ANALYZE THE TIME SERIES MODELS FOR THE FORECAST OF THE VALUE OF MUTUAL FUNDS IN ACCORDANCE WITH THE INVESTMENT POLICY

Abstract

The objective is to analyse time-series model in order to find the best model which is accurate and appropriate model for forecasting value of Thanachart Cash Management Fund (T-CASH), Thanachart Thi Ra Sombat Open-End Fund (T-TSB), Bualuang Top-Ten Fund(BTP), Bualuang Global Health Care Fund(BCARE), TMB Gold Fund(TMBGOLD). The datas are used as monthly by using ARMA, ARIMA, ARCH, GARCH, GARCH-M, T-GARCH, EGARCH, IGARCH’s concepts. The best model’s results that is most accurate from comparing the real volatilities in five different times by using MAPE statistic have answers as following below.

The most suitable models are the T-CASH based on ar(1) ar(3) ar(9)  ar(13) d(2) ma(2) and GARCH(2,1) concept, T-TSB based on ar(3) ar(11) d(1) ma(1) ma(3) and EGARCH-M(1,3) concept, BTP based on ar(1) ar(3) d(1) ma(3) and GARCH(1,2) concept, BCARE based on ar(1) ar(5) d(1) ma(1) ma(3) and TGARCH(2,1) concept and TMBGOLD based on ar(1) ar(6) ar(21) d(1) and ARCH(4) concept.