Value‐at‐risk
Authors:
- 9 November 2010
- journal article
- Published by Emerald Publishing in The Journal of Risk Finance
- Vol. 11 (5) , 464-480
Abstract
Purpose – The last few years in the financial markets have shown great instability and high volatility. In order to capture the amount of risk a financial firm takes on in a single trading day, risk managers use a technology known as value-at-risk (VaR). There are many methodologies available to calculate VaR, and each has its limitations. Many past methods have included a normality assumption, which can often produce misleading figures as most financial returns are characterized by skewness (asymmetry) and leptokurtosis (fat-tails). The purpose of this paper is to provide an overview of VaR and describe some of the most recent computational approaches. Design/methodology/approach – This paper compares the Student-t, autoregressive conditional heteroskedastic (ARCH) family of models, and extreme value theory (EVT) as a means of capturing the fat-tailed nature of a returns distribution. Findings – Recent research has utilized the third and fourth moments to estimate the shape index parameter of the tail. Other approaches, such as extreme value theory, focus on the extreme values to calculate the tail ends of a distribution. By highlighting benefits and limitations of the Student-t, autoregressive conditional heteroskedastic (ARCH) family of models, and the extreme value theory, one can see that there is no one particular model that is best for computing VaR (although all of the models have proven to capture the fat-tailed nature better than a normal distribution). Originality/value – This paper details the basic advantages, disadvantages, and mathematics of current parametric methodologies used to assess value-at-risk (VaR), since accurate VaR measures reduce a firm's capital requirement and reassure creditors and investors of the firm's risk level.Keywords
This publication has 25 references indexed in Scilit:
- Forecasting one‐day‐ahead VaR and intra‐day realized volatility in the Athens Stock Exchange MarketManagerial Finance, 2008
- A calibrated scenario generation model for heavy-tailed risk factorsIMA Journal of Management Mathematics, 2006
- An Application of Extreme Value Theory for Measuring Financial RiskComputational Economics, 2006
- VaR stress tests for highly non‐linear portfoliosThe Journal of Risk Finance, 2005
- Forecasts from biased experts: a “meta‐credibility” problemThe Journal of Risk Finance, 2005
- RELIABILITY AT RISK: The supervision of financial models as a case study for reflexive economic sociologyEuropean Societies, 2001
- From value at risk to stress testing: The extreme value approachJournal of Banking & Finance, 2000
- Extreme values from skewed distributions [mathematics education]IEEE Transactions on Education, 2000
- Evaluation of Value-at-Risk Models Using Historical DataSSRN Electronic Journal, 1996
- Generalized autoregressive conditional heteroskedasticityJournal of Econometrics, 1986