Intelligent Portfolio Theory and Application in Stock Investment with Multi-Factor Models and Trend Following Trading Strategies

Abstract This paper documents a computational form of the Intelligent Portfolio Theory for investing and trading in stock markets with two multi-factor models and two trend following trading strategies. The Intelligent Portfolio Theory for stock investment goes beyond the classical portfolio theory by using multi-factor models for stock selection and quantitative trading strategies instead of buy-and-hold. In China A-share market, a special multi-factor Model 1 is developed targeting at the food and beverage sector, which is made up of 7 effective and non-redundant factors selected from 20 candidate factors representing 6 categories: valuation, technology, size, profitability, growth ability, and solvency. A more general multi-factor Model 2 targeting at all the stocks in the whole stock market is also developed, which is constructed as a rule-based system integrating fundamental and technical analysis inspired by the empirical approach of CANSLIM. The multi-factor models are used to forecast or rate the future return of each stock in the sector, and then a small number of stocks are selected to form a portfolio. Each selected stock in the portfolio is then traded using a trend following trading strategy. Two strategies are developed, Strategy 1 with the high-low price channel breakout, and Strategy 2 with the parabolic stop and reverse indicator. The first intelligent portfolio trading system integrates Model 1 with Strategy 1, and is tested on an in-sample data of 7 years and also on an out-of-sample data of 2 years. The second system couples Model 2 with Strategy 2 and is tested on a data of 2 years. Both intelligent portfolio trading systems generated significantly useful performance in terms of annual returns and maximum drawdown. The test results confirmed the effectiveness of the Intelligent Portfolio Theory with proposed multi-factor models and trading strategies.