Sei was founded by a team of experienced individuals in the finance and technology industries, including a former engineer at Robin Hood and a veteran of Goldman Sachs. The team's diverse background in both traditional finance and cutting-edge technology helped lay the foundation for the creation of Sei, which aims to build the best infrastructure for decentralized finance. With a deep understanding of the challenges and limitations of existing solutions, the team is dedicated to unlocking the full potential of decentralized exchanges and delivering the performance, scalability, and security that the industry demands.
Sei was founded by a former Robin Hood engineer and a person who worked at Goldman Sachs, and the team has now grown to 20 individuals, with half being engineers and the other half focused on building out the infrastructure of Sei and ensuring that there are successful projects built on top of it. Decentralization, scalability, and capital efficiency
Even more impressively, in the short amount of time since Sei’s inception, they have had over 100 teams from Solana, Polkadot, Terra, NEAR, and Ethereum to migrate to their protocol. On top of this, major market makers that are essential to order book liquidity have joined like Hudson River Trading and GSR.
The exchange trilemma in the world of crypto refers to the balance between scalability, capital efficiency, and decentralization. These three elements are often at odds with each other and it can be difficult for a cryptocurrency exchange to achieve all three simultaneously. Scalability refers to the ability of a system to handle increased volume of transactions or users without slowing down or becoming bogged down. Capital efficiency, on the other hand, refers to the amount of capital that needs to be locked up in order to provide liquidity to the exchange, with a more capital efficient exchange being one that requires less capital to provide the same level of liquidity. Decentralization refers to the degree of control that users have over the exchange and their assets, with a more decentralized exchange having more distributed control and less control in the hands of a few central parties. Balancing these three elements is crucial for the success of a cryptocurrency exchange and requires careful consideration of trade-offs and compromises.
Figure 1 -- exchange trillema -- AMM/CLOB/CEX
Order Books vs. Automated Market Makers
Automated Market Maker (AMM) Model is a type of exchange model used in decentralized finance (DeFi) that relies on mathematical algorithms to determine the price of assets, instead of an order book. The prices in AMM exchanges are calculated using a formula that incorporates various factors such as supply and demand, liquidity, and other parameters. These exchanges do not require users to place limit orders or to match with other orders. Instead, users can buy or sell an asset at the current price calculated by the AMM algorithm.
On the other hand, Order Book Model is a type of exchange model that relies on a ledger of bid and ask orders from multiple users. The order book is a record of all the limit orders that users have placed to buy or sell an asset. The exchange matches the orders on the order book and executes trades based on the highest bid and the lowest ask. Order book exchanges are typically more centralized and require users to place limit orders, which can result in longer wait times for orders to be executed.
Order books are beneficial when there is deeper liquidity as they allow for more precise price discovery and better execution for large orders. Order books also provide greater transparency and allow for a wider range of trading strategies.
SEI is built on Cosmos, using the Cosmos SDK and Tendermint Core as its foundation. With these tools, SEI has been able to add specialized functionality and optimizations to provide a high-performance exchange experience. The Cosmos SDK provides a modular platform that enables SEI to design its entire end-to-end blockchain flow, while Tendermint Core provides the consensus mechanism that ensures the security and reliability of the network. By leveraging these technologies, SEI has positioned itself as a leading player in the decentralized exchange space, offering a solution that balances scalability, capital efficiency, and decentralization.
Sei uses a unique technology called Twin-Turbo Consensus to enhance its performance. This technology is a combination of two components: intelligent block propagation and optimistic block processing.
Intelligent block propagation refers to the process of efficiently and effectively distributing block information across the network. This ensures that all nodes on the network have the same information, which is crucial for maintaining network consensus.
Optimistic block processing, on the other hand, allows the network to process blocks before they are fully confirmed, thereby increasing the speed and efficiency of the network. This results in a faster processing time and faster finality for transactions on the network.
