Multi-factor Strategy Using Linear Weighting Model and Time Series Analysis of Returns

Abstract

Quantitative methods of stock picking and timing have become a popular approach today, and choosing good factors and quality models can often lead to high yields. In the current Chinese stock market, the small-cap factor has an amazing performance. In recent years, simply through the small-cap factor stock selection can get more than 40% annualized return. In this paper, the author intends to perform factor clustering through the clustering method, select quality factors from each category to form a factor portfolio, and solve the pre-coefficients of each factor through the linear regression model as the calculation of the weights, and then combine the small-cap factor with the resulting composite score to pick stocks. After the completion of the strategy, the robustness and feasibility of the strategy is analyzed through time series analysis of daily profit and loss, and the ARIMA model is used to predict the strategy's future returns in the future.