PERAMALAN HARGA SAHAM DENGAN METODE EXPONENTIAL SMOOTH TRANSITION AUTOREGRESSIVE (ESTAR) (Studi Kasus pada Harga Saham Mingguan PT United Tractors)

The stock price data series of PT United Tractors in the period of December 1th 2008 to December 29th 2014 is fluctuative. To model data nonlinear time series one method that can be used is Smooth Transition Autoregressive (STAR), if the function of an exponential transition then a method that can be used is Exponential Smooth Transition Autoregressive (ESTAR). In modelling ESTAR determined transition variable (s_t) of transition function G(s_t,γ,c). Of the research result obtained model ESTAR (1,1). With significance level of 5% obtainedthe value of the stock price data for pt united tractors in the next four to the original.It was also strengthened by the percentage of the mean absolute mape error 0,768233 % ) are relatively small. Keywords : Autoregressive,time series, nonlinearity, ESTAR, MAP