Together, these two components make up Sei's Twin-Turbo Consensus, which has been designed to optimize the performance of the network, making it faster and more efficient for order matching and transaction processing.
- Transaction received by full node: A full node receives a transaction from a user and broadcasts it to other nodes in the network.
- Transaction dissemination: Full nodes will randomly gossip this transaction to other nodes in the network.
- Transaction verification: Once a transaction is received by a validator, it verifies the validity of the transaction.
- Transaction added to mempool: The validator adds the transaction to its local mempool.
- Block proposal: Block proposers will look at the current state of their mempool and propose a block to be committed.
- Disseminating block proposal: The proposer will first disseminate the proposal to other validators in the network.
- Disseminating full block: The proposer will then disseminate the entire block (with full contents of each transaction) which gets broken up in parts and gossiped to the network.
- Reconstructing the block: If a validator has all of the transactions from the proposal in its mempool, it will reconstruct the entire block from its mempool, if not, it will wait to receive all of the block parts from the network.
- Tendermint consensus: Validators will follow the Tendermint BFT consensus to agree on the transaction ordering.
- Prevote, precommit, and commit steps: This includes a prevote step, a precommit step, and a commit step before the block and the associated state changes have been committed to the blockchain.
This process of transaction receipt, verification, and consensus helps to significantly decrease the overall time a validator waits to receive a block.
Figure 2 - Block processing with example times - (a) Processing after precommit (b) Optimistic block processing
Optimistic Block Processing
Step 1: Block Proposal and Validation - As part of Tendermint consensus, validators receive a block proposal and verify its validity.
Step 2: Optimistic Block Processing - Validators start a process to optimistically process the first block proposal they receive for any height. This process writes the candidate state to a cache.
Step 3: Committing or Discarding Data - If the block is accepted by the network, the data from the cache is committed. If the network rejects the block, the data from the cache is discarded and future rounds for that height will not use optimistic block processing.
Step 4: Latency Improvement - The improvement in latency due to optimistic block processing is min(prevote latency + precommit latency, number of transactions * average latency of a single transaction).
min(Tprevote + Tprecommit, N * T)
Figure 3 - Theoretical Latency based on Optimistic block processing
The Cosmos SDK contains several logic stages in the process of updating the network state when a block is received. The three stages are BeginBlock, DeliverTx, and EndBlock. Sei has configured DeliverTx and EndBlock to be parallelizable so that transaction processing can be done faster.
When Sei processes transactions in a block, it first goes through the DeliverTx phase. During this phase, state changes are made for most transactions such as sending tokens, making governance proposals, or executing smart contracts. However, for transactions related to central limit order book (CLOB), only basic processing occurs in the DeliverTx phase, and most state changes occur during the EndBlock logic. This is done to support frequent batch auctions, where multiple orders are grouped and a single clearing price is calculated for executing these orders. To achieve optimal performance, Sei has added parallelization to both DeliverTx and EndBlock.
It get complicated and much more technically much more quickly.
Figure 4 -- Block Processing with and without Parallelization
Native Order Matching Engine
Sei is a blockchain platform that offers general blockchain functionality such as asset transfer and smart contract deployment. It also features a matching engine that facilitates order placement and matching for decentralized exchanges building on Sei.
The matching engine allows these decentralized exchanges to deploy their own central limit order books (CLOBs). The engine maintains the order books at the chain level and provides the necessary functions for creating markets and executing trades.
Creating a new order book (i.e. a new market) is a two-step process of deploying a smart contract onto Sei and submitting a transaction to add a new order book to the registered smart contract. The order book proposal must include information about the asset, the base denomination of pricing, and the minimum price interval. The matching engine supports various order types including limit orders, market orders, fill-or-kill orders, stop-loss orders, and cancel orders. The lifecycle of an order begins with its submission and is processed by the dex module's EndBlocker hook. The chain then calls the associated smart contract which implements its custom logic and calls the matching engine. The engine first processes all order cancellations, then adds limit orders to the order book store, processes market orders, and finally processes matching limit orders.
