Inference on the Hurst parameter and the variance of diffusions driven by fractional Brownian motion (Q476364)

The Hurst exponent \(H\) (\(0<H<1\)) is called in honor of the hydrologist Harold Edwin Hurst who studied the fluctuations of the water level in the Nile river. The Hurst parameter \(H\) determines the smoothness of the trajectories of a fractional Brownian motion. \(H\) is closer to 1 for more regular trajectories. In these Lecture Notes, the authors develop estimation techniques for the Hurst parameter \(H\) and the volatility (called local variance by the authors). They use the observation of the process in a discrete mesh of points and study the asymptotic properties when the meshs distance tends to zero. For inference on \(H\) the authors use the second order increments of the process (instead of the first order increments) since variation built through them are asymptotically Gaussian in all the range \(0<H<1\). After estimating \(H\), the authors study volatility estimators in four pseudo-diffussion models. Particularly noteworthy is the simulation part of the book with a detailed description of the used algorithms. The book has eight chapters. Chapter 1 (Introduction) presents the necessary central limit theorems, Chapter 2 (Preliminaries) introduces fractional Brownian motion, stochastic integration and complex Wiener chaos. Chapter 3 contains the theoretical results on the estimators of \(H\) and the volatility and Chapter 4 the simulation studies. Chapter 5 presents all the proofs for Chapter 3. Complementary results can be found in the small Chapter 6. Tables and figures related to the simulation study are collected in Chapter 7. Chapter 8 presents some Pascal procedures used in the simulation part. The Lecture Notes are recommended for graduate courses in statistics of Gaussian processes and fractional models.