Variance–Covariance (Delta Normal) Approach of VaR Models: An Example From Istanbul Stock Exchange

Abstract

Many investors desire to know how much money they can lose for example in a day or in a ten days. In this study, variance-covariance approach of the VaR models is introduced to the reader. It estimates maximum potential loss for a given probability and time horizon. It shows money type one loss value. In a calculation process, firstly, portfolios are created. Then, returns distribution is identified. And lastly, VaR values of portfolios are measured. Daily loss is calculated with using 252 days historical data belonging to the year 2015. Stocks are chosen from Istanbul Stock Exchange (BIST 100 Index).  Calculation is made for both 95 % and 99 % confidence level and one day and ten days holding periods.

Keywords: Risk Measurement, VaR, Variance-Covariance approach, correlation, portfolio risk