Probabilistic Forecasting of German Electricity Imbalance Prices

The imbalance market is very volatile and often exhibits extreme price spikes. This makes it hard to model; however, if predicted correctly, one could make significant gains by participating on the right side of market. In this manuscript, we conduct a short-term probabilistic forecasting prices, contributing scarce literature in novel subject. performed 30 min before delivery, so that trader might still choose trading place. distribution prices modelled forecasted using methods well-known electricity literature: lasso with bootstrap, gamlss, neural networks. are compared naive benchmark meaningful rolling window study. results provide evidence efficiency between intraday balancing markets as sophisticated do not substantially overperform continuous index. On other hand, they significantly improve empirical coverage. Therefore, traders should avoid market, which inline objective current regulations analysis was conducted German market; be easily applied any similar structure.