GARCH: Long Memory/Short Memory models?
Posted: Wed Mar 03, 2010 8:52 am
Hi there,
I'm writing an assignment on GARCH models and wish to divide them into "leverage-effect models" and "long memory models" respectively. As such, I've put EGARCH, GJR and APARCH in the first category, while I intend to put IGARCH, FIGARCH, FIEGARCH, FIEAPARCH and HYGARCH in the second category.
My question is whether this is correct. I've read alot on it, but it seems there are unclear definitions of "long memory" and as thus, some argue that APARCH explains the long memory in financial timeseries, some argue that IGARCH doesn't etc.
If some of you have some knowledge/input on this, I would be very grateful!
Best regards,
Martin Falch
I'm writing an assignment on GARCH models and wish to divide them into "leverage-effect models" and "long memory models" respectively. As such, I've put EGARCH, GJR and APARCH in the first category, while I intend to put IGARCH, FIGARCH, FIEGARCH, FIEAPARCH and HYGARCH in the second category.
My question is whether this is correct. I've read alot on it, but it seems there are unclear definitions of "long memory" and as thus, some argue that APARCH explains the long memory in financial timeseries, some argue that IGARCH doesn't etc.
If some of you have some knowledge/input on this, I would be very grateful!
Best regards,
Martin Falch