Data-driven stock quant investment with Python Part 1

Invest in stocks smartly & strategically!
Take on the challenge of data-driven investing with Python!

🙌 Who is this lecture for?

Most of you interested in this course likely have some experience with stock investing. While some may have made significant profits, most have likely experienced significant losses or been caught up in a series of stocks, often against their will. While everyone's investment preferences and environments may vary, the reasons for investment failures may seem diverse. However, the reasons often appear to be somewhat consistent. This is especially true for beginners, often referred to as "ju-rin-i." What do you think are the reasons?

Humans are inherently vulnerable to investing. Even if we make a firm decision with unwavering determination, even the slightest change in our knowledge or surroundings can cause us to become psychologically unstable. We tend to only accept information that favors our thinking and judgment, or we tend to hypnotize ourselves into believing that our judgment is infallible. This happens instinctively, regardless of our will . This may be due to an unwillingness to experience the sense of defeat that comes from admitting loss, but more importantly, it's likely due to a lack of clear investment criteria or alternative options . So, how do those who invest based on such instincts fare in the stock market? Below are excerpts from reactions in stock market forums and KakaoTalk groups during past market downturns:

How was it? It probably sounds like someone else's story. Investment methods without clear standards or foundations, like the one mentioned above, will only lead to greater confusion and even greater losses when similar situations arise in the future. Regardless of the reason, persistent losses will only make us increasingly anxious, and in extreme cases, it can feel like the market is always moving against our expectations. Because humans are still more influenced by emotions than reason, we're likely to repeat the same mistakes, ultimately leading to the loss of all our hard-earned money. Many of you are probably worried about this vicious cycle and wondering how to proceed with your investments going forward.

Does that mean we really study investing? I don't think so. It's quite ironic. The passion and attitude we used to diligently study day and night to prepare for college entrance exams and land a job at our desired company vanish in the face of the stock market, replaced by speculative trading fueled by greed and desire . Without a thorough understanding of the market or stocks, we simply stare at charts that have no chance of rising, constantly agonizing over which stocks are good and when to enter. In a way, this is a place where we need to prepare for challenges far beyond our own, taking a much longer-term approach and adopting a slightly different approach to survive.

If you have experienced or are still experiencing the instinctive(?) investment as mentioned above and do not have any special plan in mind, how about trying data-based stock investment with Python together? Through this class, we will learn about “computers (programming)” that are more cool-headed, mechanical, and smarter than humans. With this in mind, I'll be using your skills to develop data-driven, systematic, and evidence-based investment habits . For those who struggle day after day, investing their hard-earned money without thorough preparation or a trusted advisor, or for those who have tried everything but struggle because Excel is the only tool they know how to use, this course will use the Python programming language to help you break free from bad investment habits and build a foundation for smart investing .

🙌 Why should you learn “data-driven stock investing” in Python?

As we enter 2020, more than 44 zettabytes of data are generated globally every day. Consequently, the value of data is increasing, and decision-making processes based on it are becoming increasingly important. Programming skills to extract insights from massive amounts of data and perform modeling are becoming a fundamental requirement.

In this day and age , is Excel still a powerful data analysis tool? Not only is Excel complex and cumbersome for processing and automating data exceeding hundreds or thousands of megabytes, but compared to programming languages, it requires a relatively large amount of time and effort to produce the same output. Furthermore, it's extremely difficult to record and organize the analysis process leading to specific results.

However, with the Python programming language , a variety of analyses and diagrams are possible with just a few lines of code. Python's rich library of data analysis tools allows for easy processing and analysis of any type or volume of data, providing an interface that easily organizes and presents the problem-solving process. Furthermore, it can be easily extended to create a single, end-to-end program (or application). Due to this appeal, even the banking and finance industries, known for their extreme conservativeness, have long since begun to use Python to replace Excel.

Python's popularity continues to grow. Its rise to the top of the programming language rankings, surpassing other languages, is largely attributed to the library Pandas . Pandas is a Python library that facilitates any analysis of two-dimensional structured data . Originally designed to facilitate the handling of financial data, it is optimized for financial data analysis more than any other existing tool. I myself use Pandas primarily for analyzing structured data in various fields, and I actively use it in my finance and stock trading projects, where the majority of my code is written in Pandas, reaping the benefits. Recently, Pandas-based libraries that support distributed processing across various computing environments and further enhance operational performance have been emerging, further solidifying Python's position as the definitive data analysis tool.

