In progress Designing Smart Electricity Markets Using Data-driven Approaches



The past two decades have seen an explosion of digital data in every sector of the global economy. Such a revolution from data scarcity to data abundance brings great opportunities for decision makers in every area. The increased accessibility of data has enabled a different way of making decisions that involves more empirical evidence rather than personal experience, intuition, or belief. In this research, we focus on the value of data-driven decision making at market level and seek to illustrate and quantify its benefits in the design and operationalization of electricity markets. Electricity day ahead auctions play a key role in the decision making process of power agents. Over the past decades, Information Systems researchers have made significant contributions to practical auction design by investigating different bidding strategies and price dynamics in real-world auctions. This PhD Project will contribute to this growing literature on design of auctions, trading strategies and strategic bidding in auctions. Our first study aims to understand agent-level determinants of price expectation formation in electricity day ahead auctions. More specifically, we answer the following question: How does an agent’s trading behaviour, market informedness and risk aversion effect his price expectation bias? In the second study we focus on tactical decision making for trading between electricity day-ahead and real-time auctions through modelling the risk premium. The third study explores heterogeneous bidding strategies in electricity day-ahead auctions using a unique and big data set. To address these research questions, we adopt a multi-method approach which consists of survey design, econometric time series models and data-mining methodologies.


Smart Markets, Energy Markets, Market Efficiency, Data-Driven Decision Making, Electricity Day-Ahead Auctions, Agents, Auction Design, Bidding Strategy, Bidder Taxonomy, Real-time Regulating Markets, Regulating Premium, Trading Inefficiency, Rational Expectations Hypothesis, Market Informedness, Risk Aversion, Hedging, Forecast Accuracy, Price Forecasting

Time frame

2016 - 2019

Supervisory Team

Wolfgang Ketter
Professor of Next Generation Information Systems
  • Promotor
Eric van Heck
Professor of Information Management and Markets
  • Promotor
Derek Bunn
Derek Bunn
  • Copromotor