Brattle senior associate Armando Levy presented a paper at the 25th Annual Center for Research Regulated Industries (CRRI) Western Conference on the topic of using a Markov Chain approach to forecast enrollments in demand response (DR) programs. In the presentation, Dr. Levy discussed how forecast models built on Markov Chains – a mathematical model that predicts the evolution of random events over time – could be used to model enrollments for a portfolio of DR programs that compete for the same customers. He suggests that the Markov Chain is a simple, flexible, and scalable approach to forecasting enrollment in a portfolio of demand response programs and that it will help to guide internally consistent forecasts in DR programs. Further, he finds it is invaluable in combining forecasts for 15 or more individual programs, even with many anticipated policy changes in the future. Dr. Levy’s presentation, co-authored by Charlie Gibbons, Jenny Palmer, and Joe Wharton of The Brattle Group, can be downloaded using the link below.

Armando Levy|Charlie Gibbons|Jenny Palmer|Joe Wharton