In this course, you'll learn how to utilize the Pandas library to analyze and process various types of real-world financial data, from extracting investment targets from annual financial statement data, backtesting them, and graphically displaying the results . We'll also explore precautions and ways to improve upon these strategies for more realistic backtesting. This course will provide a valuable opportunity to develop "data-driven investment strategies" for investors who, without the experience to personally verify their strategies, invest in stocks favored by others or simply rely on intuition, as well as those who seek a more quantitative and systematic analysis of corporate value . Furthermore, for those who have memorized Excel functions and macros but struggle to understand how to handle new data, this course will offer a chance to experience the new world of Pandas.

In an era where coding and programming skills are essential, Pandas can be an excellent choice for quickly and effectively analyzing and utilizing the overflowing data. Furthermore, gaining experience in data analysis using Pandas can serve as a great starting point for becoming more familiar with the unfamiliar world of Python. Python's popularity, vibrant community, low learning barrier, diverse libraries, and, most importantly, Pandas, optimized for financial data analysis and processing—all of these are already ready and waiting for you. With these tools, you'll be able to perform a variety of analyses on any financial data, and even backtest your desired investment strategies with complete freedom. All you need to do is prepare yourself to embrace it. This course will be a valuable asset far more valuable than two Samsung Electronics shares ( ≈ ₩100,000) . Why not start now?

🙌 What are the benefits of backtesting and investing in stocks based on your own data?

• Since you will have your own trading rules based on data , you will not need to look at the market, charts, or stock bulletin boards as often.
• Even in rapidly changing market situations like financial crises , you will never be psychologically shaken.
• Sufficient backtesting on past data will allow you to make more informed investments .
• You will be able to make decisions based on the most important factors you have identified through data verification , without being swayed by the news or public opinion.
• You will be able to create a program that is most optimized for you , tailored to your tastes.
• You can backtest tens or even millions of trades while you sleep, commute, or vacation, and even automate your actual trading.

🙌 What can you do with Python + Pandas?

• After completing this course, you will be armed with the formidable weapon of Pandas. By mastering Pandas, you will gain the fundamentals to extract and backtest various investment indicators from financial data (financial statements, daily/minute charts, individual/institutional/foreigner supply/demand, dollar/oil indices, programmatic trading trends, etc.). Furthermore, you will become an expert in analyzing not only financial data but also a variety of structured data.
• Below is a simple toy projectI implemented using only Pandas, using data I collected individually .
  • Analysis of confirmed COVID-19 cases by day of the week

• • Analysis of stock community bulletin boards (checking changes in the number of posts when stock or real estate-related events occur)

• • Analysis of cumulative net trading volume by investor in the KOSPI/KOSDAQ market

• • Create a simple strategy based on the daily gap-up/gap-down ratio (opening price compared to the previous day's closing price).

• • Data-Driven Strategies vs. Benchmark Backtesting Performance Comparison

• Most importantly(?) - You will be able to take the upcoming course "Data-Driven Stock Investment with Python Part 2" without any prerequisites.

🙋 Top 6 Questions Students Ask the Most

Q. What are the differences between this online course and previous offline courses?
Q. Can you also share strategies for making money in your lectures?
Q. Are high-frequency trading strategies like scalping and short-term swing trading included in the course content?
Q. To become a quant, do I need to know everything about mathematics/probability/statistics, including concepts like risk, alpha, beta, and factor models?
Q. Just because past investment data has proven its performance doesn't mean it will perform well in the future, right?

A. The answers to the above questions are covered in detail in ' 6. FAQ ' of ' Section 0. Orientation '.

Q1. Is any prior knowledge required? I've only read a Python grammar book once. Will I be able to follow along? Do I also need to know about stock financial statements?

A1. This course is for those who have some familiarity with programming concepts and the basics of Python . If you are new to Python, I recommend taking my introductory course, ' Anyone Can Learn Python (Click to go) ' (If you are just starting out with Python, why not take this course quickly to review?) Also, since you will be analyzing financial statement data and creating strategies based on it, it is helpful to have knowledge of indicators frequently used in investment (PER, PBR, etc.) (but it is not required ).

