Forecasting Analysis of Crude Palm Oil (CPO) Demand using Single Exponential Smoothing, Moving Average and Holt Winter’s Methods at the Palm Oil Plantation Company
DOI:
https://doi.org/10.25124/jrsi.v11i01.733Keywords:
Forecasting, Crude Palm Oil, Single Exponential Smoothing, Moving Average, Holt Winter’sAbstract
PT Steelindo Wahana Perkasa, a part of the KLK GROUP, is a palm oil plantation company located in
East Belitung. They produce olein and stearin through a refinery plant with a capacity of 1000 MT of
Crude Palm Oil (CPO) per day. These products are exported to Asia and Europe, including countries like
Malaysia, Singapore, China, and the Netherlands. Olein is also sold in the Belitung region. Currently, the
company only produces around 170 MT of CPO per day, so they purchase CPO from other companies.
Research is conducted to analyze the future demand for CPO using the Single Exponential Smoothing,
Moving Average, and Holt Winter's methods. The research's goal is to determine the most accurate method
based on the Mean Absolute Percentage Error (MAPE) value. If MAPE is <10%, the method is considered
suitable for forecasting CPO demand. The research results indicate that the most appropriate method is
Holt Winter's, with a MAPE of 8.6% and 4.4%. In period 13, company A will purchase 2,780.08 MT of
CPO, and company B will purchase 4,496.10 MT.