Modelling and predicting the Distribution of Risk Premia in Mid-Term Electricity Futures
MetadataVis full innførsel
This thesis examines risk premia in mid-term electricity futures traded in the Nordic electricity market Nord Pool using a time series of 8 years worth of data. Using OLS and quantile regression the relationship between the forward premium and several economic and physical conditions are examined. The reservoir levels and the basis are found to have a significant relationship with the forward premium both in the OLS and the quantile regression models. The volatility of the spot price is found to have negative effects on forward premia below the median and a positive effect on forward premia above the median, indicating that the effect of volatility on forward premia is highly dependent on the quantile being modelled. The OLS coefficient for volatility is very close to zero indicating that the effect on the mean is not significant. The effects of all the considered variables are considerably larger in the tails indicating that their impact may be larger than estimated in the OLS model, though the results are uncertain. Evidence is also found to support the notion that market efficiency is increasing as forward premia have decreased over time. Parameters are less significant when analyzing only the most recent data. In addition forward premia have decreased and open interest in the futures has increased in the last portion of the data sample. This suggests that speculative interest has increased and investors have gained experience.