Deterministic behavior of stochastic models to examine the volatility in Colombo stocks indices

Abstract

The study “Deterministic behavioral Stock Market Model to Examine the Volatility in Colombo Stock Indices” is done using daily stock market price indices of Colombo Stock Exchange (CSE) from 2000 to 2016. The main objective of the study is to build an appropriate model to estimate market volatility based on All Share Price Index (ASPI) and price indices of the selected sectors. Stationarity and variance patterns of the ASPI are inspected by using descriptive time series plots of the original series, log transformed series and returns series. Box-pierce LM Test and ARCH Effect Test are used to check the existing of volatility clusters in returns series. Further Statistical Tests are applied to identify the asymmetric volatility clusters. Two distinct EGARCH models are built to examine the volatility in ASPI before and after the ending of war which was occurred till May 2009, Sri Lanka. The Diagnostic Checking of the fitted models is done by using Heteroskedasticity Test, Correlogram of the squared residuals. Assumptions of the Error distribution are validated by Q-Q plot. Further, existence of volatility clusters and asymmetric patterns of price indices of Bank Finance & Insurance (BFI), Construction & Engineering (CE) and Manufacturing (MFU) sectors are tested using proper statistical tests. Diverse GARCH family models are used to inspect the variance of sector price indices. Diagnostic checking is performed for each built model and volatility of the sector indices are estimated by appropriate models.