Bayesian VAR forecasts, survey information, and structural change in the euro area

Ganics, Gergely; Odendahl, Florens

doi:10.1016/j.ijforecast.2020.11.001

Cited by 15 publications

(11 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Tilting model-based inflation forecasts to SPF expectations has been recently investigated, but for somewhat different reasons. Ganics and Odendahl (2021) focus on the euro area and show that survey forecasts can help mitigate the effects of structural breaks on the forecasting performance in a VAR. They also find that professional forecasters are better at forecasting than a standard Bayesian VAR model around the two recent euro area recessions, as well as the slow recovery thereafter.…”

Section: 'Soft Conditioning' a La Robertson Et Al (2005) Using Inflation Expectations From The Survey Of Professional Forecastersmentioning

confidence: 99%

“…Tilting towards SPF expectations can be a valid way to provide the model some degree of informed judgement, as professional forecasters do not solely rely on models (which might have been affected by COVID-19 observations), but use a substantial degree of judgement when forming their believes about the future. Also, there is evidence that such tilting improves inflation forecasting in certain challenging times, such as in the post-Great Recession period, as a way of indirectly accommodating structural changes (see Tallman and Zaman (2020), Ganics and Odendahl (2021)).…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

The COVID-19 Shock and Challenges for Time Series Models

Bobeica

Hartwig

2021

SSRN Journal

View full text Add to dashboard Cite

We document the impact of COVID-19 on frequently employed time series models, with a focus on euro area inflation. We show that for both single equation models (Phillips curves) and Vector Autoregressions (VARs) estimated parameters change notably with the pandemic. In a VAR, allowing the errors to have a distribution with fatter tails than the Gaussian one equips the model to better deal with the COVID-19 shock. A standard Gaussian VAR can still be used for producing conditional forecasts when relevant off-model information is used. We illustrate this by conditioning on official projections for a set of variables, but also by tilting to expectations from the Survey of Professional Forecasters. For Phillips curves, averaging across many conditional forecasts in a thick modelling framework offers some hedge against parameter instability.

show abstract

Section: 'Soft Conditioning' a La Robertson Et Al (2005) Using Inflation Expectations From The Survey Of Professional Forecastersmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

The COVID-19 Shock and Challenges for Time Series Models

Bobeica

Hartwig

2021

SSRN Journal

View full text Add to dashboard Cite

show abstract

“…Bayes theorem is widely used in the field of data analysis and is often used to analyze the conditional probability of numerous events, such as forecasting hierarchically structured time series data [20], seasonal time series data [21,22], multi-step-ahead time series prediction [23], general estimation and prediction [24], and statistical analysis [25,26]. A Bayesian approach has been presented to forecast univariate time series data by implementing a technique of sampling the future in [27].…”

Section: Bayesian Approachmentioning

confidence: 99%

A Bayesian Model to Forecast the Time Series Kinetic Energy Data for a Power System

2021

View full text Add to dashboard Cite

Withthe massive penetration of electronic power converter (EPC)-based technologies, numerous issues are being noticed in the modern power system that may directly affect system dynamics and operational security. The estimation of system performance parameters is especially important for transmission system operators (TSOs) in order to operate a power system securely. This paper presents a Bayesian model to forecast short-term kinetic energy time series data for a power system, which can thus help TSOs to operate a respective power system securely. A Markov chain Monte Carlo (MCMC) method used as a No-U-Turn sampler and Stan’s limited-memory Broyden–Fletcher–Goldfarb–Shanno (LM-BFGS) algorithm is used as the optimization method here. The concept of decomposable time series modeling is adopted to analyze the seasonal characteristics of datasets, and numerous performance measurement matrices are used for model validation. Besides, an autoregressive integrated moving average (ARIMA) model is used to compare the results of the presented model. At last, the optimal size of the training dataset is identified, which is required to forecast the 30-min values of the kinetic energy with a low error. In this study, one-year univariate data (1-min resolution) for the integrated Nordic power system (INPS) are used to forecast the kinetic energy for sequences of 30 min (i.e., short-term sequences). Performance evaluation metrics such as the root-mean-square error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE), and mean absolute scaled error (MASE) of the proposed model are calculated here to be 4.67, 3.865, 0.048, and 8.15, respectively. In addition, the performance matrices can be improved by up to 3.28, 2.67, 0.034, and 5.62, respectively, by increasing MCMC sampling. Similarly, 180.5 h of historic data is sufficient to forecast short-term results for the case study here with an accuracy of 1.54504 for the RMSE.

show abstract

“…By contrast, models such as quantile regressions are developed and estimated specifically to assess risks and uncertainty, as their focus of interest is not the mean or median, but other quantiles of the distribution. Many NCBs in the Eurosystem have developed this kind of models or are currently working on them (see for example Ganics and Odendahl, 2019).…”

Section: Model-based Risk Assessmentmentioning

confidence: 99%

“…All of this requires more sophisticated methods than when working with point forecasts only. Ganics and Odendahl (2019) use both entropic tilting and soft conditioning to input the distributional information from the ECB's Survey of Professional Forecasters into a BVAR model and show how this improves both the point and density forecasts from the model. Box 13 also illustrates methods to obtain risk measures from multivariate density forecasts: following Odendahl (2020), it proposes estimating joint density forecasts on the basis of univariate density forecasts available from surveys and copula functions.…”

Section: Scope For Improvementmentioning

confidence: 99%

Review of Macroeconomic Modelling in the Eurosystem: Current Practices and Scope for Improvement

et al. 2021

View full text Add to dashboard Cite

This paper provides an assessment of the macroeconomic models regularly used for forecasting and policy analysis in the Eurosystem. These include semi-structural, structural and time-series models covering specific jurisdictions and the euro area within a closed economy, small open economy, multi-country or global setting. Models are used as analytical frameworks for building baseline projections and for supporting the preparation of monetary policy decisions. The paper delivers four main contributions. First, it provides a survey of the macroeconomic modelling portfolios currently used or under development within the Eurosystem. Second, it explores the analytical gaps in the Eurosystem models and investigates the scope for further enhancement of the main projection and policy models, and the creation of new models. Third, it reviews current practices in model-based analysis for monetary policy preparation and forecasting and provides recommendations and suggestions for improvement. Finally, it reviews existing cooperation modalities on model development and proposes alternative sourcing and organisational strategies to remedy any knowledge or analytical gaps identified.

show abstract

Bayesian VAR forecasts, survey information, and structural change in the euro area

Cited by 15 publications

References 32 publications

The COVID-19 Shock and Challenges for Time Series Models

The COVID-19 Shock and Challenges for Time Series Models

A Bayesian Model to Forecast the Time Series Kinetic Energy Data for a Power System

Review of Macroeconomic Modelling in the Eurosystem: Current Practices and Scope for Improvement

Contact Info

Product

Resources

About