Stock Price Prediction using Principle Components

TL;DR

This paper introduces a PCA-based method for stock price prediction that reduces data dimensionality, leading to improved prediction stability and performance evaluated through error metrics and volatility analysis.

Contribution

It presents a novel application of PCA for stock prediction using covariance data, enhancing prediction conditioning and robustness.

Findings

Reduced mean squared error in predictions

Improved directional change accuracy

Lower volatility in forecasted prices

Abstract

The literature provides strong evidence that stock prices can be predicted from past price data. Principal component analysis (PCA) is a widely used mathematical technique for dimensionality reduction and analysis of data by identifying a small number of principal components to explain the variation found in a data set. In this paper, we describe a general method for stock price prediction using covariance information, in terms of a dimension reduction operation based on principle component analysis. Projecting the noisy observation onto a principle subspace leads to a well-conditioned problem. We illustrate our method on daily stock price values for five companies in different industries. We investigate the results based on mean squared error and directional change statistic of prediction, as measures of performance, and volatility of prediction as a measure of risk.