This paper investigates the incentives to manipulate sequential markets by strategically reneging on forward commitments. We first study the behavior of a dominant firm in a two-period model with demand uncertainty. Our results show that sequential markets may be a source of inefficiencies. We then use the model’s predictions to investigate occurrences of reneging on long-term commit- ments in Alberta’s electricity market. We develop a machine learning approach to evaluate manipulations. We find that a dominant supplier increased its rev- enues by $35 million during the winter of 2010-11, causing Alberta’s electricity procurement costs to increase by above $330 million (20%).