Surveying the best volatility measurements to forecast stock market

Abstract

This paper investigates methods to forecast future adjusted price of S&P 500 by using geometric Brownian motion (GBM) and geometric fractional Brownian motion (GFBM) for better investment decision. Four types of formulas are used to find the appropriate volatility measurement that may provide forecast value which closely resembling to actual movement of stock price. The evaluation of forecasting methods is computed by the mean absolute percentage error (MAPE).The findings showed high accuracy in all forecasting methods, with all MAPE are less than 10%, with the best forecasting method is GFBM with stochastic volatility which follow fractional Ornstein – Uhlenbeck (FOU) process.