Sale Forecasting of Merck Pharma Company using ARMA Model

Abstract

This study aims to develop a stochastic framework of model to forecast future sales for pharmaceutical industry. In this regard, the study focuses on Merck Pharmaceutical monthly sales data. This study examines the Sale forecasting models. The study includes monthly data published in the annual reports of the company from Jan. 2008 to Dec. 2012.The time series diagram shows unequal means over the time period that suggests the data is stationary. Having transformed the data, ARMA (1, 1) model is applied which shows that there will be increase in sales by $6.784m given that in the last month sales were $1bn. On the contrary, last month’s residual has an adverse effect on current month sales up to the extent of $432.942m. In this study AR (1) and MA (1) both the processes are significant at 1%

Keywords: Sales forecast, ARMA (1, 1), Pharma Industry