Balance energy in the German electricity market

Master Thesis in "Mathematical Finance", Oxford (2016)


Gaps between electricity demand and supply are closed by the responsible transmission system operator (TSO) in order to guarantee grid stability and avoid power outages. These gaps are closed by activating balancing energy that is predefined and bought in advance to the actual power delivery by the TSO. Forecasting volumes and prices of balancing energy is crucial for the profit and cost management of operators. Additionally, it can serve as a control signal for power trading strategies. This thesis focusses on the analysis balancing energy prices and volumes in the German power market. We compare statistical models, in particular different autoregressive stochastic models and a linear regression model, to analyse demand and supply balance volumes based on historical data. In a second step we use a Merit-Order model to forecast the marginal price of a balancing energy reserve. Here we also consider regulatory specifications in the calculation scheme of the energy price. With the results of these analyses we discuss their potential effect on power derivatives pricing.


Model Risks in the Valuation of Energy Derivatives

Master Thesis in "Mathematical Finance", Oxford (2016)


Commodity derivatives are usually priced with models that rely on various parameters. An accurate estimation of the model parameters may not always be possible. This gives rise to the question how the uncertainty in the parameters is re ected in the valuation, and which parameters contribute the largest amount of valuation uncertainty. The treatment of parameter uncertainty in derivative valuation has been an area of active research over the last years. One direction into considering parameter risk is given by convex risk measures. These risk measures can be translated into prices by use of risk-capturing functionals. This thesis extends the framework on risk-capturing prices presented by Bannor and Scherer to the Long-Term/Short-Term model. The impact of parameter uncertainty on the price of European call options on natural gas futures as well as the impact on the price of calendar spread options is studied. This approach allows to identify the parameters contributing the largest amount of valuation uncertainty and to quantify the valuation uncertainty caused by the respective parameter.


Valuation of a Fuel-Fired Power Plant in the Light of Emissions Trading in Alberta, Canada

Master Thesis in "Mathematical Finance", Oxford (2014)


To meet the reduction target of greenhouse gas emission fixed in the Kyoto Protocol in 1997, Alberta issued a series of new standards and regulations. One is the Environmental Protection and Enhancement Act – Emissions Trading Regulation (33/2006) from 2006, which established an emissions trading system for nitrogen oxides and sulphur dioxides. A baseline-and-credit system was introduced, in which a (annual) baseline emission is attributed to each generating unit. If a unit produces emissions less than the baseline, it can transform the unused amount into credits at the end of the year. On the other hand, those units which fail to stay within the limits (due to higher production rates or outdated pollution control technologies) are allowed to compensate their excess emissions by the purchase of credits. The trading is supervised by the central Emissions Trading Registry. Despite Canada’s withdrawal from the Kyoto Protocol in 2011, the Emissions Trading Regulation is still in place.

This new emissions trading system poses new constraints especially on fuelfired power plants. They are typically operated by producing a base load of electricity. In times of high demand when electricity prices rise, electricity generation is profitable and is therefore raised. So the amount of produced electricity is determined by the spark spread, the spread between the electricity and fuel price. With the newly installed emissions trading system additional costs or benefits from purchasing or selling emissions credits need to be considered in the operator’s production decisions. In particular this mechanism may provide additional incentives to lower electricity generation.

Using simulations of the market heat rate, which is the ratio of electricity and gas price, Cassano & Sick (2009) performed valuations of a gas-fired power plant in Alberta. Later, d-fine GmbH (2010) expanded this approach for considering a cap-and-trade emissions trading approach for the EU market.

This dissertation will focus on incorporating a baseline-and-credit mechanism into the valuation approach of Cassano & Sick (2009). Therefore, it is no longer possible to model the market heat rate directly, but to implement separate stochastic processes for electricity and gas prices comprising a modified market heat rate. Dynamic programing will be used for valuation of the power plant in combination with a Least Squares Monte Carlo approach to compute conditional expectations. The aim of this study is to analyse the impact of the baseline-and-credit scheme on the operation of fuel-fired power plants and examine whether the new system is capable of contributing to the targeted emissions reduction.

Prognose der Preisentwicklung und Risikobewertung für eine Anwendung aus dem Energieanlagebau 

Master Thesis in "Qantitative Finance", Frankfurt School of Finance & Management (2009)