I will present an alternative electricity market architecture involving ``aggregators'' who will buy power from renewable energy generators, and sell it in the day-ahead market. However, in order to proper risk management, they must truthfully elicit the probability distribution of generation of the various generators. We present an auction-based mechanism that is ``incentive-compatible'', and can truthfully elicit any finite number of moments of the distributions. I will end by describing a distributed optimal control framework that can be used to determine dynamic pricing by the distributors to the consumers in a stochastic environment.