Prediction of Intraday Trend Reversal in Stock Market Index Through Machine Learning Algorithms

Abstract

Research on stock price prediction has gained much importance in stock markets because of the immense economic value of being able to predict stock prices in advance. In this paper, we are trying to predict intraday trend reversals in a stock market index. Stock market indices often exhibit sharp intraday trend reversals of significant magnitude. If this reversal can be predicted or detected as soon as the move starts, it will be of immense economic value. Based on past data and statistical methods we can establish support and resistance levels for index and constituent shares. Usually, a sudden change in moves happens as a reversal at these levels or on crossing these levels with strong momentum. In general, traders use statistical methods like RSI (Relative Strength Index), MACD (Moving Average Convergence Divergence) to identify the trend reversals during day trading. One problem with all these methods is that they identify trend reversals with a time lag. We provide in this paper a novel approach to predict such sharp reversals by using machine learning algorithms. We analyze past and present data to predict intraday trend reversals in the index with precision and robustness.