Abstract:

 

Imagine a small inquisitive child playing with his or her toys in the backyard or garden. This is a privileged child who has many new and old playthings. The old standbys are comfortable, but this child also likes to explore the new and most recent additions. An unexpected experience with a new toy leads the child back to the familiar ones and a startling sound, like a car backfiring, may cause the child to abandon the entire ensemble altogether and head for the house.

This picture has some correspondence to the behavior of a financial market. If we think of the market as an organism responding to pleasant and unpleasant experiences, and also if we think that the organism places value on each item it encounters, this will allow us to exploit a very useful behavioral model of utility called unfolding.

Unfolding has a fairly long history. The central idea was originally proposed by Clyde Coombs in the 1950s when he considered how to model liking and preference at an individual level. According to the model, a person’s hedonic response to an item (the degree of utility placed on it) depends on the similarity of the item to an individual’s ideal. The basis for the similarity of interest may depend on several underlying drivers that are not identified in advance and are treated as latent variables. The word unfolding refers to the result of estimating the parameters associated with items’ latent variables, or coordinates, in a multidimensional space. Coombs conceived the items and ideals as deterministic (fixed) points and this led to degeneracies, which despite sometimes fitting the data well, are uninterpretable solutions.

​

This problem was solved by assuming that the items are probabilistic rather than deterministic. In 2001, The Institute for Perception introduced a novel method called Landscape Segmentation Analysis® (LSA) that solved the degeneracy problem. LSA has been applied to numerous types of hedonic data in many product categories and has also been applied to complex sensory variables.

The purpose of this report is to consider an application of LSA to financial markets as a tool in behavioral economics and to add to its already extensive range of applications.

​

​

​