Research Reports
Research on Improving Accuracy of Livestock Outlook program and Constructing Linkage Program between Livestock Outlook and Policy
-
AuthorCho, Jaesung
-
Publication Date2014.03.30
-
Original
Life cycle of Hanwoo (native Korean cattle) is longer than other livestock. In general, Hanwoo is slaughtered at the age of more than two years. Therefore, a policy for market stabilization should be implemented at least two years before that an exceed-supply or -demand is anticipated. Currently, this is difficult to implement a policy prior to detecting market problems due to the lack of future market information. Therefore, this study aims to build a model for forecasting Hanwoo market for the total number of raised Hanwoo, the number of slaughtered Hanwoo for each sex, and of raised Hanwoo in terms of each sex and age, the wholesale price of Hanwoo for each sex, the farm prices of calves 6~7 months old for each sex, the farm prices of adult cattle (600 kg) for each sex.
We identify the structure of beef market and the factors affecting the market supply and demand based on the previous research and the opinions from experts in the various fields related to Hanwoo and beef market. Equations composing the forecasting model are estimated by Excel and statistical softwares such as STATA and EViews. Forecasting error of each equation is examined by the measure of Mean Absolute Percentage Error (MAPE). The estimated results are reviewed by experts in this field.
The model is composed of a supply part, a demand part, and a price part. The model for the supply part is constructed based on the ecological equations which reflect livestock raising periods and features of production thereof. The model for the demand part is constructed based on the factors including seasonal consumption that affect beef consumption. The model for the price part is composed of equilibrium price (wholesale price) and other price equations. The equilibrium price is derived from the market clearing condition and other prices are estimated using a statistical software.
Individual equations which are components of the supply part of the model are estimated in the manner to minimize the sum of the absolute deviations (least absolute deviation) by using Excel’s solver. Other equations are estimated using a statistical software. MAPE is then used as the measure of verifying the prediction accuracy of the individual equations.
The results of prediction accuracy verification reveal high prediction capability of the estimated equations. In particular, the average MAPE of the estimated equation for the number of raised Hanwoo is only 1.12%. The average MAPE of the other equations estimating the number of animals is also very low (1.17%~2.42%), with the exception of the estimated equation for the number of raised bulls older than 2 years old (the average MAPE: 4.94%).
The average MAPE of the estimated equations for the number of slaughtered Hanwoo for each sex are 6.04% and 4.93% for female and male, respectively, which are a relatively low prediction accuracy in comparison with the estimated equations for the number of raised Hanwoo for each sex and age group. It is because the number of slaughtered Hanwoo is affected by various external factors including farmer’s profits, expected returns in the future, and government’s policy for beef cattle market. On the other hand, the number of raised animals is mainly determined by means of a standardized management system. The average MAPE of the price equations is also very low (3.34%~6.20%) which implies that the estimated equations have high prediction accuracy.
Researchers: Jae-Sung Cho, Woo-Jin Song, Yong-Geon Lee, Jung-hyun Yoon
Research period: 2013. 8. ~ 2014. 3.
E-mail address: kor0025cho@krei.re.kr
- Next
- Survey on the Korean Women Farmers, 2013
- Prev
- Study on the design for a livestock marketing survey