Component estimation for electricity prices: Procedures and comparisons

- We compare methods for component estimation in electricity price time series.
- 11 methods for long-run dynamics and 3 for periodic behavior are considered.
- 3 criteria for filtering evaluation are proposed.
- We find that a procedure based on smoothing splines and trimmed mean is the best for component estimation.

Electricity price time series usually exhibit some form of nonstationarity, corresponding to long-term behavior, one or more periodic components as well as dependence on calendar effects. As a result, modeling electricity prices requires accounting for both long-term and periodic components. In the literature, several filtering procedures have been proposed but a standard has not yet been found. Furthermore, since different procedures are applied in contexts that are not homogeneous with respect to data, periods and final goals, a fair comparison is difficult. This work considers several methods for component estimation in a homogeneous framework and compares them according to specific criteria. The final purpose is to find an estimation procedure that performs well, independently of the intended market and that can be proposed as a reference for electricity price time series filtering.