UZH-Logo

Maintenance Infos

Weighted maximum likelihood for risk prediction


Steude, Sven C (2011). Weighted maximum likelihood for risk prediction. NCCR FINRISK 689, University of Zurich.

Abstract

Most time series models used in econometrics and empirical finance are estimated with maximum likelihood methods, in particular when interest centers on density and Value-at-Risk (VaR) prediction. The standard maximum likelihood principle implicitly places equal weight on each of the observations in the sample, but depending on the extent to which the model and the true data generating process deviate this can be improved upon. For example, in the context of modeling financial time series, weighting schemes which place relatively more weight on observations in the recent past result in improvement of out-of-sample density forecasts, compared to the default of equal weights. Also, if instead of accurate forecasting of the entire density, interest is restricted to just downside risk, placing more weight on the negative observations in the sample improves results further. In this paper, a third and quite general strategy of shifting more weight towards certain observations of the sample is proposed. Weights are derived from external variables that convey additional information about the true DGP, like trading volume, news arrivals or even investor sentiment. As such, those observations are down weighted that bear a high probability of being destructive outliers with no bene¯t of using them when fitting the model. Considerable improvements in forecast accuracy for a variety of data sets and different time series models can be realized.

Most time series models used in econometrics and empirical finance are estimated with maximum likelihood methods, in particular when interest centers on density and Value-at-Risk (VaR) prediction. The standard maximum likelihood principle implicitly places equal weight on each of the observations in the sample, but depending on the extent to which the model and the true data generating process deviate this can be improved upon. For example, in the context of modeling financial time series, weighting schemes which place relatively more weight on observations in the recent past result in improvement of out-of-sample density forecasts, compared to the default of equal weights. Also, if instead of accurate forecasting of the entire density, interest is restricted to just downside risk, placing more weight on the negative observations in the sample improves results further. In this paper, a third and quite general strategy of shifting more weight towards certain observations of the sample is proposed. Weights are derived from external variables that convey additional information about the true DGP, like trading volume, news arrivals or even investor sentiment. As such, those observations are down weighted that bear a high probability of being destructive outliers with no bene¯t of using them when fitting the model. Considerable improvements in forecast accuracy for a variety of data sets and different time series models can be realized.

Downloads

1254 downloads since deposited on 27 Sep 2011
491 downloads since 12 months
Detailed statistics

Additional indexing

Item Type:Working Paper
Communities & Collections:03 Faculty of Economics > Department of Banking and Finance
Dewey Decimal Classification:330 Economics
JEL Classification:C22, C51, G10
Language:English
Date:2011
Deposited On:27 Sep 2011 09:06
Last Modified:05 Apr 2016 15:01
Series Name:NCCR FINRISK
Free access at:Official URL. An embargo period may apply.
Official URL:http://www.zora.uzh.ch/49788/
Related URLs:http://www.nccr-finrisk.uzh.ch/wps.php?action=query&id=689
Permanent URL: https://doi.org/10.5167/uzh-49788

Download

[img]
Preview
Content: Published Version
Filetype: PDF
Size: 2MB

TrendTerms

TrendTerms displays relevant terms of the abstract of this publication and related documents on a map. The terms and their relations were extracted from ZORA using word statistics. Their timelines are taken from ZORA as well. The bubble size of a term is proportional to the number of documents where the term occurs. Red, orange, yellow and green colors are used for terms that occur in the current document; red indicates high interlinkedness of a term with other terms, orange, yellow and green decreasing interlinkedness. Blue is used for terms that have a relation with the terms in this document, but occur in other documents.
You can navigate and zoom the map. Mouse-hovering a term displays its timeline, clicking it yields the associated documents.

Author Collaborations