Whitepaper
Version 1.0 – May 2023

1. Trading levels and environments

Trading levels

uquant whitepaper 1 1 trading levels

Trading environments

uquant whitepaper 1 2 trading environments
uquant whitepaper 1 2 diagram a

2. Principles of Quantitative Trading in Cryptocurrency Markets

Common trading styles often rely on the intuition and inaccurate experience of humans about the market. Trading patterns, indicators, and information are fed into traders’ minds as an unconscious routine without any discernment. They are promoted and used as magical tools that can capture the market. This mindset leads to over 905 of traders failing.

Fundamentally, the market is dominated by supply and demand, manipulation, events, behavior, and the unique architecture of each market. Countless variables affect the market, making it a difficult or impossible environment to predict, more precisely a probabilistic distribution environment. The market tends to react excessively, going up and down too much due to the psychology of humans affecting market participants, creating fat tails within the returns distribution curve. The excessive reactions and tendencies to generate sudden rewards provide a statistical basis for a method that will generate rewards over time.

The best way to explain this is to imagine taking a handful of salt and throwing them into the air. Suppose we could then measure the distance of each salt grain from the point of throwing. We will see that most of them are relatively close to ourselves, with a few exceptions that have gone further away. If you create a chart showing the distribution of results, the chart will resemble a bell curve, which is a typical and “normal” distribution.

uquant whitepaper 2 diagram a

The horizontal axis represents the distance each salt grain travels. The percentage scale shows how many grains move over a certain distance.

Suppose we continuously buy and sell randomly in the cryptocurrency market, while measuring and recording the maximum potential reward of each trade in thousands of trades and thousands of days. If we construct a version of the above chart with those results and place it on top of the previous chart, the result will look like this, with dotted lines representing the market’s reward distribution:

uquant whitepaper 2 diagram b

Therefore, it can be asserted that the market has a high speculative nature, as the cryptocurrency market generates higher rewards, both positive and negative, than expected from a “normal” reward distribution model. A large number of excessively high price events occur more than normal due to simple randomness. This means that the market brings more big winners and losers than realistically expected.

Using tight stop-loss levels will eliminate events of excessive losses and using wide reward targets will allow capturing the “fat tail” of excessively positive returns.

The focus of quantitative trading strategies is volatility, risk, and probability, through mathematical models, data, statistics, and the support of computer science and AI to create effective strategies that exist Alpha in high-latency trading levels, optimize processing speed with low-latency trading levels. In essence, quantitative trading is a scientific empirical research method, the term “quantitative” itself describes the ability to measure, calculate, represent mathematical equations, from which objective conclusions can be drawn. It is opposed to qualitative, which has the color of mystery, fascination, and blindness of belief.

Most importantly, quantitative trading believes that there is no single trading strategy that allows for generating Alpha forever, practicing quantitative trading is a meticulous, constantly self-adjusting process to fit the market or discover gaps in the structural mechanisms of different markets to find a non-contact trading method with prices. We can intuitively know that there are times when this method is highly effective, but at other times it is no longer effective and another method becomes effective, although it was previously ineffective. It is just an illusion of probability distribution and randomness.

Basic attributes of a quantitative system:

  • Risk limit: Consider risk as inherent, limiting risk at a permissible level will make the investment portfolio overcome periods of inefficient reward distribution. Simply put, the lower the risk limit, the longer you can participate in the market in the long term.
  • Maximize reward: The market itself does not specifically reflect any specific fixed reward level, determining reward needs to go through the reaction of fluctuations over time in the market and trading context. Letting rewards run freely until there is a change in the probability distribution is the method to maximize potential.
  • Identify fat tails: Only trade when there is an advantage in distribution. The number of trades is proportional to rewards in low-latency trades but often inversely proportional to rewards in high-latency trades.
  • Optimize order volume: Through formulas such as the Kelly formula, we will know how much of the capital to allocate to a trade order. If the capital allocation is less than the Kelly%, it is impossible to maximize rewards, conversely, if it is more than Kelly%, it is impossible to minimize risk and rewards will also decrease. Many people often mistakenly believe that the larger the order volume, the more rewards it creates, which is a misconception. The order volume itself causes market volatility, which can cause price slippage, causing the market to transition from the old state to a new state, no longer as initially calculated.

K = p x B (1 p) / B

Where:

f = fraction of wealth wagered or % of making the highest reward on investment or gambling.

B = fractional odds (reward to risk) or the ratio of the win to loss

p= probability of winning against the odds

q= probability of losing or (1-p)

  • Backtesting and Monte Carlo Simulation: Testing trading methods based on historical data and simulating countless market scenarios to challenge the objectivity of the method.

