Uncertainty in Electricity Markets from a Seminonparametric Approach

The spot price of electricity is highly skewed and heavy-tailed, as a result of the interaction of different variables that affect that market. Such characteristics impact the design of power plants with different technologies, fuel prices, and energy demand. This paper introduces the semi-non-parametric (SNP) approach to describe the uncertainty of different variables in an electricity market, reducing the limitations that normality and parametric density functions impose. The selection of probability density functions is achieved in terms of a finite Gram-Charlier expansion fitted by the maximum likelihood criterion. The study case is the Colombian electricity market, where the SNP distribution outperforms the normal distribution for spot price, national energy demand, the climate index ONI, and the series of hydrologic inflows of the system and some rivers. The results show that risk analysis in electricity markets requires the measurement of skewness, kurtosis, and high-order moments. The flexible methodology in our study has directly applications for implementing policies on electricity markets that improve the sustainability indicators of different systems. The particular characteristics of the series under analysis should be considered as a starting point for risk analysis and portfolio choice