ABSTRACT:
This study aimed to price croplands in Rio Grande do Sul State (southern Brazil) and point which variables had the most significant impact on prices. The main purpose was achieved using multiple linear regression and principal component analysis. The variables used in this study were planted area, production, price, and yield of the commodities soybean, wheat, and corn. The period under analysis was from January 1994 to December 2017 (biannual observations). Multiple linear regression showed that five variables contributed to land pricing, being three related to soybean and two to wheat. Multivariate analysis grouped the investigated variables into clusters and indicated their influence, in addition to providing information on land prices and reducing variable dimensionality from fourteen original variables to three principal components to be analyzed. The two analyses complemented each other so that the croplands’ price was explained by three variables, in which two corroborated in constructing the pricing model for croplands.
Key words:
linear regression; principal component analysis; cropland pricing.
Thumbnail
Thumbnail