Thus, quantitative trading has the following characteristics:

  • Based on measuring with large-scale data sets, avoiding subjective emotions with small-scale data sets.
  • Implemented by computer systems to eliminate subjective emotions.
  • Uncertainty about strategy or multiple strategy combinations to fit the market state. The
    market does not move according to our thinking.
  • Avoiding price prediction if positive. Deploying non-collision methods is optimal.

3. The formula of the central mechanism of Automated Market Maker (AMM) protocol.

uquant whitepaper 3 diagram a

In 2016, Vitalik Buterin proposed the idea for a decentralized exchange (DEX) that would use an automated market maker on the blockchain.

In 2017, Hayden Adams turned this idea into a functioning product.

In November 2018, Uniswap was launched, marking the beginning of the DeFi era. What made Uniswap unique was its solution to the high slippage problem for illiquid assets on order look exchanges. This problem existed because there were few incentives for professional market makers to provide liquidity for rare>y traded assets. Anyone can become a market maker :y depositing assets into a poo> and earning fees based on the amount of trading activity.

Thus, quantitative trading has the following characteristics:

x * y = K (K remains constant during trading).

This formula governs a pair of 2 tokens: x for token0 and y for token1. When a quantity of x0 token0 is added to the pool, we will receive y0 = y – K / (x + x0) token1; (x0, y0) is the input-output pair. The property of this formula is that it ensures that: for x0, we will receive y0 (example for Buy) and vice versa, for y0, we can get back the initial x0 (example for Sell). This is also the most important symmetric property of automatic market making.

4. Breakthrough only exists in the DeFi space – Flashloan

uquant whitepaper 4 diagram a

A flash loan is a loan that is borrowed and repaid in the same transaction. The borrower does not need to provide traditional requirements such as income proof, reserves, or collateral.

This lending protocol can be achieved using smart contracts in DeFi trading. Smart contracts establish rules for flash loans and require borrowers to repay the full amount borrowed, possibly with a fee, before the transaction is completed.

If this rule is not met, the smart contract will automatically reverse the transaction, and the loan will be cancelled as if it had never occurred.

Flash loans usually take place within seconds, which is how they can provide unsecured loans because borrowers must repay the entire amount borrowed almost immediately.

To put it simply, imagine borrowing from a bank to purchase an asset and being required to repay the entire loan within a few seconds.

Flash loans are very useful in DeFi and have three main use cases: arbitrage, collateral swap, and self-liquidation.

5. Price arbitrage trading and extracting MEV value.

Arbitrage trading is an immediate trading strategy of a certain asset on different markets to reward from small price differences. Traders can move between different exchanges to search for small differences in the quoted prices of different assets.

uquant whitepaper 5 diagram a

For example, observing the ETH/USDT pair trading on Bitstamp at $1,870.57, while the same ETH/USDT is trading on Bitfinex at $1,869.11. That’s a difference of $1.46. A trader can buy ETH on Bitfinex and sell on Bitstamp to make a reward of $1.46 per ETH.

uquant whitepaper 5 diagram b

The same applies to decentralized exchange (DEx) protocols such as UniSwap, PanCakeSwap, etc. Flash loans are used to take advantage of price discrepancies on DEx, due to their immediate nature and applicability on the Blockchain.

uquant whitepaper 5 diagram c

The above is an example of an arbitrageur on Curve and UniSwap.

Here, DAI/USDC is trading on Curve at $1, while DAI/USDC is trading on UniSwap at $0.99. A
trader can take advantage of this opportunity by using a flash loan as follows (as shown in the
figure above):

  • Take out a flash loan of 100,000 DAI from Aave.
  • Exchange 100,000 DAI for 101,010 USDC on UniSwap.
  • Exchange 101,010 USDC for 101,010 DAI on Curve.
  • Repay the 100,000 DAI loan plus a 0.09% fee, resulting in a total of 100,900 DAI for Aave.
  • Keep 110 DAI as reward.

All of the above actions are carried out through a smart contra=t in a single transaction (about 3 seconds on the BSC network). We don’t need to have capital on hand to do this, and there’s no risk of losing capital. When you send the transaction, you either win or the transaction is refunded. You only pay a small gas fee, typically less than 10 cents.

In practice, an arbitrage trade looks like this:

uquant whitepaper 5 diagram d

We can easily observe that this trade performed a series of actions: from 8,608 USDT → WBNB → LVL → 8,746 USDT. The CHI token at the end of the transaction is often used in arbitrage trading to burn and can significantly reduce Gas fees. The trade generated a reward of 138 dollars in just about 3 seconds! Of course, we need to subtract the transaction’s Gas fee, which is 26 dollars, and the final reward is 112 dollars. This is a significant amount of money in terms of time efficiency.

