Resource Allocation in the Brain and the Capital Asset Pricing Model

Abstract

What happens when information reaches the human brain? In economics, a black-box approach to information processing in the brain is generally taken with an implicit assumption that information, once it reaches the brain, is accurately processed. In sharp contrast, research in brain sciences has established that when information reaches the brain, a mental template or schema (neural substrate of knowledge) is first activated, which influences information absorption. Schemas are created through a resource intensive process in which finite brain resources are allocated to different tasks, with resource allocation in the brain having an impact on the structure of schemas. In this article, we explore the implications of this richer view from brain sciences for the capital asset pricing model (CAPM). We show that two versions of CAPM arise depending on how the brain resources are allocated in schema creation. In one version, the relationship between beta and expected returns is flat. High-alpha-of-low-beta along with effects akin to size, value and momentum (with long-run reversal) arise in this version. Furthermore, in this version, value, small size, and low beta stocks suffer a greater price decline in bad times. We conjecture that this is the version which is normally observed. In the second version, the relationship between beta and expected return is strongly positive with an implied risk-free rate which could be negative. Given recent empirical findings that the relationship between beta and average return is steeper at specific times (at market-open or when macroeconomic announcements are made), an intriguing possibility is that this second version is observed at such specific times.

Keywords: CAPM, Value Effect, Size Effect, Momentum Effect, Betting-Against-Beta, Schema, Neuroscience

JEL Classification: G12, G10