Automated stock price prediction using machine learning

Abstract

The financial market is a dynamic and composite system where people can buy and sell currencies, stocks, equities and derivatives over virtual platforms supported by brokers. The stock market – also known as the equity market – is considered one of the most dynamic components of the free-market economy. It allows investors to own shares in public companies through trading, either by exchange or over-the-counter markets. Thus, giving them the opportunity to make money by investing small amounts of capital which makes for a low-risk endeavor (initially at least), as opposed to opening a new business. Stock markets are however affected by a multitude of factors, making them uncertain and volatile. Which is why we have invested enormous amounts of time and effort trying to predict stock price movements, for by understanding those complex movements can be highly rewarding. Humans are capable of taking orders and submitting them to the market, but automated trading systems (ATS) on the other hand – implemented via a computer program – can perform quite better and with higher momentum when submitting orders. In order to evaluate and control the performance of these ATSs however, risk strategies and safety measures must be implemented; strategies that are based on human judgement. Many factors are incorporated and considered when developing an ATS, for instance, trading strategy to be adopted, complex mathematical functions that reflect the state of a specific stock, machine learning algorithms that enable the prediction of the future stock value, specific news related to the stock being traded, and so on. In order to predict market movements, investors traditionally analyzed stock prices and indicators, in addition to news about these stocks – the latter being of utmost importance when trying to understand stock price movements. Most of the previous work in this industry has focused either on labeling the released market news (positive, negative, neutral) and showing their effect on t