What is Quantitative Trading?

Quantitative trading , also known as Quant in English , is a trading method often used by financial investment institutions. It aims to conduct large amounts of data analysis in a short period of time, and then perform automatic trading through pre-written trading models to improve trading efficiency and avoid the impact of human emotions on trading.

What is quantitative trading?

Quantitative trading, in English , is a trading method based on quantitative analysis.

For ordinary financial investors, whether to hold a stock, when to hold it, how much to hold, when to sell it, and how much to sell it often determine the success or failure of an investment.

Before making a choice, ordinary investors will analyze some data, including the target company’s operating conditions, financial situation, the company’s stock price performance over the past period of time, etc. After analyzing a bunch of data, they will make a choice based on their own subjective judgment, or listen to the guidance of some masters, or follow the mainstream trend of the market.

If an investor plans to invest in multiple stocks, he also needs to collect information about the industry in which the company is located, etc., and the amount of analysis required increases. Professional stock investors spend most of their time collecting and organizing information and then making judgments.

For financial institutions, the amount of information they have to process is thousands of times that of ordinary investors, and the judgments they have to make require eliminating a lot of interference in order to find the most accurate timing.

Whether it is ordinary investors or professional institutions, the investment process can generally be divided into two steps:

• The first step is to collect and organize the given information and find out the rules;
• The second step is to develop a strategy based on the data and judge the stock trend;

The first step is to be “annoying”. The amount of information is too large and too complicated. You need to collect a large amount of data from listed companies, including company fundamentals data, stock trading data, etc.

The second part is “difficult”. After obtaining information, personal judgment on buying and selling will be influenced by personal emotions. Many investors may have regretted: I was too impulsive at the time, I should have calmed down.

With the improvement of information technology and the development of computer computing power, the content of the first step is constantly being simplified. You can easily obtain various information data tables and trend charts, and the time range of data acquisition can also reach a large span.

As for the second part, since it is impossible to conduct targeted data analysis, the investment operations given after computer data analysis are also at the reference stage.

In modern times, with the continuous development and simplification of computer languages, more and more people have mastered computer languages ​​and, relying on their  mathematical abilities, have built calculation models that suit them. After backtesting to verify the applicability of the models, they can perform target object data analysis and make trading point decisions in a very short time. In this process, there is only computer analysis and no interference from human emotions.

This trading method has been proven to be effective over time, so it is used in many large financial institutions. Financial institutions can complete a large number of transactions in an instant, which is now called quantitative trading:

Targeted trading models are written through computer languages ​​such as C/C++, MATLAB, R or Python, and digital analysis and mathematical operations are used to complete massive information collection, data analysis, purely rational strategy decisions and transaction execution.

People who are currently engaged in quantitative trading are called quantitative traders (Quants), and their main job is to:

Strategy identification: Write your own trading model, or find an existing strategy, and decide on the strategy to use and trading frequency based on your own advantages

Strategy backtesting: Apply the target strategy to historical data, and verify through model operation whether the model’s calculation results are consistent with the historical results. If they are consistent, they can continue to be used and make detailed adjustments. If they are inconsistent, the model will be abandoned.

Execution system: Apply the successfully verified trading model to actual transactions, complete automatic quantitative transactions, and minimize transaction costs.

Risk management: Track quantitative transactions based on trading models, identify possible risks in the transaction process, and conduct risk management in a timely manner.

The founder of quantitative trading

Jim Simons is the first great investor to combine computer mathematical models with financial investment. The Renaissance Technologies hedge fund he founded is currently the most successful hedge fund company focusing on quantitative trading.

Since founding his computer-based quantitative trading fund Renaissance Technologies in 1982, Jim Simons has accumulated $23 billion in wealth in just over 30 years with his quantitative trading system.

Advantages and disadvantages of quantitative trading

Quantitative trading is used by most financial institutions and is now also being used by many individual investors. Like many other trading methods, it has its own advantages and disadvantages:

advantage

Large amounts of data can be collected and analyzed quickly, greatly reducing the workload of target selection.

Once the transaction operation trigger point is set, the transaction can be carried out automatically, reducing the workload of daily investment.

Use computer mathematical models to rationally analyze market conditions, determine operating methods, and effectively avoid interference from human emotions.

shortcoming

A single quantitative trading model cannot always be effective in a dynamic market, and parameters need to be adjusted regularly to adapt to changes in the overall market environment.

What are the quantitative trading platforms?

The platforms used for quantitative trading are currently mainly major financial institutions, hedge funds, and other platforms that need to analyze large amounts of trading data.

There are also more and more individuals who, after learning computer languages, write their own trading programs to conduct financial transactions that better suit their own investment characteristics.

The platforms that can be used to write quantitative trading programs are mainly those written in several major computer languages, such as C/C++, MATLAB, or Python.

Platforms for learning quantitative trading are currently very popular and usually provide a relatively complete course system to train high-end quantitative traders from several aspects such as  mathematics , logic, computer language and financial knowledge.

How to build a Python quantitative trading system?

Quantitative trading systems can be written in a variety of computer languages, and Python, as the most popular computer language at present, ranks first in usage rate among multiple languages.

