Depreciation in the U.S. Used Tractor Market: The Roles of Brand and Technology
Abstract
This paper examines tractor depreciation through examining physical attributes, non-physical attributes, and sales venue. The goal of this thesis is to examine the effects of recent technologies on tractor depreciation patterns and update the literature regarding branding and depreciation. A significant wave of mergers and acquisitions lasting up to the early 2000’s has impacted market structure since most literature on this topic has been written. The data for this thesis was obtained from a database which has tracked agricultural auction sales since 1996 and provides a unique opportunity to observe attributes not frequently recorded in sale transactions. The analysis begins by selecting an appropriate functional form and then delves into understanding the value of various attributes and sales venue on tractor price. The primary modeling conclusion obtained in this study is that the ability for horsepower to enter the model flexibly is very important to developing a model that predicts well out of sample. This thesis supports the conclusions found by Wu and Perry (2004) that the Box-Cox and double square root models appear to function well in out-of-sample prediction. However, a flexible polynomial model provides a very similar fit for most of the horsepower range, and is probably a good option as well. Larger tractors appear to depreciate more quickly than their smaller counterparts do, possibly due to higher downtime costs. John Deere and Case IH appear to depreciate at very similar rates in relative terms in this study. Further, sales venue (online or in-person) did not seem particularly important to sale price. The value of warranties were small, indicating that information asymmetry may not impact price in an important way. A potential approach to modeling guidance system values is also discussed.
Degree
M.S.
Advisors
Gunderson, Purdue University.
Subject Area
Agriculture|Agricultural economics
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.