ABSTRACT
This chapter discusses advances in the modern portfolio theory in the context of generation self-scheduling. It explores approaches based on Value-at-Risk (VaR) and Conditional Value-at-Risk (CVaR) together with their extensions to deal with the problem of data uncertainty. In return-risk trade-off analysis, the risk is quantified by a risk measure that maps the loss to a monetary value. VaR is a risk assessment tool that is used by financial institutions to measure the minimum occasional loss expected in a given portfolio within a stated time period. CVaR is an alternative risk assessment tool that gives a better indication of risk than VaR; it quantifies the losses associated with the tail of the profit distribution. Self-scheduling formulations that attempt to balance risk and reward are based on maximizing the robust profit or conditional robust profit. VaR approaches to risk management suffer from several shortcomings.