Figure 5 -- life cycle of a CLOB transaction
Native Price oracles
Sei has a price oracle system to provide accurate exchange rate pricing for assets. Validators are required to participate as oracles to ensure reliable pricing. Voting windows can be set to as short as 1 block to quickly update prices, and validators provide their exchange rate proposals for each window. At the end of the voting period, a weighted median of all the exchange rate votes is calculated to determine the true exchange rate. Non-participation and providing incorrect data are penalized with a miss count, and if a validator has a high miss count over a certain number of voting periods, they will be slashed as a penalty for misbehavior.
Predatory MEV Protection
By executing orders in a uniform clearing price, Sei discourages frontrunning of MEV. This is because market orders are aggregated and executed at the same price, reducing the incentive for validators to manipulate the order processing for personal gain. For instance, when incoming market orders are detected, validators cannot prioritize their own orders over others, as all market orders are cleared at the same price. As a result, existing limit orders are filled at their intended price, while incoming market orders receive a fairer price, which reduces the opportunities for frontrunning and maximizes the extractable value.
Figure 6 -- equation for frequent batch auctioning
The SEI ecosystem, which consists of over 100 teams from various blockchain platforms such as Solana, Ethereum, Near, Terra, and Polkadot, is poised to have a major impact once the mainnet launches. SEI is backed by a robust 50 million dollar ecosystem and liquidity fund that is dedicated to supporting these teams and bootstrapping liquidity. Large products are already beginning to follow the wake that sei has left including Sushiswap (01/25). THe most interesting product on sei right now is probably Nitro which is a SVM dapp on Solana.
Figure 7 -- sei ecosystem segmented (01/02)
This shared liquidity model is a game changer for developers as it offers a simple way for them to plug and play into the SEI order-matching engine and access liquidity from other applications. This eliminates the need for developers to manage their own order books or liquidity pools, freeing them up to focus on creating innovative financial products.
With shared liquidity, developers can build and launch products much faster, leveraging built-in price discovery mechanisms and improved execution speed. This will likely lead to a vibrant ecosystem of diverse financial instruments that can be accessed and used by the broader community. In turn, this will drive greater adoption of SEI and help to establish it as a major player in the DeFi space.
SEI has secured an additional $20 million in ecosystem funding to enter the Asian market. With a high adoption rate of crypto in Asia and a market that is ready for disruption, SEI has the potential to become the next big layer 1 platform in Asia, capturing a significant share of the market. The new funding will support SEI's expansion into Asia, providing resources for teams and bootstrapping liquidity in the region. The shared liquidity model, which allows for simple development of complex financial instruments and access to pooled liquidity from other apps, makes SEI an attractive option for the Asia market. They are continuing a world tour that has already met in Japan and is reaching into the Indian blockchain market.
SEI has a trading experience that sets it apart from other platforms. With its shared liquidity model and seamless integration with other apps, users can access a wide range of financial instruments with ease. The integration of the order-matching engine provides a streamlined and efficient way to trade, making it an attractive platform for both seasoned and new traders. Additionally, SEI’s liquidity fund and support for teams helps to provide stability to the platform, reducing volatility and improving the trading experience for users.
The superior trading experience on SEI can act as a catalyst for future growth. With its user-friendly interface and the ability to access a wide range of financial instruments, SEI is well-positioned to attract new traders and retain existing ones. Furthermore, SEI’s support for teams and commitment to stability also makes it an attractive platform for developers looking to build new financial products. The combination of these factors makes SEI an ideal platform for future growth in the crypto trading space.
Figure 8 -- sei vs. solana vs. ethereum
SEI's lack of full decentralization with only 50 validators and being a permissioned network has caused some criticism in the decentralized finance (DeFi) community. In traditional DeFi, decentralization is considered a key aspect, with many projects aiming for full decentralization through a large number of validators or nodes. SEI, however, has a more centralized approach with a limited number of validators and being permissioned, meaning access to the network is restricted to approved participants.
