B1194
Title: Parameter estimation in high-dimensional vine copula models
Authors: Jana Gauss - LMU Munich (Germany) [presenting]
Thomas Nagler - LMU Munich (Germany)
Abstract: In certain applications, the dimension of a vine copula model is large and grows with the sample size. This leads to the question under which conditions the parameters can be estimated via stepwise ML estimation. It is shown that the stepwise MLE is consistent and asymptotically normal under certain assumptions if the number of parameters diverges. The results can also be applied to the generalized method of moments and can be extended to penalized estimation.
