A Non-Parametric Approach to Change-Point Detection in Cross-Asset Correlations

<p>In this thesis we explore the problem of detecting change-points in cross-asset correlations using a non-parametric approach. We began by comparing and contrasting several common methods for change-point detection as well as methods for measuring correlation. We finally settle on a statistic introduced in early 2012 by Herold Dehling et.al. and test this statistic against real world financial data. We provide the estimated change-point for this data as well as the asymptotic p-value associated with this statistic. Once this process was complete we went on to use simulated data to measure the accuracy, power, and type 1 error associated with this new statistic. Finally, we were able to draw conclusions on the functionality and usefulness of this statistic.</p>

Name thesis:degree_name

Master of Science in Mathematics (M.S.)

Level thesis:degree_level

Thesis (open access)

Discipline thesis:degree_discipline

Department of Mathematical Sciences

Year dc:date.available

2012

Author dc:creator

• Diamond, L. Kaili

Contributors dc:contributor

• Martha Abell
• Jonathan Duggins

• ETD
• Cross-asset correlation
• Non-parametric
• Change-point
• Diversification
• Mathematics

Repository record dc:identifier

https://digitalcommons.georgiasouthern.edu/etd/16

OAI identifier oai:identifier

oai:digitalcommons.georgiasouthern.edu:etd-1016