Abstract: This study employs tick-by-tick data to estimate realized volatility (RV), realized skewness (RSK)
and realized kurtosis (RKU) measures of 452 individual firms listed at Pakistan Stock Exchange
(PSX). Using standard HAR model and its extensions, role of realized skewness and realized
kurtosis is examined for predicting realized volatility. Both in-sample and out-of-sample forecasts
strongly support the predictability power of realized kurtosis in one, five and 22 days ahead
forecasts of realized volatility. This research provides a comprehensive empirical evaluation to
guide practitioners and applied researchers discerning the selection of variables, lag criteria and
measurement models to acquire reliable volatility predictions.
Abstract: This study employs tick-by-tick data to estimate realized volatility (RV), realized skewness (RSK) and realized kurtosis (RKU) measures of 452 individual firms listed at Pakistan Stock Exchange (PSX). Using standard HAR model and its extensions, role of realized skewness and realized kurtosis is examined for predicting realized volatility. Both in-sample and out-of-sample forecasts strongly support the predictability power of realized kurtosis in one, five and 22 days ahead forecasts of realized volatility. This research provides a comprehensive empirical evaluation to guide practitioners and applied researchers discerning the selection of variables, lag criteria and measurement models to acquire reliable volatility predictions.
Keywords: HAR model; Emerging market; Predictability power; Realized kurtosis; Realized volatility
JEL codes: G10; G12; G17; O16
DOI: ...