This article addresses well established return forecasting challenges via frameworks that focus on the sign of the change in asset index excess returns using a family of GARCH models. It investigates them in the literature's original S&P 500 index to study the predictive power of information demand proxied by Google's internet search vector index and finds evidence suggesting that an efficient trading strategy stemming from this study can be constructed. This article is aimed at academics from undergraduate level up, and thus will explain all mathematical notations to ensure that there is no confusion and so that anyone - no matter their expertise on the subject - can follow.

Find out more on the article page.