The Mystery of Market Movements: An Archetypal Approach to Investment Forecasting and Modelling

Chapter 8Developing a Conceptual Measurement Methodology Based on Archetypal ForcesPart III: The Model

When the time series of the symbol counts have been constructed and generated, they can be tested one by one by symbol for leading correlation versus time series of financial assets. However, if correlation is found, it is generally of a partial nature, given that the characteristics of archetypes are singular in direction, like that of risk averse versus risk aggressive, and will subsequently only indicate price movements either up or down and rarely both. This puts constraints on which statistical test methods one can deploy; for example, using R-square values comparing two time series with such different characteristics becomes less suitable as they are not designed to cater to one-sided leading correlations but rank correlation parallel in time and symmetric higher.

Given that the symbol words have been extracted from overwhelmingly (American) English media reflecting that particular culture and psychological environment, testing needs to be aligned with investment alternatives that typically dominate within that sphere. This would include the broader US stock market index, the S&P 500, an associated “fear” index used to gauge appetite for risk, i.e., the VIX, and some specific industrial equity index, which can be assumed to be linked to particular archetypes, such as war-related industries, gaming-related industries, and healthcare-related industries. The focus needs ...