A top-down approach to identifying bull and bear market states

Bull and bear markets receive considerable media and academic attention. It is widely believed that such states are important determinants of wider market dynamics, yet no agreed definition exists. This paper investigates frameworks for ex post classification of asset prices in two-state (bull and bear) markets. An emphasis is placed on identifying state transition points that might achieve consensus. A number of potential difficulties with existing methodologies are highlighted. A principle-based approach is adopted from which a new, flexible, hierarchical methodology is proposed that addresses these issues and permits varying degrees of resolution allowing secondary trends such as bear rallies to be incorporated. The methodology is shown to be optimal under one measure of performance.