Page 248 - Contributed Paper Session (CPS) - Volume 1

CPS1297 Chin W.C. et al.
dependence volatility (Cheong and Lee, 2018; Cheong et al, 2017) in some
empirical financial market studies.
For this specific study, we have selected the Mexican IPC index which acts
as an important indicator to reflect the general and comprehensive
performance of the Mexican Stock Exchange (BMV). In addition, as the largest
stock exchange in Mexico and the fifth stock exchange in America, BMV plays
an irreplaceable role in the financial market. Recently, Horenstein and Snir
(2017), Herrerra, et al. (2015) and Torre et al. (2016) have conducted the
empirical studies regarding the portfolio planning in this area; besides,
Choudhry (1996) and Aggarwal et al. (1999) completed relative researches
focusing on the AR-GARCH models. To the authors’ information, practical
studies about this topic are limited, especially for the highfrequency data of
the IPC index.
In our analysis, we use two high-frequency volatility estimators namely the
realized volatility (RV) and bipower variation volatility (BV), to re-examine the
HMH in the Mexican stock market. Using the Heterogeneous Autoregressive
Model (Corsi, 2009) with enhancement of asymmetric ARCH feature, the
Mexican Indice de Precios y Cotizaciones (IPC) index is modeled and estimated
using the 5-minute data. After evaluating the best forecast model for volatility,
we further examine the performances for the individual and average combined
forecasts which will be further used in determining the market risk. Volatility
usually connects with determining the market risk for investment decision. For
the application in finance, the value-at-risk is determined based on the
estimation results.
The remaining of this study is arranged as follows: Section 2 explains the
formation of high-frequency RV and BV HAR models. Section 3 discusses the
value-at-risk determination; Finally, Section 4 summarizes and concludes this
research.

2. Research Methodology
The high-frequency Heterogeneous AutoRegressive (HAR) volatility
models are based on the heterogeneous market hypothesis concepts. In this
study, we use the HAR model with the improvement of asymmetric
autoregressive conditional heteroskedastic (ARCH) impact. The specifications
for HAR(RV)-TGARCH and HAR(BV)-TGARCH models are formulated as
follows:
2,ℎ
2,
2,
2,
ln( , ) = + , ln( ,−1 ) + , ln( ,−1 ) + , ln( ,−1 ) + ,

2, 2, 2, 2,ℎ
ln( ) = + ln( ) + ln( ) + ln( ) +
, , ,−1 , ,−1 , ,−1 ,

where . follows a TGARCH model in the realized volatility (Corsi et
,
al., 2008) and each of the HAR volatility components can be computed
237 | I S I W S C 2 0 1 9