|
|
Production
Estimates and Crop Assessment Division |
|
|
Production
Estimates and Crop Assessment Division |
August 13, 2003
Since January, actual Malaysian palm oil output has exceeded the levels projected by the Malaysia Rainfall Regression Model (MRRM). The higher-than-expected levels are attributed to increased input levels and improving oil palm selections, which take advantage of more intensive management. Prices for crude palm oil increased by 67 percent from October 2001 to June 2002, which induced plantations to increase input levels.
 |
Malaysian monthly average rainfall was 148 millimeters in the second quarter of 2003, 52 millimeters below normal, and has now been below normal for six out of the last ten quarters (see graph below). The below-normal rainfall in the second quarter may have hurt pollination and could negatively affect yields in the first three months of 2004.
On the other hand, very good rainfall in the last two quarters of 2002 and the first quarter of 2003 are benefiting the crop now and will continue to do so through the remainder of 2003. The level of rainfall in the fourth quarter of 2002 (263 millimeters) was below normal, but will be favorable to production. However, below-normal rainfall from the second quarter of 2001 through the second quarter of 2002 may negatively impact palm oil output from the middle of 2003 into late 2004 due to its affect on floral sex selection and increased floral abortion.
 |
Using the Malaysia rainfall data, yield for July-September is estimated by the MRRM at 1.00 tons per hectare, slightly below the 5-year average of 1.02 tons per hectare. This would give a production level of 12.6 million tons for the 2002/03 marketing year (Oct.-Sept.), less than the 13.0 tons being forecast by USDA. The MRRM model is projecting a yield of 3.65 tons per hectare for 2003/04. This projection would give a production level of 12.2 million tons (using a projected harvested area of 3.35 million hectares), less than the 13.4 million tons currently forecast by USDA. The USDA is giving credit to increased fertilizer applications and the use of improved selections of oil palm for the higher yields that have been seen recently, which have been above what is expected from simply relying on rainfall data and a long term yield trend.
Note: The MRRM linear regression model is developed by the United States Department of Agriculture. It uses rainfall lagged 3 quarters, 1 year cumulative rainfall lagged 6 quarters, and time as independent variables regressed against yield. Because of the characteristics of linear regression, the model continues to increase projected yields even as rainfall becomes excessive. The model tends to overstate yield when average monthly rainfall lagged 3 quarters is above 300 mm, and understate yield when rainfall is optimal.
