Econometrical analyzation of day-ahead market prices

OData support
Dr. Divényi Dániel Péter
Department of Electric Power Engineering

Until the end of the last decade of the $20^{th}$ century electricity was considered to be the dominant technology, which was followed by the new wave: the information technology. At the dawn of the millennium it seems that sustainability and the depletion of the natural resources have gotten into the spectrum, therefore electricity can become mainstream again, because manpower, engines and furnaces with internal combustion retrieve energy from fossil fuels, which, the most part, come to become the form of electricity too.

It sounds a bit unusual that electricity is bought and sold on the market, but that is the case on this market, the market of electricity, which works as a stock-product. In this case it is not only essential to manage accounting in a high way, but transport of the power is significant as well in terms of logistics. Limits of storing electricity makes researching of the market even more exciting.

My thesis details how the market of electricity works, including the specialty of the Hungarian market, its technical specifications, the participants of the chain, and also a short guide of the Hungarian stock-market of the electricity. The HUPX-DAM will also be mentioned in a detailed way, with a focus on their general rules, transparencies, the daily routine of how the rates are generated, their products and the possible links between markets in terms of their advantages and disadvantages. It summarizes the types of temporal forecasts, and shows ARIMA forecast in details, and validation system (linear regression), which I would be working with.

Using this methodology I divided my thesis into two parts: the learning and the testing phases. At the learning phase I applied the mathematical methodology to the given data, which provided the ideal forecast of the learning phase, then at the test phase I validated its accuracy, and if it was necessary I modified the model so that I could be more accurate and I would be able to correctly forecast. I went on making the rates I got even more authentic and finally I was looking for the ideal rate between the two models, which could lead me to find the minimal margin of error with using both of them at the same time. During my studies I was mostly dealing with the forecast of BASE and PEAK prices, so I was trying to forecast the given hourly prices with using this method.

While researching my primary aim was to understand, and within 5% of margin of error to indicate the processes of the prices in the HUPX DAM market.

According to my hypothesis it is possible to build up a mathematical model with using DAM prices in a long run and the factors that influence the price of the electricity, so that we would be able to find a forecast that is well within the 5% of margin of error.


Please sign in to download the files of this thesis.