Advanced

Similarity-Based Grouping of a Universe of Securities

Olsen, Fredrik (2013) FMS820 20131
Mathematical Statistics
Abstract (Swedish)
Securities can be classified in multiple ways. One of the most common ways to classify securities is using Morningstar’s label system. For instance a fund can be Large-Value or Small-Growth. However, these systems have inherit limitations that makes it hard to properly maximize diversification in the selection of a portfolio. In this thesis, we present a method
to classify securities that allows flexibility in the number of groups as well as the underlying data on which we classify. The goal is to have a complete system, able to take any universe of securities as input and produce a high-quality classification of that universe of securities.
Please use this url to cite or link to this publication:
author
Olsen, Fredrik
supervisor
organization
course
FMS820 20131
year
type
H2 - Master's Degree (Two Years)
subject
language
English
id
3628866
date added to LUP
2013-04-02 10:37:07
date last changed
2013-04-03 13:33:51
@misc{3628866,
  abstract     = {Securities can be classified in multiple ways. One of the most common ways to classify securities is using Morningstar’s label system. For instance a fund can be Large-Value or Small-Growth. However, these systems have inherit limitations that makes it hard to properly maximize diversification in the selection of a portfolio. In this thesis, we present a method
to classify securities that allows flexibility in the number of groups as well as the underlying data on which we classify. The goal is to have a complete system, able to take any universe of securities as input and produce a high-quality classification of that universe of securities.},
  author       = {Olsen, Fredrik},
  language     = {eng},
  note         = {Student Paper},
  title        = {Similarity-Based Grouping of a Universe of Securities},
  year         = {2013},
}