Forecasting FAANG Stocks using Hidden Markov Model

摘要

FAANG is an abbreviation for the five major technology companies in the U.S. namely Facebook, Amazon, Apple, Netflix and Google. These firms' technological advancements have a profound influence on global economies: creating jobs, connecting people, supplying goods, and producing entertainment. Therefore, a comprehensive study of FAANG stocks is a substantial contribution to the extant literature. Moreover, since stock markets are highly volatile and depend on several uncertain factors, using appropriate systems for analyzing their behavior becomes critical. The Hidden Markov Model (HMM) conforms well to such a realworld problem: it works on the underlying hidden states (invisible to investors) to predict future stock values (visible to investors). Therefore, the paper introduces a novel research on FAANG stocks by using the HMM to forecast potential stock market prices. Training the HMM with historical data (Open, High, Low, Close (OHLC) values) and testing the recent actual observations of these stocks helps to fulfill the study. The Mean Absolute Percentage Error (MAPE) calculates the efficiency of the model in predicting next day's Close price to be nearly 97% - 99%.