Reduced Order Representation of Implied Volatility Surfaces

Speaker: Andrew Papanicolaou, NYU Tandon

Location: Courant Institute of Mathematics, 251 Mercer St, Warren Weaver Hall 1302, New York, NY 10012

Date: Tuesday, April 30, 2019, 5:30 p.m.

Synopsis:

We consider a Principal Component Analysis of implied vol surfaces (IVS) for US equities using data from OptionMetrics, available through Wharton Research Data Services (WRDS). Using higher-order decomposition methods for tensor data, we are able to address a number of important questions about data of this nature. In particular, we find the IVS to contain a non-trivial tensor structure, and that tensor methods such as multi-linear SVD and canonical polyadic decomposition have the ability to extract an orthogonal portfolio among delta-neutral option positions.