Chinese Crypto Market
Deep Dive into the Chinese Crypto Market: Exploring Trading Habits, MBTI Profiles, and Trending Sectors
Aug 08, 2024 16:10
TechFlow: Sunny and David
Monad Labs: Keone Hon
"DeFi's current inefficiency can't effectively compete with traditional finance's billions of daily orders."
--Keone, CEO of Monad Labs
"In traditional finance, the S&P 500 mini futures contract, a high-volume trading instrument with a nominal value of $250,000 per contract, trades 2 to 4 million contracts daily, leading to a nominal trading volume of up to $1 trillion.
In contrast, DeFi platforms like Uniswap have only reached a total historical trading volume of $1 trillion. Trade sizes in DeFi are also much smaller."
Keone is the founder and CEO of Monad Labs. Prior to founding Monad Labs, Keone served at Jump Trading for 8 years, a leading proprietary traditional and crypto trading company focusing on algorithmic and high-frequency trading strategies. There, Keone managed vast unstructured market data from various exchanges by leading a team of high-frequency traders and engineers, turning chaos into useful information.
At Jump Trading, Keone met his co-founder James Hunsaker, both challenging the current status quo: "Ethereum cannot efficiently support the scale and user experience of future on-chain financial markets." Hence, they are embarking on the reconstruction of a Layer 1 smart contract platform compatible with the Ethereum Virtual Machine (EVM).
With years of formal financial engineering experience, Keone provides valuable insights into the problem of low capital efficiency in DeFi, while also having a keen understanding of technically feasible solutions.
Recently, TechFlow invited Monad Labs' founder Keone to share his observations in the field of DeFi infrastructure. These observations have inspired him to boldly rethink the architecture of EVM Layer 1 and provide more optimized solutions.
Monad Labs completed a $19 million seed round in March of this year, led by Dragonfly Capital. Keone said he and his co-founder place a lot of importance on the startup team, and this seed funding will mainly be used to expand the team and focus on developing a more efficient EVM L1 to prepare for more emerging assets to be chained.
Monad Labs plans to launch a test network by the end of the year and aims to launch the main network in early 2024. The entire team is working hard to complete improvements and launch the test network. Monad Labs will be able to collaborate with developers in different regions in 2024 to support the applications they are building.
Now, let's delve into the interview with Keone's perspective.
TechFlow: In past interviews, you've said that the impetus for founding Monad came from recognizing the huge gap between traditional finance and crypto finance.
How big is this gap, particularly when comparing decentralised exchanges to traditional financial markets or centralised crypto exchanges?
Keone:
Quantitative trading in traditional finance is extremely competitive. Information from exchanges arrives in packets, which multiple competitors receive simultaneously. Machines then recalculate and decide whether to send back an order. In this environment, speed is a decisive factor for winning a trade. This competition has spurred innovation and a focus on low-level details to maximize system performance.
The crypto space is also more volatile, featuring numerous exchanges and less mature technology.
In terms of maturity, traditional finance is at the top, followed by centralised crypto finance and then decentralised finance (DeFi).
In terms of exchange
From a trading perspective, the crypto space is still maturing compared to traditional finance. Major crypto exchanges gained significant volume and usage mainly around 2017-2018, making them relatively new. Traditional exchanges have evolved over a longer period, and implementing new technology is a multi-year effort.
In terms of infrastructure determinism
Many centralised crypto exchanges are hosted on AWS, introducing more variability in system performance. This makes it harder for participants to co-locate near the matching engine server, leading to increased unpredictability.
In terms of user experience
In DeFi, it's common for users to experience significant slippage, sometimes up to 1%. This inefficiency results in higher costs for users, not only in gas fees but also in the actual cost of trade execution. These inefficiencies are notable, especially for trades that would be considered small in traditional finance.
In terms of scale
In traditional finance, the S&P 500 mini futures contract is a high-volume trading instrument, with each contract having a notional value of $250,000. It trades between 2 to 4 million contracts per day, resulting in up to a trillion dollars in notional volume.
In contrast, DeFi platforms like Uniswap have only reached a trillion dollars in total historical volume. The trading sizes in DeFi are also much smaller; a $100,000 trade in DeFi can experience significant slippage.
The inefficiency in DeFi is evident when compared to centralized crypto finance or traditional markets. Users commonly face slippage of 1% or 2% on DeFi platforms, a scenario that would be rare in traditional finance.
Monad Labs aims to bridge this execution gap and elevate DeFi to levels of efficiency seen in traditional markets.
TechFlow: Currently, we're also seeing innovation at the DEX protocol level from platforms like Uniswap and dYdX.
How are these current DEX protocols narrowing the gap between DEX and centralized exchanges (CEX), and why are they still insufficient to bridge the divide between DeFi and traditional finance?
Keone:
Professional traders, often high-frequency traders, provide most of the liquidity in traditional markets. These firms manage risk across various assets and venues, executing hundreds of millions or even billions of orders per day to keep the market liquid and competitive. These firms also compete to narrow the market spread, aiming to make it as small as possible for a given quantity.
In traditional finance, limit order books foster competition among market makers to reduce the spread (Bid-Ask Spread = Highest Bid - Lowest Ask), enhancing user experience. In contrast, DeFi platforms rarely use limit order books. Notable exceptions like dYdX operate on separate Layer 2 solutions, limiting their composability with other DeFi applications.
To bridge the gap between centralized finance, traditional finance, and DeFi, it's essential to create an environment where professional market makers can efficiently update quotes to minimize spreads. For this to happen, on-chain transactions must be cost-effective, as market makers pay gas fees for each quote update.
