Friday, December 17, 2010

Market Ecology

AAA - Applications of Ant Algorithms

Trying to work on a model that pulls from ant behavior...looking for parallels for application to fund selection (net long/short positions). A simple model for trail selection and formation in ant colonies is based on how an ant travels over or through a substrate as a function of pheromone trails, presumable laid down by predecessors (this is discussed in detail in ch.13 of Self-Organization in Biological Systems, ed. Camazine, et al, 2001). The equations I am trying to replicate are shown below (disregard matrix):

Replicating the simple probability of an ant choosing the left branch vs. the right branch, can a similar approach be applied to fund management, or even sector selection? Instead of using pheromones as the independent variables, could an analogous equation be derived based on momentum in and/or out of long/short positions? Ie., what is the probability of a new participant (fund manager) entering into a position as a function of the velocity/momentum established by predecessors in that sector? Can a signal be extracted that can let the new participant evaluate if their position is supported?

Thoughts/suggestions welcome, aeroculus at gmail dot com

No comments:

Post a Comment


About Me

My photo
geek, scientist, data farmer