trades in a batch share the same price, no one receives a worse price due to being a few milliseconds slower. In fact, liquidity providers frequently see price improvement.

Evermore transforms the nature of competition itself. Traders no longer gain an advantage by milliseconds. Instead, they must compete for fills by providing better prices. This approach shifts market dynamics from speed-based strategies to price-based competition, thereby promoting fairer and more effective trading.

By enshrining the order book at the protocol level, Evermore ensures consistent, reliable, and unbiased order execution for all traders, enhances liquidity and encourages market stability, and facilitates more accurate price discovery while eliminating front-running and stale quote sniping.

The innovations introduced by Evermore, like highly-parallelized batch processing and an enshrined order book architecture, require deep integration at the blockchain level. Implementing these features on existing high-performance Layer 1s would result in inefficiencies, higher operational costs, and other complexities, such as cranking and additional layers of economic abstraction, that would compromise scalability and performance.

A dedicated Layer 1 allows Evermore to optimize components for maximum efficiency. This approach ensures parallelization of batch order processing and exceptional throughput with minimal latency, which are critical for achieving global scale. Moreover, Evermore achieves true composability and interoperability and prioritizes the specific needs of complex DeFi applications.

By adapting the latest research in market design into the lowest levels of a new Layer 1, Evermore enhances market quality and sets a new standard for decentralized trading and exchanges.

In subsequent posts, we will delve deeper into Evermore’s mechanism design, architecture, and the technical foundations of a discrete-time approach to markets.