Research on the design of equity investment based on dynamic panel data and a support vector machine: from thinking to time asynchrony between the publication of macroeconomic data and the disclosure of accounting information

摘要

The main purpose of this paper is to design strategies of equity investment on the basis of analyzing the impact of time asynchrony between the publication of macroeconomic data and the disclosure of accounting information on stock prices. We found that most of only fundamental analyses or technical analyses applied in the literature had limitations. In the light of the stocks' logarithmic rate of return, we built use of the theory of Brown motion to calculate the value-at-risk thresholds of stocks as the benchmark of selecting stock. We established support vector machine models to train four decision-making functions of screening stocks. Finally, we designed the strategies of dynamic stock allocations. In this paper, the fact that we use dynamic panel data and a support vector machine to integrate fundamental and technical analyses is new and original.