Dynamic Portfolio Selection Methods for Power Generation Assets

In this paper we start off by reviewing the literature on how to extend the meanvariance portfolio model to multi-stage portfolio problems. We then apply a multiperiod portfolio selection model to power generation assets, which is based on a reallocation methodology with scenario tree. Two solution approaches are used: the multi-period rebalancing model and the global solution one. These approaches are contrasted with the efficient frontier obtained for a “buy-and-hold policy”, thus helping to illustrate the effect of portfolio dynamization. The study covers all major electricity generation technologies in Germany and investigates the impact of offshore wind and solar power plants on existing power generation portfolios. We find that solar power technology has a positive impact on the efficiency of the portfolios, and the analysis underlines the advantages of using a multi-period rebalancing model for decision-making.