Forecasting Using Relative Entropy

Robertson, John C.; Tallman, Ellis W.; Whiteman, Charles H.

doi:10.1353/mcb.2005.0034

Cited by 119 publications

(94 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As noted by Robertson et al , (), there are several ways to interpret the entropic tilt and the information associated with it. Recall that the initial weights π i embody information from the likelihood function and any prior information regarding the parameters θ .…”

Section: Best‐performing Priors Determined By Entropic Tiltingmentioning

confidence: 99%

See 1 more Smart Citation

Empirical Bayesian Density Forecasting in Iowa and Shrinkage for the Monte Carlo Era

Lewis

Whiteman

2014

Journal of Forecasting

Self Cite

View full text Add to dashboard Cite

Non-technical summaryThe track record of a sixteen-year history of density forecasts of state tax revenue in Iowa is studied, and potential improvements sought through a search for better performing "priors" similar to that conducted two decades ago for point forecasts by Doan, Litterman, and Sims (Econometric Reviews, 1984). Comparisons of the point-and density-forecasts produced under the flat prior are made to those produced by the traditional (mixed estimation) "Bayesian VAR" methods of Doan, Litterman, and Sims, as well as to fully Bayesian, "Minnesota Prior" forecasts.The actual record, and to a somewhat lesser extent, the record of the alternative procedures studied in pseudo-real-time forecasting experiments, share a characteristic: subsequently realized revenues are in the lower tails of the predicted distributions "too often". An alternative empirically-based prior is found by working directly on the probability distribution for the VAR parameters, seeking a betterperforming entropically tilted prior that minimizes in-sample meansquared-error subject to a Kullback-Leibler divergence constraint that the new prior not differ "too much" from the original. We also study the closely related topic of robust prediction appropriate for situations of ambiguity. Robust "priors" are competitive in out-of-sample forecasting; despite the freedom afforded the entropically tilted prior, it does not perform better than the simple alternatives. Nicht-technische ZusammenfassungEs wird die Treffsicherheit der seit sechzehn Jahren durchgeführten Dichteprognosen zu den Steuereinnahmen des US-Bundesstaates Iowa analysiert und nach möglichen Verbesserungen gesucht; dies geschieht durch die Suche nach besseren "Priors" ähnlich wie vor zwei Jahrzehnten bei den Punktprognosen von Doan, Litterman und Sims (Econometric Reviews, 1984). Die Punktund Dichteprognosen auf der Grundlage des flachen Priors werden mit jenen der traditionellen Bayes'schen VAR-Methoden nach Doan, Litterman und Sims sowie mit den reinen Bayesianischen "Minnesota Prior"-Prognosen, verglichen. Das tatsächliche Ergebnis und -in etwas geringerem Umfang -auch jenes der alternativen Verfahren, die anhand von Experimenten in Pseudoechtzeit untersucht werden, haben eines gemeinsam: Die tatsächlich erzielten Einnahmen liegen "zu oft" am unteren Rand der vorausgesagten Verteilung. Ein alternativer, empirisch-basierter Prior lässt sich ermitteln, indem die Wahrscheinlichkeitsverteilung für die VAR-Parameter direkt verwendet wird und ein besserer Prior gesucht wird, der den quadrierten mittleren "in-sample"-Fehler minimiert, und zwar unter der Bedingung, dass der neue Prior nicht "zu stark" vom Original abweicht. Wir untersuchen auch das eng verwandte Thema einer stabilen Vorhersage, die für nicht eindeutige Situationen zweckmäßig ist. Stabile "Priors" sind in "out-of-sample"-Prognosen kompetitiv; trotz des Freiraums dieses echteren Priors schneidet er nicht besser ab als die einfachen Alternativen. EMPIRICAL BAYESIAN DENSITY FORECASTING IN IOWA AND SHRINKAGE FOR THE ...

show abstract

Section: Best‐performing Priors Determined By Entropic Tiltingmentioning

confidence: 99%

“… 11 See Robertson et al , () for a related way of modifying forecasts to reflect moment conditions not used directly in their production.…”

mentioning

confidence: 99%

Empirical Bayesian Density Forecasting in Iowa and Shrinkage for the Monte Carlo Era

Lewis

Whiteman

2014

Journal of Forecasting

Self Cite

View full text Add to dashboard Cite

show abstract

“…In this paper we let f be the main scenario's predictive density and we will measure the distance between this distribution and the predictive densities from all scenarios. The empirical usefulness of the Kullback–Leibler information criterion (KLIC) has been thoroughly established, and it has been employed as an evaluation measure in recent work by, for example, Cogley, Morozov and Sargent (2005) and Robertson, Tallman and Whiteman (2005) 12…”

Section: Combining Predictive Densitiesmentioning

confidence: 99%

Incorporating Judgement in Fan Charts

Österholm

2009

Scandinavian J Economics

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 AbstractWithin a decision-making group, such as the monetary-policy committee of a central bank, group members often hold differing views about the future of key economic variables. Such differences of opinion can be thought of as reflecting differing sets of judgement. This paper suggests modelling each agent's judgement as one scenario in a macroeconomic model. Each judgement set has a specific dynamic impact on the system, and accordingly, a particular predictive density -or fan chart -associated with it. A weighted linear combination of the predictive densities yields a final predictive density that correctly reflects the uncertainty perceived by the agents generating the forecast. In a model-based environment, this framework allows judgement to be incorporated into fan charts in a formalised manner.JEL Classification: C15, C53, E17, E50

show abstract

“…Recently, a number of studies have used exponential tilting (Robertson et al , ) to incorporate moment restrictions—e.g. from survey forecasts—into predictive densities obtained from macroeconomic time series models.…”

Section: Introductionmentioning

confidence: 99%

Forecasting with Bayesian Vector Autoregressions Estimated Using Professional Forecasts

Frey

Mokinski

2015

J of Applied Econometrics

View full text Add to dashboard Cite

SUMMARYWe propose a Bayesian shrinkage approach for vector autoregressions (VARs) that uses short-term survey forecasts as an additional source of information about model parameters. In particular, we augment the vector of dependent variables by their survey nowcasts, and claim that each variable modelled in the VAR and its nowcast are likely to depend in a similar way on the lagged dependent variables. In an application to macroeconomic data, we find that the forecasts obtained from a VAR fitted by our new shrinkage approach typically yield smaller mean squared forecast errors than the forecasts obtained from a range of benchmark methods.

show abstract

Forecasting Using Relative Entropy

Cited by 119 publications

References 23 publications

Empirical Bayesian Density Forecasting in Iowa and Shrinkage for the Monte Carlo Era

Empirical Bayesian Density Forecasting in Iowa and Shrinkage for the Monte Carlo Era

Incorporating Judgement in Fan Charts

Forecasting with Bayesian Vector Autoregressions Estimated Using Professional Forecasts

Contact Info

Product

Resources

About