Time-Varying Combinations of Predictive Densities Using Nonlinear Filtering

Billio, Monica; Casarin, Roberto; Ravazzolo, Francesco; Dijk, Herman K. van

doi:10.2139/ssrn.2172254

Cited by 46 publications

(68 citation statements)

References 83 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As a result, for notational ease only, we temporarily drop the subscript t on w. Note that, unlike Billio et al (2013) and Del Negro et al (2014), we do not explicitly parameterise the time-variation. Our time variation arises due to the need to normalise the combined density to integrate to one; and this normalisation is by construction time-varying.…”

Section: Generalised Density Combinations: Theorymentioning

confidence: 99%

“…This allows for the possibility that while one model may be particularly useful (and receive a high weight in the combination) when the economy or market is in recession or a bear market, for example, another model may be more informative when output growth is positive or there is a bull market. The idea of letting the weights on the component densities vary according to the (forecast) value of the variable of interest contrasts with recent suggestions to let the weights in the combination follow a Markov-switching structure (Waggoner & Zha (2012)) or to evolve over time (Billio et al (2013) and Del Negro et al (2014)). Accommodating time-variation in the combination weights mimics our approach to the extent that over time one moves into different regions of the forecast density.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Generalised density forecast combinations

Kapetanios

Mitchell

Price

et al. 2015

Journal of Econometrics

View full text Add to dashboard Cite

Density forecast combinations are becoming increasingly popular as a means of improving forecast 'accuracy', as measured by a scoring rule. In this paper we generalise this literature by letting the combination weights follow more general schemes. Sieve estimation is used to optimise the score of the generalised density combination where the combination weights depend on the variable one is trying to forecast. Specific attention is paid to the use of piecewise linear weight functions that let the weights vary by region of the density. We analyse these schemes theoretically, in Monte Carlo experiments and in an empirical study. Our results show that the generalised combinations outperform their linear counterparts.JEL Codes: C53

show abstract

Section: Generalised Density Combinations: Theorymentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Generalised density forecast combinations

Kapetanios

Mitchell

Price

et al. 2015

Journal of Econometrics

View full text Add to dashboard Cite

show abstract

“…Another strand of literature related to our work is the focus on density forecast combinations from a Bayesian point of view (see, e.g., Aastveit, Ravazzolo, & Van Dijk, 2016;Billio, Casarin, Ravazzolo, & van Dijk, 2013;Del Negro, Hasegawa, & Schorfheide, 2016;Waggoner & Zha, 2012). These papers treat the combination weights as (time-varying) random variables.…”

Section: Introductionmentioning

confidence: 99%

Combining density forecasts using focused scoring rules

Opschoor

Dijk

Wel

2017

J of Applied Econometrics

View full text Add to dashboard Cite

SummaryWe investigate the added value of combining density forecasts focused on a specific region of support. We develop forecast combination schemes that assign weights to individual predictive densities based on the censored likelihood scoring rule and the continuous ranked probability scoring rule (CRPS) and compare these to weighting schemes based on the log score and the equally weighted scheme. We apply this approach in the context of measuring downside risk in equity markets using recently developed volatility models, including HEAVY, realized GARCH and GAS models, applied to daily returns on the S&P 500, DJIA, FTSE and Nikkei indexes from 2000 until 2013. The results show that combined density forecasts based on optimizing the censored likelihood scoring rule significantly outperform pooling based on equal weights, optimizing the CRPS or log scoring rule. In addition, 99% Value-at-Risk estimates improve when weights are based on the censored likelihood scoring rule.

show abstract

“…This leads to a more complex latent variable model, both in terms of modeling and computation. Alternatively, one can use completely unobserved combination weights (e.g., in the modeling strategies of Billio et al (2013a)) or weights which are partially observed and driven by one or some of the variables mentioned above. Given the already large number of latent variables in our model, the latter weight specification strategy is preferred in order to avoid overfitting problems.…”

Section: Interaction Mechanismsmentioning

confidence: 99%

Interconnections between Eurozone and US Booms and Busts Using a Bayesian Panel Markov-Switching VAR Mode

et al. 2015

Self Cite

View full text Add to dashboard Cite

Standard-Nutzungsbedingungen:Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden.Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen.Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in AbstractInterconnections between Eurozone and United States booms and busts and among major Eurozone economies are analyzed using a Panel Markov-Switching VAR model. The model accommodates changes in low and high data frequencies and incorporates endogenous time-varying transition matrices of country-specific Markov chains. These country-specific Markov chains depend on their own past history and the history of other chains, thus allowing for interconnections between cycles, and an endogenous common Eurozone cycle is derived by aggregating the country-specific cycles. The model is estimated using a simulation based Bayesian approach in which an efficient multimove sampling algorithm is defined to draw time-varying Markov-switching chains. Using industrial production growth and credit spread data for all countries, several empirical results have emerged. Recession, slow growth and expansion are empirically identified as three regimes with slow growth becoming persistent in the Eurozone in recent years different from the US. The Eurozone and the US regimes appear not fully synchronized, with evidence of more recessions in the Eurozone. Second, turning point analysis indicates larger synchronization at the beginning of the Great Financial Crisis: this shock affects the US first, leading the Eurozone cycle, and spreads then rapidly among these economies. Third, amplification effects influence recession probabilities for Eurozone countries when shocks occur. The evidence is different for the US where this reinforcement does not exist. In recent years there are more imbalances among regimes in Eurozone countries. Fourth, a credit shock results in substantial negative industrial production growth for several months in Germany, Spain and the US. Our empirical results point towards the existence of a group of core countries: Germany, France and * We thank the co-editor Fabio Canova, three anonymous referees and conference and seminar participants

show abstract

Time-Varying Combinations of Predictive Densities Using Nonlinear Filtering

Cited by 46 publications

References 83 publications

Generalised density forecast combinations

Generalised density forecast combinations

Combining density forecasts using focused scoring rules

Interconnections between Eurozone and US Booms and Busts Using a Bayesian Panel Markov-Switching VAR Mode

Contact Info

Product

Resources

About