SBIR/STTR Phase II: Next Generation Component Software for Simulation-Based Econometric Estimation

Period of Performance: 01/01/2002 - 12/31/2002


Phase 1 SBIR

Recipient Firm

Insightful Corporation
Seattle, WA 98109
Principal Investigator


This SBIR Phase II research project proposes to develop user-friendly component software for classical econometric estimation and inference based on simulation methods, such as maximum simulated likelihood, method of simulated moments, and efficient method of moments. In the last decade different simulation-based methods have been developed to tackle complex economic/statistical models which cannot be estimated by conventional methods such as Maximum Likelihood Estimation (MLE) and Generalized Method of Moments (GMM). Although these simulation-based estimators have desirable theoretical properties, they have remained as research topics in academia and have not become useful tools for practitioners because of the lack of user friendly software. Building upon the Phase I research and development, Insightful (formerly MathSoft) plans to study two classes of models: mixed logit models for discrete choice analysis which represent cross sectional and panel data problems, and models for term structure of interest rates which represent discrete time and continuous time structural models. Extensive Monte Carlo experiments will be used to explore finite sample properties of various aspects of simulation, estimation and forecasting, with an aim of improving and stabilizing the current algorithms. The user-friendly component software will be developed using both object oriented S-Plus language and the state-of-art JavaBean technology, and it will provide intuitive graphical user interface. The S-Plus functions of the technology proffered by Insightful for econometric estimation and inference will serve the purpose of quickly gaining a broad user base, while the JavaBeans can be used to develop custom applications. The software will help economists and practitioners in other fields such as the financial industry, social sciences, and biotechnology to conduct flexible and extensible model estimation and inference.