Arbitrage trading on Defi protocols has the following characteristics:

  • It is performed by exchanging tokens through multiple similar and cyclical protocols: A → B → C → A. The length of this cycle is not fixed and is proportional to the computational complexity. The longer the cycle, the harder it is to detect due to higher computational processing requirements.
  • Gas fees have a significant impact on the reward of each trade. Some trades only generate a few cents in reward due to Gas fees. Even for trades with higher Gas fees than the reward received, it is no longer a rewardable arbitrage opportunity.
  • Competition is increasing over time, but that does not mean opportunities will not exist. The nature of the Defi space is vast, with a large number of trading pairs operating on protocols that are difficult to quantify. Therefore, popular trading pairs are often reserved for experienced and powerful arbitrage traders, while less popular or new trading pairs still create many opportunities for new traders to enter the market.

The arbitrage trading process is carried out through 5 steps:

  • Step 1: Infrastructure and network preparation. Investment costs for RPC providers, Mempool, and high-computational concurrent computer systems, as well as strong bandwidth and geographic distance in peer-to-peer connections between peers (Nodes, Peers), are all required. This affects the speed of transactions.
  • Step 2: Perform simulation or heuristic estimation of the potential reward generated by each transaction. The Bot system listens to events in the Mempool and analyzes to detect price difference opportunities, determine the sequence of actions needed to achieve reward from the price difference. This step will perform data extraction techniques, comparing a batch list of similar or interconnected trading pairs in different protocols to calculate rewards and determine the existence of reward. This is the most difficult step as it requires a lot of computing power in the infrastructure.
  • Step 3: Use smart contracts to synthesize the sequence of actions needed into a single transaction. Each individual operation command to exchange tokens through protocols will be compressed.
  • Step 4: Put the aggregated transaction into the Mempool and compete for fees, speed, and infrastructure power, including gas fee calculation. This is the most important step as it determines the success of the transaction. Bidding high gas fees to take the top position in the mempool can erode all rewards, while the speed of sending requests to the mempool creates a significant ranking difference. Arbitrageurs will have to race to optimize in the above aspects or perform special techniques to gain an advantage.
  • Step 5: The transaction is added to the blockchain by a block miner or validator, and success results in a reward or a failed transaction due to competition failure or calculation error, which may result in a small gas fee loss.

During the period from January 1st, 2021 to July 26th, 2022, arbitrageurs earned $138.05 million on the Binance Smart Chain. On May 12th, one arbitrageur earned $320,000. What happened in the cryptocurrency market at that time was LUNA and UST. This was the day when insufficient liquidity in the LUNA protocol caused the UST stablecoin to lose value compared to USD. Then, automatic minting of batches of LUNA tokens caused their price to plummet. When things like this happen, a massive sell-off occurs. Large sales volumes on DEX protocols mean significant price changes. This creates large price differences between different DEXs (e.g. Pancakeswap and BiSwap). And this is what arbitrageurs expect. When people start to panic, they tend to sell faster rather than at the optimal price. During normal times, they would split their orders, start selling in pairs with higher liquidity, wait for the price to recover, etc. But that day, they didn’t have time to do all of that. Moreover, many positions in the lending protocols on the chain were over-collateralized (often used for short selling) and were closed by those liquidating. Collateral assets were automatically sold to cover the loans. These massive automated sales also created inefficiencies. The total reward for all arbitrageurs on that day was over $2 million.

The more volatile the market, the higher the efficiency of arbitrage trading, especially during market downturns. The arbitrage trading model can be expanded to many different chains.

6. Automatic token supply management mechanism

We will consider another use of the formula x * y = K. Because it is symmetrical, it has the ability to create a market equilibrium. But its most important but lesser-known function is to convert a fixed-value investment into a variable investment, which can be applied to mana-in- the token supply of a project.

We need to issue a token with a maximum supply of y = 18 * 10^6 (MUQT) and assume that we start with x = 1.8 * 10^6 (USDT):

1.8 * 10^6 * 18 * 10^6 = 32.4 * 10^12

Note that at this point, no tokens are in circulation on the market: the circulating token amount N = 0.

An investor wants to buy MUQT with a capital of $1 million (x0 = 10^6). We can calculate y0 = 16,363,637 (MUQT), and the contract will mint 16,363,637 MUQT tokens → N = 16,363,637. Conversely, if the person sells the purchased tokens, they will be burned, and the circulating token amount will return to N = 0. If there are many people participating in the buying and selling process, the minting and burning actions will occur continuously. Based on the order and holding time of the tokens, they will have different rewards or losses. The circulating token amount directly reflects the trading volume of the token on the market, and the token supply automatically balances with its value in the pool. This mechanism unifies the Swap feature with supply management. The value of x is crucial in this mechanism: the lar-er the value of x, the smaller the speed and number of tokens minted. The x/y ratio reflects the initial token price.

