THE MODELING OF RICE AND LOCAL SUGAR PRICES THROUGH TIME SERIES HYBRID MODEL
A staple crop production is one of gauges a countrys prosperity, especially the materials primary commodities such as rice and sugar. The staple price observations based on time form a time series data. Since both commodities prices may be suddenly increased or decreased then those phenomena shou...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/19517 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | A staple crop production is one of gauges a countrys prosperity, especially the
materials primary commodities such as rice and sugar. The staple price observations
based on time form a time series data. Since both commodities prices may be
suddenly increased or decreased then those phenomena should be analyzed by hybrid
time series ARIMA-GARCH model. The average of each rice and sugar prices in
five main traditional market of Bandung, are used as case study. The plots of
data show that both rice and sugar prices are non-stationary processes. Those
processes have normal and spiky behaviors. It makes the necessity to renew the
modeling using hybrid model, which considers the homoscedastic and heteroscedastic
processes. Two considered models are ARIMA for homoscedastic part and GARCH
for heteroscedastic part, whose mean and variance are not constant. It is obtained
that both rice and sugar prices observations have similar pattern, so that they have
the same time series models, which are ARIMA(2,1,1), GARCH(2,1), and ARIMA-
GARCH. |
|---|