Overview
Ruey S. Tsay is a statistician and financial econometrician (University of Chicago Graduate School of Business). He is known for time series analysis of nonlinear models, outlier detection, regime change in AR processes, and for influential textbooks. His 1993 paper with McCulloch introduced the RLAR/RVAR random shift framework for Bayesian structural change detection using the Gibbs sampler.
Key Contributions
- RLAR and RVAR models (with McCulloch 1993): Random level-shift and variance-shift AR models with Gibbs sampler inference. Each observation can independently sustain a shift; a probit extension allows shift probability to depend on exogenous variables. Applied to monthly US gasoline price changes. See Structural Break Testing and Gibbs Sampler.
- Outliers and level shifts (1988): "Outliers, Level Shifts, and Variance Changes in Time Series," Journal of Forecasting 7: 1–20 — deterministic shift model and detection procedures that the 1993 paper generalizes to the stochastic RVAR framework.
- Analysis of Financial Time Series (textbook): Widely used textbook covering ARMA, GARCH, nonlinear models, realized volatility, and option pricing.
Related