Most of the theories in behavioral finance rely on some kind of psychological biases. However, the potentially boundless set of psychological biases that theorists can use to build behavioral models and explain observed phenomena creates the potential for “theory dredging.” We develop a unified theory of human psychology based on entropy law, the most universal natural law. This unified theory of human psychology will help us determine whether patterns discussed in the finance literature are genuine or the result of data mining. It will also greatly reduce the possibility of “theory dredging” in the future works on behavioral finance.
Because of the inability of efficient market theory to explain the persistence of some patterns in the financial markets, many new theories have emerged to understand these patterns. Most of these theories rely on some kind of human psychological biases and are generally grouped under the category of behavioral finance. However, “the potentially boundless set of psychological biases that theorists can use to build behavioral models and explain observed phenomena creates the potential for ‘theory dredging.’” (Chan, Frankel and Kothari, 2002) Thus it is difficult to distinguish data mining from genuine patterns. It would be very helpful to develop a unified theory of human psychology based on a sound foundation to understand market patterns.
The patterns in financial markets reflect the patterns of information processing by the investment public. Since information is the reduction of entropy and all human activities are essentially entropy processes, it is natural to understand human psychology and market patterns from the viewpoint of entropy theory.
Recently, Chen (2003) showed how entropy theory provides clear understanding of some of the most common patterns in psychology and financial markets. In this paper, we will provide a systematic discussion about the patterns of human psychology and explain how entropy theory offers a simple and unified understanding of these patterns. Since entropy process is the most universal physical process, a unified framework built on entropy theory will greatly reduce the possibility of “theory dredging” in the future works on behavioral finance. It will also help us determine whether some patterns frequently discussed in the finance literature, such as equity premium puzzle, are genuine or the result of data mining.
The remainder of the paper is structured as follows. Section I introduces the entropy theory of information. Section II explains how entropy theory offers a unified understanding of the patterns of human psychology. In section III, we apply the entropy theory of psychology to examine whether a perceived pattern in the financial market is genuine or the result of data mining. Section IV concludes.
Download
An Entropy Theory of Psychology and its Implication to Behavioral Finance
