发表期刊：Journal of Econometrics
发表时间：March 2026
作者及单位：Peter Reinhard Hansen, Chen Tong*(Department of Finance, School of Economics,Wang Yanan Institute for Studies in Economics, Xiamen University)
摘要：We introduce a family of multivariate heavy-tailed distributions, termed convolution-t distributions, constructed as convolutions of heterogeneous multivariate t-distributions. Unlike commonly used heavy-tailed distributions, this family captures nonlinear dependencies, accommodates heterogeneous marginal distributions, and reveals cluster structures prevalent in economic data. Importantly, convolution-t distributions admit simple closed-form densities that facilitate estimation and likelihood-based inference. The characteristic features of convolution-t distributions are shown to be important in an empirical analysis of realized volatility measures and help uncover their underlying factor structure.
关键词: Multivariate heavy-tailed distributions; Convolutions of t-distributions; Voigt profile
