|
|||||
|
|
Moment-Implied Densities: Properties and Applications |
|
| Authors: | Eric Ghysels Fangfang Wang |
| Institution: | 1. Department of Economics, University of North Carolina at Chapel Hill and Department of Finance, Kenan-Flagler Business School at UNC, Chapel Hill, NC 27599 (eghysels@unc.edu);2. Department of Information and Decision Sciences, University of Illinois at Chicago, Chicago, IL 60607 (ffwang@uic.edu) |
| Abstract: | Suppose one uses a parametric density function based on the first four (conditional) moments to model risk. There are quite a few densities to choose from and depending on which is selected, one implicitly assumes very different tail behavior and very different feasible skewness/kurtosis combinations. Surprisingly, there is no systematic analysis of the tradeoff one faces. It is the purpose of the article to address this. We focus on the tail behavior and the range of skewness and kurtosis as these are key for common applications such as risk management. |
| Keywords: | Affine jump-diffusion model Feasible domain Generalized skewed t distribution Normal inverse Gaussian Risk neutral measure Tail behavior Variance Gamma |