Chair in Statistics and Empirical Economics Research
Raum 304
Phone:
+049 431 880-3810
Telefax:
+049 431 880-7605
mdeme@stat-econ.uni-kiel.de
Office hours:
By appointment
Lectures:
For current lectures see Univis
Brief profile:
Biosketch
I studied engineering and business administration at the "Politehnica" University in Bucharest and graduated in 2000 as best of my class. After a brief period at the Technical University in Darmstadt as a research and teaching assistant, I started to work on my PhD under the supervision of Uwe Hassler at the Goethe University in Frankfurt.
In 2005, I obtained my PhD in econom(etr)ics; the PhD thesis consists of four papers dealing with misspecifications in models with seasonal deterministics and issues in forecasting under asymmetric loss functions. Having chosen to pursue an academic career, I went on as a post-doc researcher in Frankfurt. During this period I started focussing on panels with cross-unit dependence.
During the academic year 2007-2008, I was a Max Weber fellow at the European University Institute in Florence. My mentor during the stay in Tuscany was Helmut Lütkepohl. In summer 2008, I returned to Frankfurt on a position as junior professor for applied econometrics, and moved to the University of Bonn in 2010, where I was a professor at the Hausdorff Center for Mathematics. I joined the University of Kiel in April 2014.
Teaching
A student's understanding of applied statistics (and also econometrics) can benefit a lot from treating the relevant models and methods with some rigor. Rigor, however, is just means to an end, namely the successful application of statistical methods to real-life data. Please consult the teaching section for more details on specific courses; see above for courses I'm teaching this semester.
Research focuses
Forecasting, in particular with many predictors; Financial data; Complexity reduction.
Third-party funding
- DFG: "Zeitvariierende Dynamik in Paneldaten mit stochastischen Trends," gemeinsam mit Christoph Hanck (Uni Duisburg-Essen); 2016-2018, Projektmitarbeiterin: Danvee Floro, M.Sc.
- DFG: "Zeitvariierende Volatilität in Paneldaten mit stochastischen Trends," gemeinsam mit Christoph Hanck (Uni Duisburg-Essen); 2014-2016, abgeschlossen. (Ehemalige Projektmitarbeiterin: Danvee Floro, M.Sc.)
- DFG: "Approximation und Aggregation bei der Modellierung und Vorhersage persistenter Zeitreihen,'' gemeinsam mit Uwe Hassler (Uni Frankfurt); 2012-2015, abgeschlossen. (Ehemaliger Projektmitarbeiter: Nazarii Salish)
Editor
Statistics: A Journal of Theoretical and Applied Statistics (co-editor)
Computational Statistics
Selected publications (see CV for complete list):
- Demetrescu, M. and B. Hillmann (2020), Nonlinear Predictability of Stock Returns? Parametric vs. nonparametric inference in predictive regressions. Journal of Business and Economic Statistics forthcoming.
- Demetrescu, M., I. Georgiev, P.M.M. Rodrigues and A.M.R. Taylor (2020), Testing for Episodic Predictability in Stock Returns. Journal of Econometrics forthcoming.
- Demetrescu, M. and U. Homm (2016), Tests for No Cross-Sectional Error Correlation in Large-N Panel Data Models. Journal of Applied Econometrics 31 (1), 4-31.
- Demetrescu, M. and U. Hassler (2016), (When) Do Long Autoregressions Account for Neglected Changes in Parameters? Econometric Theory 32(6), 1317-1348.
- Breitung, J. and M. Demetrescu (2015), Instrumental Variable and Variable Addition Based Inference in Predictive Regressions. Journal of Econometrics 187 (1), 358-375. Companion working paper.
Working papers
- Robust Inference under Time-Varying Volatility: A Real-Time Evaluation of Professional Forecasters, with C. Hanck and R. Kruse (latest version)
- On the use of VAR models with ignored changes in mean and variance, with N. Salish
- Inference on the Asymmetry of Loss Functions with Persistent Instruments, with C. Roling and A. Titova (working paper version); published as "Re-Evaluating the Prudence of Economic Forecasts in the EU: The role of instrument persistence", Journal of Applied Econometrics forthcoming.
- Testing for constant correlation of filtered series under structural change, with D. Wied; published as "Testing for constant correlation of filtered series under structural change", The Econometrics Journal 22(1) 2019, 10–33.
- Finite-sample size control of IVX-based tests in predictive regressions, with M. Hosseinkouchack (accepted for publication in Econometric Theory)
- Robust fixed-b inference in the presence of time-varying volatility, with C. Hanck and R. Kruse
- Residual-augmented IVX predictive regression, with P.M.M. Rodrigues (accepted for publication in Journal of Econometrics) working paper version.
- Homogenous vs. Heterogenous Transition Functions in Smooth Transition Regressions: An LM-Type Test, with J. Leppin and S. Reitz (published in Econometric Reviews 40(2) 2021, 177–196)
- Bias Corrections for Exponentially Transformed Forecasts: Are they worth the effort? with V. Golosnoy and A. Titova; published as Bias corrections for exponentially transformed forecasts: Are they worth the effort?, International Journal of Forecasting 36(3) 2020, 761–780.
- Testing the Fractionally Integrated Hypothesis using M Estimation, with P.M.M. Rodrigues and A. Rubia
- On the Long-Memory Properties of Modulated Processes, with P. Sibbertsen; published as "Inference on the Long-Memory Properties of Time Series with Non-Stationary Volatility"; Economics Letters 144 2016, 80–84
- Nonstationary-Volatility Robust Non-Cointegration Tests with an Application to Cross-Dependent Panels, with C. Hanck; published as "Multiple Testing for No Cointegration under Nonstationary Volatility"; Oxford Bulletin of Economics and Statistics 80(3) 2018, 485-513
- Testing the Predictive Ability of Possibly Persistent Variables under Asymmetric Loss, with C. Roling (we're currently revising)
- Regression-Based Fractional Cointegration Testing under Estimated Degree of Integration, with V. Kuzin and N. Salish (we're currently revising)
- Factor-Based Forecasting under Asymmetric Loss, with S. Hacioglu Hoke; published as "Predictive regressions under asymmetric loss: Factor augmentation and model selection;" International Journal of Forecasting 35(1) 2019, 80-99
Work in progress & other
- Fast Extraction of Common Factors from Heterogeneous Data
- IVX-Based Panel Predictive Regressions for Stock Returns, with C. Hanck
- Testing the Predictability of Stock Returns with Smoothly Varying Deterministic Mean, with M. Hosseinkouchak
- Cross-Sectional Error Dependence in Panel Quantile Regressions, with M. Hosseinkouchak
- Identification and Estimation of Dynamic Factor Models, with J. Breitung
- Monitoring Value-at-Risk and Expected Shortfall Forecasts, with Y. Hoga
- Spurious Long-Range Dependence in Realized Volatilities? A panel examination, with C. Pigorsch and A.I. Tarcolea
- Long Autoregressions under Asymmetric Loss, with A. Titova
- Testing the Marginal Normality of Heteroskedastic Series, with R. Kruse
- Do persistent predictors predict stock returns directionally? with C. Leschinski
- Fractional Integration and Cointegration Testing Using the Sample Mean (not working on it anymore)
- Corrigendum to "Testing predictive regression models with nonstationary regressors" (Zongwu Cai and Yunfei Wang, Journal of Econometrics 178, 4-14)
Further information: