Matthew Flynn, PhD, Director and Head of Machine Learning, Aetna
This presentation walks through retrieving current interest rate data from FederalReserve.gov, then explores the dramatic drop in short-term rates after the 2009 recession. We use JMP scripting to automate the process of retrieving publically available data history. Next we explore and transform the data in JMP to generate interactive visualization and analysis reports. Next we identify common factors within the yield curve data via Principal Components to identify parallel shifts, twists and butterfly effects. We then model the yield curve data via the Time Series and Nonlinear platforms to estimate polynomial and Nelson-Siegel-Svensson functions for forecast interest rates. In short, we demonstrate that JMP can be an attractive and powerful analysis platform for yield curve modeling.