An Approach to Analysis and Prediction of Portfolio Budget Variances Using JMP®
Prakash Shrivastava, Oakland University, MI
Many businesses are concerned about variations in expenditures against forecast. Understandably, a company’s finance leadership frowns on “spending over budget” but of equal concern is “spending under budget” as the unspent money could have been better utilized elsewhere in the company. Using a class project1 , we review and analyze this typical scenario found in an IT organization. It looks at the problem from Diagnosis and Prescriptive perspectives. In the diagnosis phase, it uses JMP® to (a) explore and visualize patterns in data and (b) use quality control methods for root cause analysis. In the prescriptive phase, it uses Data Mining methods using JMP® to explore possibility of building models to classify/predict variances. This paper attempts to show use of JMP® in developing a draft methodology for potential improvements in quality of forecasts thereby reducing variations in expenditures.