An analysis of portfolio VaR: Variance-covariance approach

Abstract

With the growing exposure and linkages of Indian financial markets with the international financial markets, a rational investor (individual or institutional) would opt to reap the benefits of international investment opportunities by constructing a portfolio which would generate good returns with least risk. At the same time, the investor is unaware of the expected degree of return and risk inherent in the portfolio. This requires predicting the market risk of a portfolio using appropriate model. As such, the study attempts to calculate the portfolio market risk of domestic and international hypothetical portfolio using VaR-CoVaR (Variance-Covariance) model. The daily closing prices for a period ranging from 2000 to 2014 of Nifty Spot (NSR), Nifty Future (NFR), INRUSD currency pair Spot (USR) and INRUSD currency pair Future (UFR) are considered for building hypothetical domestic portfolio. The daily closing prices of BRICS nations, US and UK equity market indices from January 2000 to December 2014 have been considered for international portfolio. The investors are classified as risk-averse, risk-neutral and risk-takers. The study concludes that VaR-CoVaR model provides accurate results at 95 percent and 90 percent confidence intervals.