| dc.description.abstract | Per capita consumption of yogurt in the United States has increased dramatically over the last 40 years. Yogurt was named one of the most popular foods over the last ten years. This increase in growth partially can be attributed to an increasingly health-conscious population. In order to provide an in-depth analysis of the yogurt industry as a whole by brand, this study uses weekly point-of-sale information from Nielsen for five major yogurt brands over the period January 2009 to December 2011. During this period, the most dominant brands in the yogurt industry were Dannon, Yoplait, Stonyfield, Private Label, and Chobani. Chobani entered the market in 2009 as the first Greek yogurt brand.
Single-equation regression demand models and a seemingly unrelated regression (SUR) demand systems model were constructed to provide own-price, cross-price, and expenditure elasticities of demand for yogurt by brand. These models also identified the impacts of income, the Great Recession, seasonality, and habit formulation for each of the five brands. The drivers of demand in the single-equation models were own-price, habit formulation, tastes and preferences, seasonality, and Christmas. The SUR model takes into account correlation of the error terms as well as the incorporation of the homogeneity restriction on the own-price, cross-price, and income elasticities. Additionally, the SUR model was used to re-examine some of the results from the single-equation models since the latter were at odds with conventional economic theory. Similar to the single-equation models, own-price, income, tastes and preferences, seasonality, and Christmas were found as drivers in the SUR model. Cross-price effects were not found to be a significant driver of demand for the respective yogurt brands in both models. To analyze the performance of the models, metrics such as R-Squared, Adjusted R-Squared, standard error of the regression (SER), and Durbin-Watson (DW) statistics were used. The SUR model performed better for the Chobani, Stonyfield, and Private Label brands, while the single-equation models performed better for the Dannon and Yoplait brands.
Using ex-post forecasting, single-equation and SUR models were evaluated to determine their ability to generate accurate forecasts. The last thirteen weeks of 2011 were withheld to test the forecast performance of the respective models. Metrics such as root mean squared error (RMSE), mean absolute error (MAE), mean absolute percent error (MAPE), bias proportions, variance proportions, and covariance proportions were used to assess the ability of the single-equation and SUR models to generate accurate predictions. Forecast accuracy was better for the Chobani, Yoplait, and Private Label brands when using the SUR model, but forecast accuracy was better for the Dannon and Stonyfield brands when using the single-equation model.
The analysis provides a view of the yogurt industry from 2009 to 2011 in the United States. To date, there is not much research dealing with the demand for yogurt by brand. This examination of the yogurt industry will benefit analysts and manufacturers of these major yogurt brands as well as improve their understanding of the driving forces of demand for yogurt. In particular, the estimated own-price elasticities will help manufacturers maximize their revenues, with appropriate pricing strategies. | en |
