Overview
Richard T. Baillie is an econometrician at Michigan State University specialising in financial time series, long memory processes, and exchange rate dynamics. He is best known for introducing the FIGARCH model (with Bollerslev and Mikkelsen, 1996), which extends IGARCH by replacing the unit-root polynomial with a fractional differencing operator, and for the 1996 Journal of Econometrics survey that established the standard reference framework for long memory and fractional integration in econometrics. Earlier work with Bollerslev documented common stochastic trends and GARCH-in-mean risk premia in exchange rate systems.
Key Contributions / Features
- Long memory survey — Baillie (1996) — Comprehensive review of ARFIMA processes, R/S and GPH estimation methods, and applications of long memory to inflation, interest rates, exchange rates, forward premia, and volatility. See Baillie (1996) and Long Memory and Fractional Integration.
- FIGARCH — Baillie, Bollerslev, and Mikkelsen (1996) — Fractionally Integrated GARCH model: replaces (1−L) in the IGARCH variance equation with (1−L)d, 0<d<1, yielding hyperbolic decay of lagged squared innovations; estimated d≈0.4–0.6 for USD exchange rates. Key bridge between purely GARCH (geometric decay) and IGARCH (unit root). See GARCH and BEKK-GARCH.
- Exchange rate cointegration — Baillie-Bollerslev (1989a) — Documented one cointegrating vector among seven nominal spot exchange rates using the Johansen procedure.
- Long memory in forward premia — Baillie-Bollerslev (1994b) — Found strong evidence for long memory (d^≈0.45–0.80) in forward premia; showed standard unit root vs. stationarity tests have low power to distinguish I(1) from I(d) in exchange rates.
Related