Book Synopsis Analysis of Value at Risk Models Based on the Shanghai Stock Index by :
Download or read book Analysis of Value at Risk Models Based on the Shanghai Stock Index written by and published by . This book was released on 2003 with total page pages. Available in PDF, EPUB and Kindle. Book excerpt: Over the last few years, Value at Risk has been universally accepted as a measure of market risk in financial institutions. A lot of research has been done in the field of Value at Risk leading to the development of differing approaches to estimate Value at Risk. However each method has its own set of assumptions and there is very little consensus on the preferred method to estimate Value at Risk. Since all existing methods involve some tradeoff and simplifications, determining the best methodology for estimating Value at Risk becomes an empirical question for implementing the most suitable model. The challenge of this work is to come up with the best and easily implementable approach suitable to Shanghai Stock index data and apply time series models for calculating Value at Risk and compare their performance with current models. Several sketches of current methods are introduced with open issues associated with each method. The study identifies the path for future research to improve the performance of models. The Value at Risk models are evaluated over the two sample periods. The two periods serve to validate the performance of models over time. The best models (EWMA and GARCH) models were reevaluated for the extended forecast sample period and it was found that GARCH models performed consistently over the time. The study makes use of both parametric and non parametric models and also proposes some of the models to estimate Value at Risk. Performance evaluation of the risk metrics, Garch models and historical simulation Value at Risk models are outlined and assumptions tested on Shanghai stock exchange index. The risk metrics and the Garch models incorporate volatility updating as well as clustering phenomenon. It does a poor job in capturing the extreme tail region as compared to historical simulation models. On the other hand historical simulation models capture the tail of the empirical distribution, but are practically insensitive to periods of sudden volatility. Time series models fail to reject the random walk hypothesis and perform poorly in comparison to the current model .Overall the risk metrics model with decay factor of 0.90 performs better than all other models when comparing the accuracy of Value at Risk estimates in first sample period. However over the both forecast sample periods the GARCH models perform consistently. The performance of EWMA marginally deteriorates for the second sample period. It is felt that the conditioning on the past movement of the stock or assert in the previous period will significantly improve the performance of current Value at Risk models. The movements on the positive side should produce less volatility than the movements of equal magnitude on the negative side. This can be taken care of by conditioning of variance of returns on the direction of movement of asset.