Forecasting Volatility of Chinese Stock Market by Using GARCH Models
Huafei Ren
Supervisor: Zexi Wang
Master Thesis: Msc in International Business
NORGES HANDELSHØYSKOLE
This thesis was written as a part of the Master of Science in Economics and Business Administration program - Major in International Business. Neither the institution, nor the advisor is responsible for the theories and methods used, or the results and conclusions drawn, through the approval of this thesis.
Abstract
This paper aims to forecast the monthly volatility of Chinese stock market by using various GARCH models. We also examine the performance of these GARCH models by comparing the predicted monthly volatility to a proxy of actual monthly volatility calculated using daily data. The evaluation result shows that the GARCH-in-mean model is slightly superior to GARCH model and TGARCH model in term of the ability to deliver “accurate” volatility forecast. But EGARCH model seems a lousy approach in our case. We also construct the upper and lower bound of the predicted volatility for GARCH, TGARCH and GARCH-in-mean models. All of these three models seem to be able to create a reasonable range of the predicted volatility. But GARCH and GARCH-in-mean models are better than TGARCH model in this respect. An interesting by-product is that good news generates greater shocks to the market than bad news, which is quite contrary to the widely accepted view that bad news usually creates more volatility to the market than good news.
Table of Contents
ABSTRACT………………………………………………………… 2
CONTENTS……………………………………………………………... 3
1. INTRODUCTION…………………………………………………….4
2. LITERATURE REVIEW…………………………………………….7
3. DATA DESCRIPTION……………………………………………….10
4. METHODOLOGY……………………………………………………15
4.1 SOME WELL-KNOWN CHARACTERISTICS OF FINANCIAL TIME SERIES……….15
4.2 AUTOREGRESSIVE CONDITIONAL HETEROSKEDASTICITY (ARCH) MODEL…..16
4.3 GENERALIZED AUTOREGRESSIVE CONDITIONAL HETEROSKEDASTICITY
(GARCH) MODEL………………………………………………………………………………18
4.4 THE THRESHOLD GARCH (TGARCH) MODEL………………………………………..22
4.5 GARCH-IN-MEAN MODEL AND TIME VARYING RISK PREMIUM………………….25
4.6 THE EXPONENTIAL GENERAL AUTOREGRESSIVE CONDITIONAL HETEROSKEDASTIC (EGARCH) MODEL………………………………………………….28
4.7 MAXIMUM LIKELIHOOD ESTIMATION…………………………………………………31
5. EVALUATING GARCH MODELS…………………………………..34
5.1 ORDINARY LEAST SQUARES (OLS) METHOD………………………………………….37
5.2 CONSTRUCTION OF UPPER AND LOWER BOUND……………………………………..37
6. CONCLUSIONS……………………………………………………….43
APPENDIX……………………………………………………………….45
REFERENCES…………………………………………………………...54
References
Robert F. Engle (1982). Autoregressive Conditional Heteroscedasticity with Estimates of Variance of United Kingdom Inflation, Econometrica 50: 987-1008.
Tim Bollerslev (1986). Generalized Autoregressive Conditional Heteroskedasticity, Journal of Econometrics, 31: 307-327.
Tim Bollerslev, Wooldridge M. Jeffrey (1992). Quasi-Maximum Likelihood Estimation and Inference in Dynamic Models with Time-varing Covariances, Econometric Reviews 11: 143-172.
Nelson, D. B. (1991). Conditional heteroskedasticity in asset returns: A new approach, Econometrica 59: 347-370.
Sentena, E. (1995). Quadratic ARCH models, The Review of Economic Studies, 639–61.
Glosten, L. R., R. Jagannathan, www.51lunwen.org D. Runkle (1993). On the Relation between the Expected Value and the Volatility of the Nominal Excess Return on Stocks, Journal of Finance, 48: 1779-1801.
Zakoian, J. (1994). Threshold Heteroskedastic Model, Journal of Economic Dynamics and Control 18: 931-955.
Robert F. Engle, Tim Bollerslev (1986). Modelling the persistence of conditional variances. Econometric Reviews, Volume 5, Issue 1: 1 – 50
Chou, R. F. (1988). Volatility Persistence and Stock Valuations: Some Empirical Evidence Using GARCH, Journal of Applied Econometrics, Vol.3: 279-294.
Amit Goyal (2000). Predictability of Stock Return Volatility from GARCH Models, working paper, Anderson Graduate School of Management, UCLA