This brings us to Monad's vision. While at Jump Trading and specifically Jump Crypto, my co-founder James and I identified a need for highly efficient Ethereum Virtual Machine (EVM) execution. Existing environments supporting the EVM were limited in their capacity to process transactions and offer low gas fees.
The best available options for users and developers provide only 100 to 200 transactions per second, or 10 to 20 million transactions per day.
This is insufficient to allow DeFi to effectively compete with traditional finance, which processes billions of orders per day.
There is a big gap between the execution quality of users experiencing a DeFi versus traditional finance. The source of it was ultimately the fact that the underlying blockchain was just very expensive to use.
Ultimately, uniswap is designed the way it is because gas is really expensive. It means really expensive fees for liquidity providers to update their quotes all the time. Therefore, it's better to just allow them to set it and forget it. That means that there is not good capital efficiency.
The capital is spread across a wide AMM curve because the capital is not efficient. When a user makes a trade, they experience a lot of slippage because there's not a lot of capital near the inside fair value in order to trade against.
All this can be resolved if you go fix the underlying premise, which is that gas is expensive.
TechFlow: How does Monad solve the source problem of expensive blockchain and hence high gas fees?
Keone:
Monad has made four major improvements to advance the Ethereum space:
parallel execution of transactions,
deferred execution relative to consensus,
high-performance state access,
performant consensus mechanism based on HotStuff with additional research improvements.
These optimizations address various bottlenecks observed when resimulating Ethereum transactions on a distributed network.
Monad is a fully bytecode-compatible EVM Layer 1, offering 10,000 transactions per second. This throughput considers average transactions from Ethereum's history, not just simple transfers.
In tech terms, it can handle one billion transactions per day, supporting apps with large numbers of daily active users. This marks a significant increase compared to Ethereum's one million transactions per day and other EVM-compatible platforms, which can only offer roughly 10 million transactions per day.
Developers can easily port Ethereum-based applications to Monad without making changes, maintaining bytecode and RPC compatibility. This ensures tools like MetaMask and Etherscan can also be deployed seamlessly.
One of Monad's key innovations is parallel execution. Transactions remain linearly ordered, but during execution, the system parallelizes the work. For example, if the first and fourth transactions are interdependent because they affect the same state, the system will execute transactions 1, 2, and 3 in parallel but will sequence transaction# 4 after 1.
From a user perspective, the only change is increased throughput. There's no risk of transactions interfering with each other. The system uses optimistic parallel execution, ensuring that transactions committed in parallel align with their original order. If unexpected common dependencies arise, the latter transaction will be rolled back and rescheduled.
Monad is the first to introduce parallelization to the Ethereum Virtual Machine. While other blockchains like Solana have implemented parallelism, they operate under different assumptions and standards.
Challenges mainly involve proper scheduling or "pipelining," akin to how modern CPUs manage a pipeline of instructions and execute them in parallel to enhance speed and capacity. We believe this approach is fundamental to efficient execution and will likely be adopted by other blockchains in the future, most of which currently use a single-threaded execution model.
TechFlow: How do we understand Monad's intrinsic business drive in challenging the EVM status quo?
Keone:
In my mind there's a couple of major business models in Defi right now.
One of them is exchanges to allowing people to transfer risk. That's clearly a valuable service. People are willing to pay for that.
Another one is just the actual interest. The interest rate on U.S. dollars are relatively high, and there are business models like MakerDAO that enable the creation of synthetic dollars in the form of Dai through their protocol. With MakerDAO, the ability to create synthetic dollars allows for the issuance of loans that earn interest, making it a viable and robust business model.
ETH staking. Ethereum protocol generates several billion dollars worth of fees per year.
Understanding these different business models and identifying real user demand helps to predict other valuable services DeFi can offer. While short-term incentives like yield farming may attract users temporarily, long-term sustainable business models matter most.
TechFlow: We have seen non-financial applications other than DeFi emerge lately. What is your view on non-financial applications and the subsequent machine learning required to power their user experience?
Keone:
My experience with machine learning has shown me its potential as a powerful tool for precise predictions in consumer-facing applications, enhancing user experience. Examples include the Twitter feed algorithm and Tinder's matchmaking.
In the crypto space, while decentralized finance is a major focus, machine learning can extend its utility to various applications, ultimately aiming to improve user experience. The Monad team is actively exploring ways to enable ML-powered apps on chain. The key challenge is integrating machine learning outputs into blockchain. Although computation is relatively inexpensive in the Ethereum ecosystem compared to storage, some hurdles still need to be overcome to realize this integration.
TechFlow: We are a crypto media that focuses on the Asian market. What are your views on the state of the DeFi market in Asia?
Keone:
Major factors influencing blockchain adoption include institutional efforts by countries, user demand for personal finance, and active developer communities. In some developing countries, people are using crypto for personal finance, a significant catalyst for building infrastructure and payment rails. This is particularly true for those dealing with high inflation rates and seeking stable assets like USDC.
One interesting observation is that some Asian countries excel in several of these factors simultaneously. For instance, Hong Kong is making concerted efforts to integrate crypto into various governmental services, making it strong in the institutional adoption front.
The developer community in Asia, particularly in Southeast Asia, is robust with a significant number of web 3 developers. User behavior varies by country, but where decentralized finance already offers better options than traditional finance, it acts as a strong catalyst for growth.
Monad docs:https://docs.monad.xyz/
Recommendation
Lets Make the World a Better Place with Raullen Chai, IoTeX CEO
Jun 06, 2024 13:18
Interview with Wormhole Foundation COO: Uniswaps choice, founderless organisation, and 900m message milestone
Jun 06, 2024 23:21
Interview with Arweave Core Contributor and everVision CEO: Discussing the Future Roles of Ethereum, Bitcoin, and Arweave
Jun 06, 2024 23:07