High-low price prediction and technical analysis : a thesis presented in partial fulfilment of the requirements for the degree of Master of Business Studies in Finance at Massey University, Albany, New Zealand

Abstract

High and low prices convey additional information beyond closing prices, based on the fact that the improved high-low volatility estimator enjoys higher efficiency than the standard close-to-close variance. Empirical results show that a positive risk-return relationship is exhibited more frequently when predicted high-low prices are applied rather than historical data by taking advantage of the new volatility estimator in 48 countries. In this study, models using historical data contain(1) historical closing prices, (2)historical high-low prices, and (3) the Risk Aversion method, while the three high-low prices prediction approaches include (1) Engle and Granger two-step linear model, (2) Engle and Granger non-linear model, and (3)MIDAS technique. Corporate governance variables, associated with laws and enforcement, have weak explanatory power over investors’ perception of risk. This study also contributes to the validity of technical analysis by showing that high and low prices forecasts are able to generate valuable trading signals and positive returns based on range-based strategy and midpoint strategy. The superior investment performance benefits investors in trading both stock indices and options in U.S. financial markets.