Dynamic Risk Management of Electricity Contracts with Contingent Portfolio Programming

Where a service provider faces substantial operational risk in meeting demand, as well as price risk in resourcing it from a wholesale market, conventional risk management optimization methods can be quite inefficient. As a leading example of this problem, we analyze the management of an electricity contracts portfolio. We develop a contingent optimization approach that accounts for the stochasticity of both electricity prices and loads, and which permits the specification of various Conditional Value-at-Risk requirements to optimize hedging across intermediate stages in the planning horizons. Our experimental results based on real data from Nordpool and Finland suggest that the modeling of price and load correlations is particularly important for this class of problems. We also show that contingent optimization yields higher expected return at a lower risk level than conventional approaches (e.g., periodic optimization or fixed allocation strategies). The sensitivity analysis is extended to characterize some aspects of behavior. Thus, we observe that a competitive player is more susceptible to price-related than load-related uncertainties and that a risk averse player is especially sensitive to the drivers of the forward risk premium.