방법 비교

	GARCH-MIDAS	컴포넌트 GARCH	Unrestricted MIDAS Regression
분야	계량경제학	계량경제학	계량경제학
계열	Regression model	Regression model	Regression model
기원 연도	2012	1999	2007
창시자	Engle and Ghysels	Engle and Lee	Eric Ghysels
유형	Time-varying variance model	Decomposed variance model	Time-series regression
원전	Engle, R. F., & Ghysels, E. (2012). GARCH for long memory. Journal of Econometrics, 164(2), 385-391.	Engle, R. F., & Lee, G. (1999). A permanent and transitory component model of stock return volatility. Journal of Political Economy, 107(6), 1363-1384.	Foroni, C., Ghysels, E., & Marcellino, M. (2015). Mixed-frequency vector autoregressive models. International Journal of Forecasting, 31(4), 1051-1070.
별칭	Mixed-frequency volatility model	Volatility components model	Unrestricted Mixed Data Sampling
관련	3	3	3
요약	GARCH-MIDAS decomposes volatility into short-term (GARCH) and long-term (MIDAS) components, allowing low-frequency macroeconomic variables to drive medium-term volatility while high-frequency returns govern daily fluctuations. Introduced by Engle and Ghysels (2012), this framework elegantly separates volatility time scales. The approach is powerful for understanding how macro conditions (growth, inflation) drive risk premia and for improved volatility forecasting.	Component GARCH decomposes conditional variance into transitory (short-term) and permanent (long-term) components with different dynamics, allowing flexibility in capturing volatility behavior at multiple frequencies. Introduced by Engle and Lee (1999), it elegantly models the empirical finding that volatility exhibits both rapid mean-reversion (daily shocks) and slow mean-reversion (level shifts). This framework is crucial for understanding volatility persistence and improving long-horizon volatility forecasting.	U-MIDAS (Unrestricted MIDAS) is a regression framework designed to handle mixed-frequency data—when explanatory variables arrive at different sampling frequencies (e.g., monthly GDP mixed with daily stock returns). Introduced by Ghysels and colleagues (2007), it eliminates the restrictive lag-structure polynomial constraints of the original MIDAS approach, allowing fuller use of high-frequency information. This flexibility makes it ideal for nowcasting and real-time economic forecasting.