![IconGems-Print](images/IconGems-Print.png)
519 – SPEED: Methodological Advances in Time Series: BandE Speed Session, Part 2
Application of Linear and Non-Linear Models Into Trend Analysis of U.S. Cotton Export (1996–2017)
Marguerite Moore
North Carolina State University
Lori Rothenberg
North Carolina State University
Zahra Saki
North Carolina State University
Cotton fiber is the largest source of U.S. textile and apparel export advantage. Rises and falls in cotton export advantage over the examination period (1996-2017) embolden research into past trend analysis of cotton fiber. UN-COMTRADE database is used to extract export values of products including HS5201, Cotton; not carded or combed, to calculate the Normalized Revealed Comparative Advantage (NRCA) index. Simple linear regression (SLR), modified SLR following outlier removal and a non-linear model (i.e., cyclical) are used to fit the data. A cyclical model suggesting a repeated period of seven years best describes variation of U.S. cotton export competitiveness.