This design mechanism creates the following advantages:

  • No mining rewards (No Pre-Mine)
  • No market advantages, so it is fair for everyone
  • Token distribution based on equal investment
  • Sustainable token value if combined with the Lock mechanism because there is no concept of circulating tokens outside the protocol.

MuSwap is a decentralized protocol built on the above mechanism, and MUQT is the token representing the μQuant platform that is directly distributed on MuSwap.

Token information:

  • Token name: MUQT
  • Maximum supply: 18,000,000
  • Decimal: 18
  • Starting price: $0.1
  • Token type: Community Fund Share
  • Protocol: Automatic balance supply (ABS)
muquant mint and burn flow

7. What is μQuant?

μQuant is a decentralized quantitative trading fund that operates reward mining on DeFi, starting with the arbitrage trading system on the BSC chain. uQuant officially pre-launched in June, 2023 with a reward-sharing model and is operated by data analysts and blockchain experts. Decentralized finance, or DeFi, has become increasingly popular in recent years and the use of mathematical models and algorithms can help identify patterns and opportunities that human analysis may overlook. This is a true goldmine to explore and earn enormous rewards. The difference from the rest of the market is that we share rewards with everyone based on a unique architecture. We believe that the enormous rewards from the decentralized finance space are a fair opportunity for everyone.

8. Reward-sharing model and features of the μQuant platform

muquant flow chart

Reward-sharing model:

  • Automatic supply and price balancing
  • Real-time transaction data
  • Verifiable rewards
  • Compatibility with all types of wallets
  • Available to everyone

9. The price scale of the token in correlation (capital – reward) with its value.

uquant whitepaper 9 diagram a

The x-axis reflects capital and reward, and the y-axis reflects the token price

The simplified function for the relationship between the token price and the total invested capital and reward is:

uquant whitepaper 9 diagram b

  • At a capital and reward scale of 3.892 million US dollars, the token price reaches 1 dollar.
  • At a capital and reward scale of 16.2 million US dollars, the token price reaches 10 dollars.
  • At a capital and reward scale of 55.12 million US dollars, the token price reaches 100 dollars.

uquant whitepaper 9 diagram c

10. MuSwap – A Standard for Fair Value Distribution

In previous tokenomics models, tokens were issued and distributed by project management teams, with a clear fan chart outlining the distribution purpose. However, the inherent advantage has been completely biased towards developers, who own a large supply of tokens before they are traded on the market, making the use of capital unverifiable. This problem has not been solved for a long time. Then, another problem arose when Uniswap and similar protocols came into operation, making Defi develop strongly. To operate, protocols need to be provided with liquidity, initial supply, or to liquidity for outside or useless tokens. This means that the project will have a large amount of capital that cannot be used for a long time or it becomes a loophole for Rug pull behavior by ghost projects. This mechanism is not beneficial for the development of a project, while it can cause serious harm to the community. Based on the mechanism of automatic management of token supply, MuSwap is designed to become a standard for fair value distribution, ensuring the rights of the entire community while providing development capital for the issuing project. This is achieved through the consistency between token distribution and transaction activity from the beginning.

The integrated tokens on MuSwap must meet the following standard set:

  • No initial supply
  • Launch time or start vote publicly announced in advance
  • Fees and fee distribution are pre-set
  • Limits are adjusted appropriately through voting by the community

With the above standard, the protocol achieves:

  • No market cap tokens
  • No initial locked liquidity
  • No Rug pull
  • No bias

The value of tokens can be predicted and the value received by the community depends on their expectations and participation time in the project, avoiding unexpected and unfair factors.

11. Voting proposals in the MuSwap protocol

A decentralized autonomous organization (DAO) built into the protocol enables community voting on the following proposals:

  • Proposal regarding fees
  • Proposal regarding total supply limit
  • Proposal regarding account-based limits
  • Proposal regarding transaction frequency
  • Proposal regarding moving funds to external protocols

Voting is based on token-weighted voting, which ensures transparency and protects the interests of the community through community consensus.

12. Roadmap

  • 04/2023: Launch MuSwap protocol and price difference trading system on the BSC chain.
  • 08/2023: Expand price difference trading system on Arbitrum and Sui networks.
  • 11/2023: Expand MuSwap protocol on the SUI network.
  • 2024: List MUQT token on leading global exchanges.
  • 2025: Announce the quant trading platform for the global community.