Q2. Is the main focus of this course trading or data analysis?

A2. If I had to be specific, I'd say this course focused more on "financial data processing and analysis." While developing my own investment system, the area I spent the most time and effort on was, "How can I quickly and easily process financial data, which represents the same information but has various forms, to fit my desired strategy ?" Initially, I didn't prioritize this aspect. My top priority was quickly implementing programs and strategies that would allow me to buy and sell at any given time. Therefore, I built the system using my limited programming knowledge. Consequently, implementing new strategies often required extensive revisions to the existing code. I kept encountering obstacles along the way, and after much googling, I finally concluded that " data preprocessing" and " removing potential discrepancies in real-world applications (e.g., processing past data to avoid looking into the future) were the most important. I believed that if I could solve this problem, many other aspects could be implemented with less effort. Ultimately, I adopted Pandas to create a realistic and robust investment system. You can think of the content of this class as a curriculum that has been compiled from the experiences I have had during this process .

Q3. Do you actually make data-driven investments?

A3. I'm actually implementing the strategies I developed based on the content covered in this course , and I'm personally seeing satisfactory results. The course content focuses on the elements I found most challenging and the tools I found most useful in developing my investment system. Therefore, every single piece of information and technique covered in this course is applied seamlessly within my investment system. Therefore, I believe the course content is highly credible, and with just a little bit of practical application, I'm confident you'll gain the fundamentals and skills to freely create your own investment strategy.

Q4. There's a lot to build if you want to create an investment system. I'm still a Python beginner, so it'll probably take at least a year to build it. Isn't that too late?

A4. Anyone who tries to do everything at once will feel overwhelmed and frustrated. I've been working on this for over two years now, and I haven't even completed half of the system I ultimately want to build. I'm still steadily adding, testing, and verifying new features. If you were planning to invest for a year or two and then leave the stock market, I wouldn't have anything to say. However, for the average retail investor, who plans to invest for the rest of their lives, the mindset of "I need to quickly complete all the groundbreaking tools I'll use for the next few decades" is similar to the mindset of "I need to hit the jackpot with stocks and make a fortune quickly." From this perspective, the most important thing is to program each essential part of your current investment process, step by step. For example, if your strategy involves trading only one or two stocks, with ample time between buys and sells based on conditions and infrequent cycles, there's no need to build a system that automatically buys and sells through a brokerage API first. This will most likely work well with the ordering function of HTS or MTS, or even by manually placing a buy/sell order. The important thing is how the (financial) data is (pre)processed during this process . There is a huge difference in productivity, and Pandas will greatly help you speed up your work and improve the quality of your work.

Q5. If this is Part 1 of the course, is there also a Part 2 class? What does Part 2 cover?

A5. Part 1 covers the basics of Pandas, applying it to various financial data, focusing on financial statement data , and even learning how to perform simple backtesting using it. Since this lecture (Part 1) is the first lecture, it may feel a bit introductory. Part 2 will cover Pandas techniques that focus on time-series price data (OHLCV) data , and based on this, we will cover static and dynamic asset allocation strategies that require adjusting weights at various periods (monthly rebalancing, 60:40, all-weather, VAA, DAA, etc.) . In addition, we will delve into how to utilize various types of return data (e.g., log return) and implement evaluation indicators related to backtesting (e.g., annualized return, Sharpe, MDD, CAGR, Std) .

Q6. After attending the lecture, I feel like I'm getting better at handling financial data. But obtaining financial data is also a lot of work. 😭 What should I do?

A6. As mentioned in the orientation, data collection and processing is the most difficult and time-consuming task. We also offer a class that teaches you how to freely retrieve and automate the data you need from various sources: Python Web Crawling & Automation to Replace Your Work (click here).

🙌 Things to note before taking the class

• The data provided in this lecture is sample data intended to demonstrate the application of what you've learned. Therefore, it may not be immediately applicable to real-world situations.
• Responsibility for your investments always lies with the investor.

📚 We've prepared this orientation video with as much care as the course content itself. While it's a bit long, it covers the direction and purpose of not only this course but also future courses, and covers topics you might be curious about. Let this orientation spark your desire to learn!

📚 No more one-man lectures where the instructor stands in front of students! Classes where everyone understands the principles and works together! Shall we get started now?

Until the day when everyone in the country invests in Python!