trading-simulator

Basic components

• data
• portfolio
• weights estimator
• knapsack solution

Overview

This project aims to generate profits by trading five different stocks throughout the year, beginning with an initial capital of $1000. It employs a dynamic programming algorithm to identify the optimal trades for each day, and the relevant data is extracted from CSV files and stored in suitable objects.

Motivation

I have extensively studied C++ through various books and completed fundamental courses to solidify my understanding of the language. While I have successfully implemented numerous small-scale C++ programs, I recognize the importance of gaining experience in developing larger, more complex projects. To bridge this gap, I actively sought out interesting projects on platforms like GitHub to further enhance my skills.

Given my passion for the fintech industry and its potential to make a tangible impact on the economy and society, I decided to focus on a trading simulator project. This choice allows me to apply the knowledge and techniques I acquired during my university studies to a practical scenario. By immersing myself in this project, I aim to sharpen my C++ skills and gain valuable insights into the fintech industry.

I firmly believe that engaging with real-world projects, such as the trading simulator, not only strengthens my technical capabilities but also provides me with a solid foundation to pursue a career in the dynamic field of fintech.

Trading strategy

In pursuit of simplicity, I opted to formulate a trading strategy based on moving averages (2, 7, 14, and 30 days).

Knapsack solution for stock exchange

Limitation

• only one CPU, weights estimator could be very slow if the granularity is high

Future work

• another trading strategy
• realtime data every day

How to use

1. Setup:

• Clone the project and navigate to the trading-simulator directory
• Download basic compiler tools. Install cmake and clang using brew install cmake clang for macOS
• If you use Linux or Window, you need to download these compiler tools compatible to your computer system, like sudo apt-get install cmake gcc for Linux

2. Compile and Run:

• Create a build folder with mkdir build && cd build
• Compile the program with cmake .. && make
• Run the program using ./trading_simulator, and observe the generated profit information
• For weights estimator program:
  • cd estimator && clang weights_estimator.cpp -o weights_estimator
  • Run ./weights_estimator to get the result
  • You can also adjust the range and granularity of weights to generate your own result

3. Adding Stocks:

• All the stock data are downloaded from Yahoo Finance
• We use the historical daily price data of SPY (10.01.2020 - 09.30.2022) to calculate the weights
• We will do the trading for these stocks: APPL, AMD, BP, META, NVDA and TSLA
• You can also download stock data from Yahoo Finance, and put them into data folder