Reality Check for the Value-at-Risk Estimates of the Energy Commodities
(2014) NEKN01 20142Department of Economics
- Abstract
- The fluctuations of the price in the energy market affect the households, firms and the government intuitions. We can perceive the information of the energy market from daily economic news. The entire society is concerned for the events that affect the energy market and the changing prices of the energy resources. It is thus meaningful and interesting to study the risk of the energy market. This paper provides empirical study for three representative energy resources from the year 1997 to 2013 by using the value-at-risk (VaR) estimation. The representative energy resources are natural gas, (Brent and WTI) crude oil and propane. In order to generate a serious study and consider both calm periods and volatile periods, the sample period is... (More)
- The fluctuations of the price in the energy market affect the households, firms and the government intuitions. We can perceive the information of the energy market from daily economic news. The entire society is concerned for the events that affect the energy market and the changing prices of the energy resources. It is thus meaningful and interesting to study the risk of the energy market. This paper provides empirical study for three representative energy resources from the year 1997 to 2013 by using the value-at-risk (VaR) estimation. The representative energy resources are natural gas, (Brent and WTI) crude oil and propane. In order to generate a serious study and consider both calm periods and volatile periods, the sample period is divided into 12 subsamples by using “rolling window” method. The investigation is designed to select the most adequate VaR estimates by applying three types of non-parametric approaches, namely the standard historical simulation (HS), the historical simulation with ARMA forecasting (HSAF) and the volatility weighted historical simulation (VWHS). In light of my empirical study, value-at-risk estimates at the 95% confidence level (VaR_(95%)) generally perform poorly in explaining the risk of the three representative energy resources, and value-at-risk estimates at the 99% confidence level (VaR_(99%)) are generally capable to explain the risk of the three representative energy resources (except for the financial crisis year 2008). Meanwhile, the results show that the HSAF approach and the VWHS approach perform slightly better than the standard HS approach, and the VaR_(99%) estimates of VWHS approach can explain the risk occurred in natural gas and Brent crude oil for all the subsample periods. More importantly, it seems that VaR_(99%) estimates of student t-distributed asymmetric VWHS models are qualified for both calm and volatile periods for natural gas and Brent crude oil. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/student-papers/record/4699075
- author
- Miao, Dingquan LU
- supervisor
- organization
- course
- NEKN01 20142
- year
- 2014
- type
- H1 - Master's Degree (One Year)
- subject
- keywords
- Energy commodities, Value-at-Risk (VaR), Historical Simulation (HS), Historical simulation with ARMA forecasting (HSAF), Volatility weighted historical simulation (VWHS)
- language
- English
- id
- 4699075
- date added to LUP
- 2014-10-30 13:23:23
- date last changed
- 2014-10-30 13:23:23
@misc{4699075,
abstract = {{The fluctuations of the price in the energy market affect the households, firms and the government intuitions. We can perceive the information of the energy market from daily economic news. The entire society is concerned for the events that affect the energy market and the changing prices of the energy resources. It is thus meaningful and interesting to study the risk of the energy market. This paper provides empirical study for three representative energy resources from the year 1997 to 2013 by using the value-at-risk (VaR) estimation. The representative energy resources are natural gas, (Brent and WTI) crude oil and propane. In order to generate a serious study and consider both calm periods and volatile periods, the sample period is divided into 12 subsamples by using “rolling window” method. The investigation is designed to select the most adequate VaR estimates by applying three types of non-parametric approaches, namely the standard historical simulation (HS), the historical simulation with ARMA forecasting (HSAF) and the volatility weighted historical simulation (VWHS). In light of my empirical study, value-at-risk estimates at the 95% confidence level (VaR_(95%)) generally perform poorly in explaining the risk of the three representative energy resources, and value-at-risk estimates at the 99% confidence level (VaR_(99%)) are generally capable to explain the risk of the three representative energy resources (except for the financial crisis year 2008). Meanwhile, the results show that the HSAF approach and the VWHS approach perform slightly better than the standard HS approach, and the VaR_(99%) estimates of VWHS approach can explain the risk occurred in natural gas and Brent crude oil for all the subsample periods. More importantly, it seems that VaR_(99%) estimates of student t-distributed asymmetric VWHS models are qualified for both calm and volatile periods for natural gas and Brent crude oil.}},
author = {{Miao, Dingquan}},
language = {{eng}},
note = {{Student Paper}},
title = {{Reality Check for the Value-at-Risk Estimates of the Energy Commodities}},
year = {{2014}},
}