Synergizing ARDL and LSTM methods for enhanced crude palm oil price forecasting in Malaysia / Mohd Shahrin Bahar ... [et al.]
Understanding the volatile nature of palm oil prices is essential due to its profound economic and market implications. The complexity of forecasting palm oil prices stems from the interplay of various demand and supply forces, making it challenging for scholars to pinpoint the key determinants. Thi...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | en |
| Published: |
Universiti Teknologi MARA Selangor
2024
|
| Subjects: | |
| Online Access: | https://ir.uitm.edu.my/id/eprint/107934/1/107934.pdf https://ir.uitm.edu.my/id/eprint/107934/ https://journal.uitm.edu.my/ojs/index.php/JEEIR/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Understanding the volatile nature of palm oil prices is essential due to its profound economic and market implications. The complexity of forecasting palm oil prices stems from the interplay of various demand and supply forces, making it challenging for scholars to pinpoint the key determinants. This study addresses the intricate challenges in the palm oil industry, including price volatility, shifting consumer preferences, and environmental sustainability, by analysing the factors influencing Malaysian Crude Palm Oil (CPO) pricing dynamics. Utilising data from the Malaysian Palm Oil Board, covering the period from January 2004 to December 2021, we examined the impact of these variables on CPO prices. Methodologically, we employed Autoregressive Distributed Lag (ARDL) and Long Short-Term Memory (LSTM) models to evaluate and forecast CPO prices. Our findings indicate that the LSTM method outperformed the ARDL method in terms of forecasting accuracy. Specifically, the LSTM model showed superior performance when using a selection of ten independent variables identified through LASSO and SHAP estimation, compared to using eleven or four variables based on ARDL regression results. The analysis underscores the significant influence of weather conditions and macroeconomic factors, particularly tax rates, on CPO prices. These findings contribute to a deeper understanding of market dynamics and enhance the accuracy of CPO price forecasting. |
|---|