Document Type

Honors Project

Abstract

Asymmetric Correlation is an empirical observation that correlations between security returns are significantly higher during market downturns compared to upturns. Previous literature has focused on examining this asymmetry by way of a rigorous statistical analysis. However, the sources of this phenomenon had not been explored yet.

In this paper, I propose a simply hypothesis: asymmetric correlations can be explained in terms of the securities' underlying fundamental factors. Using the recently proposed Dynamic Conditional Correlations (DCC) class of GARCH models, I estimate the covariance of the fundamental factors and model securities' returns.

I find that fundamental factors can only partially explain the asymmetric correlation. In my final model, I show that despite taking fundamental factors into account, the correlations still exhibit asymmetry. Therefore behavioral factors also play a significant role in causing this phenomenon. Modeling behavioral causes will be the next important step towards understanding asymmetric correlation.

Share

COinS
 
 

© Copyright is owned by author of this document