Estimating Conditional VaR Using Symmetric and Non-Symmetric Autoregressive Models in Old and Oil Markets

Subject Areas : Financial Knowledge of Securities Analysis

Saeid Fallahpour ¹ , Fatemeh Rezvani ² , Mohammadreza Rahimi ³

1 - استادیار مدیریت مالی، دانشکده مدیریت دانشگاه تهران
2 - کارشناس ارشد حسابداری، دانشکده مدیریت دانشگاه تهران
3 - دانشجوی دکتری مدیریت مالی، دانشکده مدیریت دانشگاه تهران

Received: 2014-06-07 Accepted : 2014-07-16 Published : 2015-06-22

Keywords: Value at risk, Conditional Value at Risk, Autoregressive Conditional&#x0D, &#x0D, H, Coherent risk measure, backtesting,

Abstract :

Price volatility on gold and oil market is the top news all the time. Global economy isaffected by those markets volatility. Because of the domestic investor tendency in goldmarket and feasibility of investing on oil by Energy Exchange in Iran, in this paper wefocus on the volatility of gold and oil return. The implemented method is one-day aheadout of sample forecast by the conditional value at risk.The goal of this paper is to answer which of the models; GARCH, ECHARCH, andTARCH is best at forecasting the CVaR for gold and oil return. We estimate the value byassuming normal and t-student distribution.The results show that the TGARH(1,1) model specifications are good option forforecasting the CVaR in oil market by t-student distribution