An EA for portfolio selection over multiple investment periods with exponential transaction costs

Authors:
Matthew J. Craven;Henri C. Jimbo
Affiliations:
University of Abertay Dundee, Dundee, United Kingdom;American University of the Middle East, Dasman, Kuwait
Venue:
Proceedings of the 15th annual conference companion on Genetic and evolutionary computation
Year:
2013

Citing 4
Cited 0

Heuristics for cardinality constrained portfolio optimisation

Computers and Operations Research
An Evolutionary Approach to Multiperiod Asset Allocation

Proceedings of the European Conference on Genetic Programming
Steepest ascent hill climbing for portfolio selection

EvoApplications'12 Proceedings of the 2012t European conference on Applications of Evolutionary Computation
A comparative study of multi-objective evolutionary algorithms to optimize the selection of investment portfolios with cardinality constraints

EvoApplications'12 Proceedings of the 2012t European conference on Applications of Evolutionary Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

We study a matrix representation for an EA attack on the CCPOP with transaction costs. The representation is based on portfolio sequences which change over the investment lifetime in response to asset price changes. We show the approach is effective and that EA performance is directly related to asset price correlation. We compare the EA with a matrix hillclimber and show some common results of vector representations do not hold for a matrix one, potentially providing a step forward in performance of such algorithms.