Algorithmic Trading System Development

Abstract

This project aims to create an automatic trading system based on mean-variance portfolio theory and technical analysis techniques in the aspects of stock selection, risk management and buy-and-sell strategy. The system focuses on global stock market. For building a practical system for real trading, the studied for stock data extraction has been carried out. In stock selection, a tailored stock filter was created to expose some potential stocks in the market, getting insights from market psychology based on the price movements and volume. In risk management, mean-variance portfolio theory was applied in risk analysis, weighing the expected returns and price volatility, to generate an optimal portfolio solution with maximum Sharpe ratio. The concept of covariance shrinkage was also included. In buy-and-sell strategy, three regular technical indicators including Relative Strength Index (RSI), Commodity Channel Index (CCI) and Moving Average Convergence / Divergence (MACD) were utilized to evaluate the trend of a stock. In the report, these three indicators were explored deeply, concluded what information can be extracted from. The result is a web application that provides a user interface for implements of stock filter, mean-variance portfolio management, trend prediction and strategic back-testing.