Python is a cross-platform compatible high-level programming language. The open source environment has many proprietary professional library functions, such as:

Scipy, numpy, pandas, matplotlib, quantopian, Zipline, TA-Lib, Pybacktest, etc. can quickly develop barrier-free quantitative trading strategies.

Tensorflow, seaborn, scikit learn, Keras, plotly, and stats can help trading models perform more efficient data mining and trading execution.

SpyderIDE optimizes data visualization in trading models, making financial analysis more intuitive and easy.

As Python’s exclusive algorithmic trading library function, PyAlgoTrade focuses on paper trading, backtesting, real-time trading and technical analysis, bringing more efficient quantitative trading.

Using Python as a computer language to write trading models is the same as the process of formulating all quantitative trading models, which consists of strategy identification, strategy backtesting, execution system and risk management.

But the advantage of Python is that in all processes, its computer language is easier to understand, the logical order is more organized, and it provides multiple exclusive library functions that can be called directly.

During the strategy identification stage, you can call multiple library functions according to the trading characteristics you need to write a trading strategy that is more suitable for you.

During the strategy backtesting phase, professional library functions can conduct more comprehensive data backtesting to obtain more accurate backtesting results, ensuring that the trading model written in the early stage is more effective.

In terms of the execution system, due to the clarity of the language logic, the probability of bugs occurring during model execution is greatly reduced, and no investment benefit points are missed.

During the risk management process, because the language is clear, it is easy to find the adjustment points and make minor adjustments to the data to control the necessary risk management without affecting the complete operation of the entire trading model.

What are the quantitative trading strategies?

Quantitative trading strategy means that quantitative traders write targeted trading models based on the characteristics of their trading styles, collect and integrate the required information, and monitor data and make decisions based on the proposed different trading judgment points. Currently, the more successful quantitative trading strategies verified by the market are:

Alpha Hedging Strategies

Investors face systematic risk (Beta/β risk) and non-systematic risk (Alpha/α risk) in market transactions. The strategy of measuring and separating systematic risk to obtain excess absolute returns, i.e. alpha returns, is called alpha hedging strategy.

Turtle Trading Strategy

The Turtle Strategy is a trend-following quantitative trading strategy that sets parameters for entry conditions, position control, capital management, stop loss and take profit, etc. to conduct automated trading. This strategy can be used as a basic template for the design of complex trading strategies.

Multi-factor stock selection strategy

The multi-factor stock selection strategy is to find some indicator parameters related to the rate of return and build a stock portfolio based on the indicator. If the stock portfolio outperforms the market index, continue to go long and short the futures to earn alpha returns. If it underperforms, go long the futures and short the current stock portfolio to earn reverse alpha returns. It is an important model in current quantitative stock selection.

Double Moving Average Strategy

The basic idea of ​​the double moving average strategy is to establish an m-day moving average and an n-day moving average respectively. The two moving averages will definitely have an intersection. If m>n, the point where the n-day moving average “crosses” the m-day moving average is a buy point, and vice versa, it is a sell point. This strategy uses the intersection of different days’ moving averages to capture the strong and weak moments of the stock to conduct automated quantitative trading.

Cross-product arbitrage strategy

The basic idea of ​​this strategy is to trade using the price difference between two different but interrelated index futures products. Interrelation means that they are mutually substitutable or affected by the same supply and demand factors, such as arbitrage between related commodities or arbitrage between raw materials and finished products. For the market, this strategy can help distorted market prices return to normal levels and increase market liquidity.

Cross-period arbitrage strategy

Similar to the cross-quality arbitrage strategy, the cross-period arbitrage strategy is also a quantitative trading strategy applicable to futures. Cross-period arbitrage is to make arbitrage profits by trading futures contracts of the same index but different delivery months on the same exchange.

Index Enhancement Strategy

This strategy is suitable for index investors. Fund managers use this strategy to keep the characteristic parameters of their recommended investment portfolios higher than the return level of the benchmark index in order to maintain their good investment performance.

Grid Trading Strategy

This strategy is an active trading strategy that uses market fluctuations to make profits. The basic idea is to use the price difference of the investment target within the preset grid range to repeatedly increase and reduce positions. For example, increase positions when the target price breaks through the grid, and reduce positions when it returns to the grid, so as to maximize investment returns.

Industry rotation strategy

This strategy aims to automatically switch between different industries according to the strong periods of different brands in different industries to maximize investment returns.

High Frequency Trading Strategies

High-frequency trading strategies can help investors earn profits in extremely short market changes. Computers can track market trends in real time according to set programs, automatically buy or sell within the set price range, and earn huge profits from price fluctuations through a large number of transactions.

R-Breaker Strategy

R-Breaker is an intraday trading strategy. Based on the closing price, highest price and lowest price data of the previous trading day, with the help of a specific  mathematical model, six price levels are established, from high to low: Breakthrough Buy Price, Observation Sell Price, Reversal Sell Price, Reversal Buy Price, Observation Buy Price and Breakthrough Sell Price. These six price levels are the different operation trigger points of the current transaction. Investors can adjust the gap between the price levels by adjusting the parameters in the model to change the automatic operation trigger conditions. This strategy was rated as one of the most profitable strategies by Future Thruth magazine.