| dc.description.abstract | Because of the increasing volatile power infeed and the uncertainty associated, the temporal granularity of the auctions in the European energy markets has increased in recent years. Examples for this are the introduction of quarter hourly intraday trading and the shortening of contracts for balancing reserve. This leads to an increased number of profit opportunities, and therefore, a multitude of complex bidding strategies arise. An optimal bidding strategy for a power plant can only be derived by taking all the following profit opportunities in the consecutive energy markets into account. Such a bidding behaviour is also known as coordinated bidding. Nevertheless, power traders currently still consider the markets in a deterministic and sequential manner and base bidding decisions on rules of thumbs.
Only some studies in the literature try to capture the interdependencies of sequential markets while deriving bidding strategies. Even fewer quantify the difference in profit induced by the coordination of bidding compared to bidding based on heuristics. This is also referred to as the gain of coordinated bidding. Assessing that gain, however, appears to be of practical use to traders that want to maximise their profit, since it supports an informed decision whether or not to switch to coordinated bidding.
This thesis hence aims to answer the research question “What effect does coordinated bidding have on the generated profit of a trading party in comparison to heuristic bidding for different portfolio compositions varying by flexibility, size, price impact and risk preferences?”. Therefore, factors influencing the gain of coordination are identified from literature and a comparison between a multi-stage stochastic coordinated bidding model by Kraft et al. (2021) and a developed heuristic is carried out. The coordinated bidding model depicts three markets in the common European market setting, namely the secondary reserve capacity market, the day-ahead and intraday market. In this market setup, the uncertainty of the realised prices is considered in a three-stage scenario tree. In order to analyse all the influencing factors, the price taker assumption in the original model is relaxed, by modeling a price effect.
From the results, it can be summarised that the gain of coordination in the analysed market setting is substantially higher than the one found in similar analyses for the Nordic markets. It reaches up to 25% for a portfolio comprising 100 MW PV and 100 MW biomass compared to the 3% found by Aasgård (2022) for a 50 MW hydro power plant. The modeling of a price maker in the intraday market in comparison to the price taker assumption in Kraft et al. (2021) has a concise impact on the results, even for the small portfolio considered. The price taker assumption seems to overestimate the attractiveness of the intraday market, which shows the importance of a price impact modeling. From a sensitivity analysis, it can also be concluded that a rise in portfolio size decreases the relative gain of coordination significantly. The statistical arbitrage between the two spot markets can be exploited less, due to the higher price decrease in the intraday market. Similarly, an increase in risk aversion of the trader, the price impact and the flexibility of the power plants decrease the gain associated with coordinated bidding. | |