2020
DOI: 10.2139/ssrn.3593124
|View full text |Cite
|
Sign up to set email alerts
|

Higher Moment Constraints for Predictive Density Combination

Abstract: The majority of financial data exhibit asymmetry and heavy tails, which makes forecasting the entire density critically important. Recently, a forecast combination methodology has been developed to combine predictive densities. We show that combining individual predictive densities that are skewed and/or heavy-tailed results in significantly reduced skewness and kurtosis. We propose a solution to overcome this problem by deriving optimal log score weights under Higher-order Moment Constraints (HMC). The statis… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1

Citation Types

1
2
0

Year Published

2020
2020
2022
2022

Publication Types

Select...
3
2

Relationship

0
5

Authors

Journals

citations
Cited by 5 publications
(3 citation statements)
references
References 39 publications
1
2
0
Order By: Relevance
“…Our results con-tribute to the active literature on frequentist estimation of linear predictive combinations via predictive criteria (e.g. Hall and Mitchell, 2007, Ranjan and Gneiting, 2010, Clements and Harvey, 2011, Geweke and Amisano, 2011, Gneiting and Ranjan, 2013, Kapetanios et al, 2015, Opschoor et al, 2017, Ganics, 2018and Pauwels et al, 2020. In particular, our results provide a possible explanation behind the often mixed out-of-sample performance of optimal weighting schemes.…”
Section: Introductionsupporting
confidence: 78%
“…Our results con-tribute to the active literature on frequentist estimation of linear predictive combinations via predictive criteria (e.g. Hall and Mitchell, 2007, Ranjan and Gneiting, 2010, Clements and Harvey, 2011, Geweke and Amisano, 2011, Gneiting and Ranjan, 2013, Kapetanios et al, 2015, Opschoor et al, 2017, Ganics, 2018and Pauwels et al, 2020. In particular, our results provide a possible explanation behind the often mixed out-of-sample performance of optimal weighting schemes.…”
Section: Introductionsupporting
confidence: 78%
“…The tail of the distribution is also the main feature of interest when measuring downside risk in equity markets. Besides, Pauwels et al (2020) proposed an approach to computing the optimal weights by maximizing the average logarithmic score subject to additional higher moments restrictions. Through the constrained optimization, the combined probability forecast can preserve specific characteristics of the distribution, such as fat tails or asymmetry.…”
Section: Linear Poolingmentioning
confidence: 99%
“…The frequentist literature includes work on linear combinations (or linear pools), in which various measures of predictive accuracy, including scoring rules, are used to define the criterion that is optimized to estimate the weights. Relevant references here include Hall and Mitchell (2007), Kascha and Ravazzolo (2010), Geweke and Amisano (2011), Ganics (2017), Opschoor et al (2017), Martin et al (2020) and Pauwels et al (2020). Non‐linear weighting schemes (or non‐linear transformations of linear schemes)—either estimated via optimization of prediction‐based criteria, or drawing on the principle of predictive calibration (Dawid, 1982; 1985; Gneiting et al, 2007)— are explored in Ranjan and Gneiting (2010), Clements and Harvey (2011), Gneiting and Ranjan (2013) and Kapetanios et al (2015).…”
Section: Simulation Study: Financial Returnsmentioning
confidence: 99%