AI-powered identification of anomalies and manipulation in electricity markets

The world’s energy system is undergoing a major shift towards renewable energy sources. Renewable energy are known for their intermittency, non-dispatchability and unpredictability since they depend on weather conditions. The errors between forecasted energy production and actual production make electricity prices more volatile and unstable than other commodities. This market volatility presents opportunities for traders to generate profits through traditional trading methods or algorithmic strategies. Algorithmic trading increases liquidity, narrows the bid-ask spread, and reduces the market impact of large trades. However, algorithmic trading can be used intentionally or unintentionally to manipulate the market illegally. The high-frequency nature of some of these algorithms makes it difficult for regulators to detect and react to such manipulations, highlighting the need for more sophisticated models. The objective of this PhD research is to build a detection framework of illegal manipulations in electricity markets. The research is divided into the following studies.