Please use this identifier to cite or link to this item:
Title: Evaluation of value-at-risk models using historical data in Asia.
Authors: Lin, Xiuwen.
Li, Xiang.
Tse, Kit Lam.
Keywords: DRNTU::Business::Operations management::Risk management
Issue Date: 2007
Abstract: The concern of the study is the performance assessment of Value-at-Risk (VaR) models when applied to emerging markets in Asia. The VaR models are assessed on their performances under different parameter settings and under different market conditions. The models investigated include Historical Simulation (HS) and the different volatility models – Simple Moving Average (SMA) and Exponentially Weighted Moving Average (EWMA) under Variance-Covariance (VC). The evaluation is implemented through the use of backtesting, introduced by The Basel Committee on a Banking Supervision in the framework of Basel II. Several potential pitfalls associated with each methodology are highlighted. The results from our empirical study propose that from the regulatory point of view, 99% confidence level is more desired to suffice risk coverage. Small window length is concluded as most favorable when differing weights are not assigned to historical data. Our findings also suggest that HS is the most accurate VaR model which provides adequately for the intended risk coverage. During adverse market conditions, HS may be supplemented with EWMA for better market risk measurement. However, further exploration for the most appropriate model under such conditions should be carried out. Overall, the results presented have important implications for risk managers and market regulators.
Rights: Nanyang Technological University
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:NBS Student Reports (FYP/IA/PA/PI)

Files in This Item:
File Description SizeFormat 
  Restricted Access
2.39 MBAdobe PDFView/Open

Page view(s) 50

Updated on May 14, 2021


Updated on May 14, 2021

Google ScholarTM


Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.