A Portfolio Selection Strategy Based on the Peak Price Involving Randomness

Abstract

We propose a system based on the peak price involving randomness (PPR) for the portfolio selection. In the light of the relative price forecast in the paper entitled “reweighted price relative tracking system for automatic portfolio optimization” by Lai et al., which automatically assigns different weights to the predicted relative price based on the performance of each asset, we will determine the proportion of each stock by the three factors: The average price of the daily price divided by current price, the ratio of the peak price of daily price to current price, and the random value of the relative price of stock in the time window. A large number of the experiments on the five datasets show that the PPR is superior to some recent portfolio selection system in many aspects. These results suggest that the PPR is an efficient automatic portfolio optimization system, at least on the datasets.

Keywords

• Automatic portfolio optimization
• average price
• peak price
• random value