Putri Dwi Pradina, - and Scolastika Mariani, - and Sugiman, - (2014) Pemodelan Arfima Melalui Alfa Hill Stable Sebagai Penentu D Dan Aplikasinya Dalam Estimasi Harga Saham. Unnes Journal of Mathematics, 3 (2). pp. 62-67. ISSN 2460-5859

PDF - Published Version Download (392kB)

Abstract

ARFIMA (Autoregresive Fractionally Integrated Moving Average) model was developed to model the long memory in time series with differencing (d) as a real number, ­0,5<d<0,5. Differencing operation is related to Hurst exponent (H). In the terms of Hurst exponent, there is a term what expressed relation between the fractal dimension of time series and the fractal dimension of probability space , i.e . The fractal dimension of probability space is Levy index of the Stable distribution. This research was aimed to assess ARFIMA(p,d,q) model through to determined and determined ARFIMA(p,d,q) model to estimate a stock price (Gowa Makasar Tourism(Close) Stock) with help software OxMetrics 4.0 which result the smallest MSE value. Thus, ARFIMA(3;0,16398;2) model without constant is a right mode to estimte the data. The estimated model equation of data is i where e .Estimation of the stock price for period 3rd July 2013, 4th July 2013, 5th July 2013, and 8th July 2013 respectively is 6158,20; 4763,00; 4278,60; and 5855,90.

Actions (login required)

View Item