AI 용어집
인공지능 완전 사전
ARCH Model
Autoregressive Conditional Heteroskedasticity model introduced by Engle (1982) where conditional variance is a linear function of past squared errors.
GARCH Model
Generalized Autoregressive Conditional Heteroskedasticity model developed by Bollerslev (1986) extending ARCH by including past conditional variances in the variance equation.
Leverage Effect
Phenomenon where negative shocks to returns increase future volatility more than positive shocks of the same magnitude, captured by EGARCH or TGARCH models.
Volatility Persistence
Measure of how quickly volatility shocks dissipate, determined by the sum of ARCH and GARCH parameters in a GARCH(p,q) model.
EGARCH Model
Exponential GARCH model by Nelson (1991) that captures the leverage effect and ensures positive variance by modeling the logarithm of conditional variance.
GJR-GARCH Model
GARCH model by Glosten, Jagannathan and Runkle (1993) that incorporates an asymmetric term to differently model the impact of positive and negative shocks on volatility.
IGARCH Model
Integrated GARCH model where the sum of parameters equals 1, implying infinite persistence of volatility shocks and a unit root in the variance process.
APARCH Model
Asymmetric Power GARCH model by Ding, Granger and Engle (1993) that generalizes several GARCH models by including a power parameter and a leverage effect.
FIGARCH Model
Fractionally Integrated GARCH model that captures long memory in volatility by allowing hyperbolic rather than exponential persistence.
Multivariate GARCH Model
Extension of univariate GARCH models to simultaneously model conditional variances and covariances of multiple time series.
Variance Forecasting
Main application of GARCH models involving projecting future conditional variance over different time horizons for risk management.
CGARCH Model
Component GARCH model that separates volatility into a transitory component and a long-term component, allowing for better modeling of persistence.
HARCH Model
Heterogeneous GARCH model that includes volatility terms at different time horizons to capture long memory effects in financial markets.