Modeling cost behavior: Linear models for cost stickiness
Ramon V. Del Rosario College of Business
Academy of Accounting and Financial Studies Journal
Literature acknowledges that costs might not be linear and proportional with activity levels. However, conjectures about the sticky behavior of costs are largely based on anecdotal and empirical evidence despite sufficiently advanced economic theory that explains cost behavior (Cooper and Kaplan, 1998; Noreen and Soderstrom, 1997; Banker and Johnston, 1993). For instance, while Noreen and Soderstrom (1997) find no evidence of stickiness, Anderson, et al (2003) find that SG&A costs are sticky that is, they increase, on the average, by 0.55% per 1% increase in revenues, but decline by 0.35% per 1% decrease in revenues. Subramaniam and Weidenmier (2003) confirm cost stickiness, finding that total cost increase 0.93% per 1% increase in revenues but decrease only by 0.85% per 1% decrease in revenues. Both studies used data from US firms. This paper derived a basic cost behavior model and used this model to test whether asymmetric cost behavior in Philippine firms is also prevalent, using different linear models such as OLS and GLS regression analyses. It concluded that GLS regression analysis is not more efficient than OLS regression analysis.
Uy, A. O. (2011). Modeling cost behavior: Linear models for cost stickiness. Academy of Accounting and Financial Studies Journal, 15 (SUPPL.1), 25-34. Retrieved from https://animorepository.dlsu.edu.ph/faculty_research/162