Hypernormal Densities

We derive a new family of probability densities that have the property of closed-form integrability. This flexible family finds a variety of applications, of which we illustrate density forecasting from models of the AR-ARCH class for U.S. inflation. We find that the hypernormal distribution for the model's disturbances leads to better density forecasts than the ones produced under the assumption that the disturbances are Normal or Student's t.

MoreLess

Year of publication:	2002-09-01
Authors:	Giacomini, Raffaella ; Gottschling, Andreas ; Haefke, Christian ; White, Halbert
Institutions:	Department of Economics, Boston College
Subject:	ARMA-GARCH models \| neural networks \| nonparametric density estimation \| forecast accuracy

freely available

Full text |

More access options

Check Google Scholar

In German libraries (KVK)

I need help

More details

Extent:	application/pdf
Series:	Boston College Working Papers in Economics.
Type of publication:	Book / Working Paper
Notes:	The text is part of a series Boston College Working Papers in Economics Number 584 42 pages
Classification:	C63 - Computational Techniques ; C53 - Forecasting and Other Model Applications ; C45 - Neural Networks and Related Topics
Source:	RePEc - Research Papers in Economics

Persistent link: https://www.econbiz.de/10004968837