2015Munich Personal RePEc Archive (Ludwig Maximilian University of Munich)Open access

Identifying Dependence Structure among Equities in Indian Markets using Copulas

Vaibhav Grover

Open full text 5 citations

Abstract

In this study we have examined that assets returns in Indian markets do not follow an elliptical dependence structure; asymmetric tail dependence can be observed among asset returns particularly when the assets exhibit downside returns in a bearish market. We have used Elliptical, Archimedean and Canonical Vine copulas to model such dependence structure in large portfolios. Using certain goodness-of-fit tests we find that Archimedean copulas are insufficient to model the dependence among assets in a large portfolio. We have also compared copula models using an out-of-sample Value-at-Risk (VaR) calculation and comparing results to the historical data. It is observed that the Canonical Vine copulas consistently capture the variation in weekly and daily VaR values.

Open-access reader

About this research paper

What this paper is about

In this study we have examined that assets returns in Indian markets do not follow an elliptical dependence structure; asymmetric tail dependence can be observed among asset returns particularly when the assets exhibit downside returns in a bearish market. We have used Elliptical, Archimedean and Canonical Vine copulas to model such dependence structure in large portfolios. Using certain goodness-of-fit tests we find that Archimedean copulas are insufficient to model the dependence among assets in a large portfolio. We have also compared copula models using an out-of-sample Value-at-Risk (VaR) calculation and comparing results to the historical data. It is observed that the Canonical Vine copulas consistently capture the variation in weekly and daily VaR values.

Why it matters

OpenAlex reports 5 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

In this study we have examined that assets returns in Indian markets do not follow an elliptical dependence structure; asymmetric tail dependence can be observed among asset returns particularly when the assets exhibit downside returns in a bearish market. We have used Elliptical, Archimedean and Canonical Vine copulas to model such dependence structure in large portfolios. Using certain goodness-of-fit tests we find that Archimedean copulas are insufficient to model the dependence among assets in a large portfolio. We have also compared copula models using an out-of-sample Value-at-Risk (VaR) calculation and comparing results to the historical data. It is observed that the Canonical Vine copulas consistently capture the variation in weekly and daily VaR values.

Key concepts: Vine copula, Copula (linguistics), Downside risk, Tail dependence, Econometrics, Portfolio, Economics, Mathematics

Related papers

Back to paper searchBrowse research topicsOriginal source
Identifying Dependence Structure among Equities in Indian Markets using Copulas — Research Paper | ScholarLens