James D. Hamilton
James D. Hamilton is an econometrician at the University of California, San Diego, and a Research Associate of the National Bureau of Economic Research. He is best known in this vault for Hamilton 1989, “A New Approach to the Economic Analysis of Nonstationary Time Series and the Business Cycle” (Econometrica 57(2), 357-384) — the paper that introduced autoregressive models with Markov-switching parameters and founded the Markov Regime-Switching Model family. With more than 10,000 citations, that paper supplied both the model class and the recursive-filter estimation method that made it practical, and the family is now commonly called the “Hamilton model”.
Hamilton’s broader work makes him not just the originator of the model but its principal surveyor and methodological referee. His textbook Time Series Analysis (Princeton University Press, 1994) is a standard graduate reference; his Palgrave Dictionary of Economics entry “Regime-Switching Models” (2005) is the concise survey-of-record; and his NBER Working Paper 21863, Macroeconomic Regimes and Regime Shifts (2016, published in the Handbook of Macroeconomics Vol. 2), is the authoritative modern treatment. Across these he documents the lineage explicitly — crediting Markov-switching regressions to Goldfeld and Quandt 1973 and switching regressions to Quandt 1958 1972 — while his own contribution was extending the model to serially dependent (autoregressive) data.
Importantly for this vault’s skeptical brief, Hamilton is also the source of the model family’s central cautions. In WP 21863 he warns that richly parameterised switching specifications “could easily result in an overfitted and misspecified model”, recommends the parsimonious time-invariant Markov chain as the robust default, notes that “there can be multiple local maxima to the likelihood function” so EM estimation must be started from many points, and stresses that testing the number of regimes is non-standard because nuisance parameters become unidentified under the null. He also notes for several specifications that “it’s not clear how to form out-of-sample forecasts”. Hamilton’s entire body of work in this area is econometrics and macroeconomic forecasting — it founds and explains the model but tests no trading strategy and reports no trading profitability.
James D. Hamilton [defines] Hamilton 1989 James D. Hamilton [defines] Macroeconomic Regimes and Regime Shifts Hamilton 1989 [defines] Markov Regime-Switching Model
Connections
- Hamilton 1989 — proposes_model, source: https://www.jstor.org/stable/1912559
- Macroeconomic Regimes and Regime Shifts — proposes_model, source: https://www.nber.org/system/files/working_papers/w21863/w21863.pdf
- Markov Regime-Switching Model — proposes_model, source: https://www.jstor.org/stable/1912559
- Goldfeld and Quandt 1973 — relates, source: https://econweb.ucsd.edu/~jhamilto/palgrav1.pdf
- Overfitting in Quantitative Trading — relates, source: https://www.nber.org/system/files/working_papers/w21863/w21863.pdf
Sources
- Hamilton, J.D. (1989), “A New Approach to the Economic Analysis of Nonstationary Time Series and the Business Cycle”, Econometrica 57(2), 357-384 — https://www.jstor.org/stable/1912559
- Hamilton, J.D. (2016), “Macroeconomic Regimes and Regime Shifts”, NBER Working Paper 21863 — https://www.nber.org/system/files/working_papers/w21863/w21863.pdf
- Hamilton, J.D. (2005), “Regime-Switching Models”, Palgrave Dictionary of Economics — https://econweb.ucsd.edu/~jhamilto/palgrav1.pdf