Is it possible to break the «curse of dimensionality»? Spatial specifications of multivariate volatility models

The article is devoted to the estimation of multivariate volatility of a portfolio consisted from twenty American stocks. The six specifications of multivariate volatility models are formulated and estimated. It’s demonstrated that spatial specifications of multivariate volatility models allow not only reduce the dimension of the problem, but in some cases outdo original specifications at in-sample and out-of-sample comparison.

MoreLess

Year of publication:	2014
Authors:	Lakshina, Valeriya
Published in:	Applied Econometrics. - Publishing House "SINERGIA PRESS". - Vol. 36.2014, 4, p. 61-78
Publisher:	Publishing House "SINERGIA PRESS"
Subject:	multivariate volatility models \| curse of dimensionality \| weight matrix \| spatial autoregression \| forecasting

Check Google Scholar |

More access options

In German libraries (KVK)

I need help

More details

Type of publication:	Article
Classification:	C01 - Econometrics ; C51 - Model Construction and Estimation ; c58 ; G17 - Financial Forecasting
Source:	RePEc - Research Papers in Economics

Persistent link: https://www.econbiz.de/10011106004