Liebl, Dominik (2010): Estimation of the Semiparametric Factor Model: Application to Modelling Time Series of Electricity Spot Prices.

PDF
MPRA_paper_26800.pdf Download (514kB)  Preview 
Abstract
Classical univariate and multivariate time series models have problems to deal with the high variability of hourly electricity spot prices. We propose to model alternatively the daily mean electricity supply functions using a dynamic factor model. And to derive, subsequently, the hourly electricity spot prices by the evaluation of the estimated supply functions at the corresponding hourly values of demand for electricity. Supply functions are price (EUR/MWh) functions, that increase monotonically with demand for electricity (MW). Apart from this new conceptual approach, that allows us to represent the auction design of energy exchanges in a most natural way, our main contribution is an extraordinary simple algorithm to estimate the factor structure of the dynamic factor model. We decompose the time series into a functional spherical component and an univariate scaling component. The elements of the spherical component are all standardized having unit size such that we can robustly estimate the factor structure. This algorithm is much simpler than procedures suggested in the literature. In order to use a parsimonious labeling we will refer to the daily mean supply curves simply as price curves.
Item Type:  MPRA Paper 

Original Title:  Estimation of the Semiparametric Factor Model: Application to Modelling Time Series of Electricity Spot Prices. 
Language:  English 
Keywords:  Factor Analysis; functional time series data; sparse data; electricity spot market prices; European Electricity Exchange (EEX) 
Subjects:  C  Mathematical and Quantitative Methods > C1  Econometric and Statistical Methods and Methodology: General > C14  Semiparametric and Nonparametric Methods: General C  Mathematical and Quantitative Methods > C2  Single Equation Models ; Single Variables > C22  TimeSeries Models ; Dynamic Quantile Regressions ; Dynamic Treatment Effect Models ; Diffusion Processes C  Mathematical and Quantitative Methods > C1  Econometric and Statistical Methods and Methodology: General C  Mathematical and Quantitative Methods > C0  General > C01  Econometrics 
Item ID:  26800 
Depositing User:  Dominik Liebl 
Date Deposited:  17. Nov 2010 19:12 
Last Modified:  21. Feb 2013 11:07 
References:  J Antoch, L Prchal, MR Rosa, and P Sarda. Functional linear regression with functional response: Application to prediction of electricity consumption. Functional and Operatorial Statistics, pages 52195219, 2008. M Forni, M Hallin, M Lippi, and M Reichlin. The generalized dynamic factor model: Identication and estimation. The Review of Economics and Statistics, (82): 540554, 2000. W K Härdle and S Trück. The dynamics of hourly electricity prices. 2010. B U Park, E Mammen, W Härdle, and S Borak. Time Series Modelling With Semiparametric Factor Dynamics Journal of the American Statistical Association, 104(485): 284298, 2009. J G Staniswalis and J J Lee. Nonparametric Regression Analysis of Longitudinal Data. Journal of the American Statistical Association, 93(444): 14031418, 1998. F Yao, H G Müller, and J L Wang. Functional Data Analysis for Sparse Longitudinal Data. Journal of the American Statistical Association, 100(470): 577590, 2005. 
URI:  https://mpra.ub.unimuenchen.de/id/eprint/26800 