Date of Graduation

5-2019

Document Type

Thesis

Degree Name

Master of Science in Agricultural Economics (MS)

Degree Level

Graduate

Department

Agricultural Economics and Agribusiness

Advisor/Mentor

Michael P. Popp

Committee Member

Bruce L. Dixon

Second Committee Member

Lawton Lanier Nalley

Keywords

artificial neural networks, cow-calf production, herd size management strategies, price signals

Abstract

This thesis is comprised of two studies examining the effects of price signal based herd size management strategies on profitability of cow-calf operations. Herd size management strategies were evaluated across the previous two cattle cycles, 1990-2014, using a fixed land resource and included a variety of production scenarios. These scenarios varied in terms of stocking rates, fertilizer applications rates, and calving season. Each scenario was also analyzed both with and without weather effects on forage production. Weather effects were simulated using a production index derived from satellite imagery across the observed 25-year period. Three herd size management strategies: i) constant herd size; ii) dollar cost averaging; and iii) price signal-based, anticipatory counter-cyclical expansion/contraction, were evaluated on the basis of net present value of cash operating profits as well as on the basis of risk in terms of range of yearly cash operating profit. This analysis revealed fall calving herds with increased forage production and hay sales through medium fertilizer application in conjunction with a counter-cyclical herd size strategy to be the profit-maximizing management choice regardless of inclusion/exclusion of weather effects or time period. However, a constant herd size strategy was shown to create little regret in terms of net present value of cash operating profit. The second study attempts to rank causal variables that drive the differences in profitability across herd size strategies as well as land use intensities revealed in the first study. Two techniques, linear regression and artificial neural networks (ANNs), were compared and contrasted on the basis of relative variable impact rankings as well as goodness-of-fit. This analysis showed cattle price and head sold to be the largest drivers of profitability across the study period. In addition, fall calving was reinforced as the profit-maximizing decision while optimal choices regarding fertilizer application and stocking rate were not apparent. While ANNs were shown to be superior in terms of goodness-of-fit, linear regression provided coefficients, which allowed for more meaningful examination of tradeoffs between calving seasons, stocking rates, and fertilizer rates.

Share

COinS