Fractionally Integrated Models for Volatility: A Review
NONLINEAR FINANCIAL ECONOMETRICS: MARKOV SWITCHING MODELS, PERSISTENCE AND NONLINEAR COINTEGRATION, pp. 104-123, Palgrave Macmillan, 2011
Posted: 31 Jan 2011
Date Written: January 30, 2011
Many empirical studies showed the strong degree of persistence of shocks to the conditional variance process. In this case, the distinction between stationary and unit root processes may be too restrictive, since the propagation of shocks occurs at an exponential rate of decay in a stationary process, while the persistence of shocks is infinite for a unit root process. Fractional models have been recently proposed in the financial literature to fill the gap between short and complete persistence. The main motivation to use this kind of models is that the propagation of shocks occurs at a slow hyperbolic rate of decay, as opposed to the exponential decay associated with the stationary and invertible ARMA class of processes, or the infinite persistence resulting from non-stationary processes. We review the main developments of this growing field of research, trying to highlight the advantages and disadvantages of the main approaches proposed so far.
Keywords: Fractional Integrated Models, GARCH, EGARCH, Threshold-GARCH, APARCH, IGARCH, FIGARCH, FIEGARCH, FIAPARCH, HYGARCH
JEL Classification: C22, C51, C52
Suggested Citation: Suggested Citation