Development and application of new methodologies of combining time series and expert survey data for economic forecasting.

• Applicant: Professor Dr. Winfried Franz Xaver Pohlmeier
• Subject Area: Statistics and Econometrics
• Term: from 2012 to 2019
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 219805061

• Final Report Year: 2019

Final Report Abstract

In this project a forecasting toolbox was developed, which helps to combine the information of traditional financial and macroeconomic time series with the information in panel data of expert surveys. The merits of the strategy were shown to be threefold: (i) Assessment of the quality of various expert surveys, (ii) understanding of the pitfalls and opportunities of information resulting from expert surveys and (iii) improvement of forecasts by the use of micro-information. For example it could be shown, how inclusion of most recent information at the micro level through the time stamped information can used to adjust and improve macroeconomic nowcasts. Bayesian methods were shown to be useful to stabilize macroeconomic forecasts by incorporating time series information of the expert surveys. Moreover, several papers contributed to a better understanding on the dynamics of individual expectations and the underlying individual learning mechanism of experts.

Publications

• (2013): Four Essays on Probabilistic Forecasting in Econometrics, Ph.d. thesis, University of Konstanz, Konstanz
  Krüger , F.
  
• (2014): Three Essays on the Econometrics of Survey Expectations Data, Ph.D. thesis, University of Konstanz, Konstanz
  Mokinski , F.
  
• (2015): “Measuring Disagreement in Qualitative Expectations,” Journal of Forecasting, 34 (5), 405–426
  Mokinski , F., X. S. Sheng , & J. Yang
  (See online at https://doi.org/10.1002/for.2340)
• (2016): “Disagreement versus uncertainty: Evidence from distribution forecasts,” Journal of Banking & Finance, 72, S172 – S186
  Krüger , F. & I. Nolte
  (See online at https://doi.org/10.1016/j.jbankfin.2015.05.007)
• (2016): “Forecasting Conditional Probabilities of Binary Outcomes under Misspecification.” Review of Economics and Statistics, 98 (4), 742 – 755
  Elliott, G., D. Ghanem , & F. Krüger
  (See online at https://doi.org/10.1162/REST_a_00564)
• (2016): “Forecasting with Bayesian Vector Autoregressions Estimated Using Professional Forecasts,” Journal of Applied Econometrics, 31 (6), 1083–1099
  Frey, C. & F. Mokinski
  (See online at https://doi.org/10.1002/jae.2483)
• (2016): “Using time-stamped survey responses to measure expectations at a daily frequency,” International Journal of Forecasting, 32 (2), 271 – 282
  Mokinski , F.
  (See online at https://doi.org/10.1016/j.ijforecast.2015.06.004)
• (2017): Three Essays on Bayesian Shrinkage Methods, Ph.d. thesis, University of Konstanz, Konstanz
  Frey, C.
  
• (2018): “Shrinkage for Categorical Regressors,” Working Paper, University of Konstanz
  Heiler , P. & J. M Arecková
  
• (2019): “Particle learning for Bayesian semi-parametric stochastic volatility model,” Econometric Reviews, 38 (9), 1007– 1023
  Virbickait E , A., H. F. Lopes , M. C. Ausín , & P. Galeano
  (See online at https://doi.org/10.1080/07474938.2018.1514022)
• (2019): “What determines forecasters’ forecasting errors?” International Journal of Forecasting, 35 (1), 11 – 24
  Nolte , I., S. Nolte , & W. Pohlmeier
  (See online at https://doi.org/10.1016/j.ijforecast.2018.07.007)