On the other hand, this limited number of validators can help SEI in terms of network congestion and improved performance. With a smaller number of validators, the network can operate more efficiently and handle more transactions, leading to a better user experience. Additionally, not all DeFi users prioritize decentralization as a key factor. Some believe that the benefits of a more centralized network, such as improved performance and ease of use, outweigh the benefits of decentralization. As such, SEI's approach could potentially attract a different group of users who prioritize these features.
Saturated L1 Market
SEI faces a headwind in the form of saturation in the Layer 1 market, as there are already numerous existing blockchains vying for space and attention. Despite this, it is believed that SEI's competitive advantage and innovative approach sets it apart from the rest, making it a desirable platform for the next wave of decentralized finance (DeFi) applications to be built on. With its unique shared liquidity model, developers can build products faster and more efficiently, benefitting from improved execution speed and built-in price discovery mechanisms. This, combined with its focus on creating a smooth and streamlined user experience, makes SEI a compelling option for the future of DeFi and a catalyst for growth.
One of the challenges for Sei in the current decentralized finance (DeFi) landscape is the popularity of automated market makers (AMMs). Many of the largest DeFi protocols, such as Uniswap, operate on an AMM model rather than an orderbook model. This shift in focus towards AMMs has resulted in a general lack of understanding of orderbook-based trading systems among retail investors.
This lack of understanding can be a hindrance to Sei's liquidity as investors may not be as familiar or comfortable with the orderbook-based trading experience. Additionally, the reliance on AMMs by the larger DeFi protocols may result in a lack of demand for the more complex and nuanced orderbook-based systems like Sei. While Sei's competitive advantage in terms of performance and scalability may attract more advanced traders, the wider retail audience may be deterred by the unfamiliar trading experience. In order for Sei to continue growing its user base and liquidity, it may need to find ways to educate and onboard retail investors who may be unfamiliar with orderbook-based systems.
First, the tokenomics of Sei will be critical to monitor upon launch. This includes the inflation rate, fully diluted value, and the game theory behind the token. Additionally, it will be important to understand the value accrual mechanism for the token and how it drives network effects. This will give insight into the potential future value of the token and the network as a whole.
Second, the ecosystem around Sei will also be key to monitor. This includes the number of projects that are being built, the number of developers contributing to the ecosystem, and the overall level of activity and engagement. This will give an idea of the potential for growth and adoption in the future.
Finally, it will be important to monitor the actual performance of Sei relative to its stated goals and objectives. This includes the network's speed and scalability, its ability to handle large amounts of transactions, and any backlash to its centralization in validators. It will also be interesting to see how the network performs in comparison to other DeFi protocols and its ability to maintain its position as a leading player in the space.
The final remarks about Sei emphasize its uniqueness as a protocol that is attractive to developers and protocols seeking to build on top of a robust, efficient and highly scalable infrastructure. Sei's shared liquidity model, combined with its orderbook-based system and a unified clearing price, provide a competitive advantage over other existing DeFi protocols. These features make Sei a prime candidate for the next wave of DeFi growth and innovation.
In terms of technical improvements, Sei is dedicated to continuously improving its protocol to meet the evolving needs of the DeFi ecosystem. The recent acquisition of additional ecosystem funding for its Asia market entrance is a testament to the team's commitment to expanding and growing the Sei network. As the DeFi space evolves, Sei's technical innovations will be a catalyst for growth and set the protocol apart from others in the market.
Finally, the team and funds behind Sei are a significant strength for the protocol. With investment from major players such as Multicoin in the seed round and market making support from Hudson River Trading and GSR, Sei is well positioned to continue to grow and evolve. With a focus on innovation and a commitment to building a robust and scalable DeFi ecosystem, Sei has the potential to be a major player in the DeFi space for years to come.