Academia.eduAcademia.edu

IJMSS10Nov5784.docx

point out that a volatility model must have the forecasting ability, this is the central requirement. They explore the stylized factors of volatility and observe the ability of GARCH type models to capture those features. In this paper, we aim to evaluate the ability of GARCH type models to capture the stylized factors of Dhaka Stock Exchange (DSE) returns volatility. We consider the sample period from 27 th November 2001 to 31 st July 2013 for DSE general index and estimate GARCH type models. We made a comparative of different GARCH type models for capturing the stylized factors of the stock index return's volatility.

How efficient the GARCH type volatility models are? Evidence from Dhaka Stock Index Dr. Md. Ashraful Islam Khan Associate Professor, Department of Population Science and Human Resource Development, Rajshahi University Md. Dalowar Hossain Post Graduate Researcher Department of Population Science and Human Resource Development, Rajshahi University Abstract Engle and Patton (2000) point out that a volatility model must have the forecasting ability, this is the central requirement. They explore the stylized factors of volatility and observe the ability of GARCH type models to capture those features. In this paper, we aim to evaluate the ability of GARCH type models to capture the stylized factors of Dhaka Stock Exchange (DSE) returns volatility. We consider the sample period from 27th November 2001 to 31st July 2013 for DSE general index and estimate GARCH type models. We made a comparative of different GARCH type models for capturing the stylized factors of the stock index return’s volatility. Keywords: Volatility, GARCH type Models, DSE general Index. Introduction As Engle and Patton (2000) mention, volatility models must have the ability to forecast future volatility. At the same time, volatility models must have the capacity to capture stylized factors of volatility series. To capture the stylized factors, theoretical researchers are engaged in developing different types of time varying volatility models immediately after the development of Engle(1982)’s ARCH model. To capture different stylized factors exhibited by the volatility series, ARCH type models have been extended in various dimensions including Bollerslev (1986)’s GARCH, to capture the important nonlinearily, asymmetry, and long memory properties in the volatility process (see, e.g., Andersen and Bollerslev, 1998). Other popular extensions of GARCH type models to improve the flexibility of the basic ARCH model are EGARCH (Nelson, 1991) model, GJR-GARCH (Glosten, Jaganathan, and Runkle, 1993), AGARCH (Engle, 1990), APARCH (Ding et al., 1993), TGARCH (Zakoian, 1994) and QGARCH (Sentana, 1995). However, there are few other models including MCMC (see, e.g., Verhofen, 2005 ;Chib, Nardari and Shephard, 2002), support vector machine (see, e.g., Khan, 2011) are also available in literature to capture such stylized factors. We confined our study on GARCH type models only. The rest of the paper is arranged as follows. A brief review of literatures will be in section 2. Section 3 contains the data description along with summary statistics. Section 4 describes computing models. Section 5 discusses the empirical results and section 6 conclusion. Literature Review Dhaka stock market is not yet a well established yet. Still the numbers of quality researches on DSE index is limited in the literature. Very limited numbers of scholars tried to modeling and forecasting DSE index return series but there observed plenty research gaps in their works. Basher et al. (2007), which is one of the best ever works on Bangladesh Stock Market, investigate the issue of market efficiency and time-varying risk-return relationship by using ARMA(p,q), GARCH(1,1)-M models. They consider Gaussian error distribution while the returns series is non Gaussian according to their findings. Chowdhury (1994) observes stock return behavior by using EGARCH-M model considering GED distribution. Chowdhury et al. (2006) investigate how predicted macroeconomic volatility is related to the predicted stock market volatility in Bangladesh by considering only the GARCH (1,1) model to their study. Molla (2009) uses GARCH (1,1), GARCH (2,1) and GARCH (2,2) models to investigate the time varying risk return relationship and persistence of shocks to volatility in Bangladesh stock market and observe that positive skewness and excess kurtosis reveal the non normality of the DSE return series though they do not mention the underlying distribution(s) the considered. Other studies include Hossain and Uddin (2011), Rayhan et al. (2011), Islam et al. (2012), Alam et al. (2013), Muktadir-al-Mukit (2013) and Islam et al. (2014) can be mentioned here. While thinking about the above mention literatures, there observed significant divergence among model selection, underlying innovation distribution selection. Therefore, this study aims to fit the GARCH type models for Dhaka Stock Indices by considering all possible innovation distributions. R program and E-views software will be used for both estimation and forecasting purpose. The GARCH type models The general form of the ARCH type models are (1) (2) (3) (4) where and are functions of , the information set at time and depend on an unknown vector parameters , is a i.i.d. process, independent of with and . and are the conditional mean and conditional variance respectively. The extensions of ARCH type models are: GARCH Model Bollerslev (1986) proposed the GARCH model. We use GARCH (1,1) models as (5) IJMSS Vol.04 Issue-11, (November, 2016) ISSN: 2321-1784 International Journal in Management and Social Science (Impact Factor- 6.178) A Monthly Double-Blind Peer Reviewed Refereed Open Access International Journal - Included in the International Serial Directories International Journal in Management and Social Science http://www.ijmr.net.in email id- [email protected] Page 2