Forecasting volatility is one of the major challenges in the field of finance. Calvet and Fisher present a powerful, new technique for volatility modeling. Drawing on insights from the use of multifractals in the natural sciences and mathematics, they show how to construct regime-switching models that contain multiple durations, are easy to estimate, and outperform some of the best traditional forecasting models such as GARCH. Their preliminary work has been well-received in the top academic journals and this is the first time they present their research in a comprehensive way. The book is suitable for a PhD courses in Economics/ Finance, or a short course for practitioners on volatility modeling.