• My UNC Charlotte

  • Directory

  • Campus Events

  • Library

  • Prospective Students

    • About UNC Charlotte
    • Campus Life
    • Admissions
    • Graduate Admissions
  • Faculty and Staff

    • Human Resources
    • Auxiliary Services
    • Inside UNC Charlotte
    • Academic Affairs
  • Current Students

    • Athletics
    • Financial Aid
    • Advising
    • Student Health Center
  • Alumni and Friends

    • Alumni Association
    • Advancement
    • Foundation
    • Make a Gift
Probability Seminar, Department of Mathematics & Statistics
Probability Seminar, Department of Mathematics & Statistics
  • My UNC Charlotte

  • Directory

  • Campus Events

  • Library

  • Prospective Students

    • About UNC Charlotte
    • Campus Life
    • Admissions
    • Graduate Admissions
  • Faculty and Staff

    • Human Resources
    • Auxiliary Services
    • Inside UNC Charlotte
    • Academic Affairs
  • Current Students

    • Athletics
    • Financial Aid
    • Advising
    • Student Health Center
  • Alumni and Friends

    • Alumni Association
    • Advancement
    • Foundation
    • Make a Gift
  • Past Talks

Contact Me

Michael Grabchak

Links

  • Dept Site

Wed Oct 18, 2023 at 4:00PM in Fretwell 315

October 15, 2023 by Michael Grabchak
Categories: Probability Seminar

Sébastien Bossu, UNC Charlotte

Title: Spanning multi-asset option payoffs with ReLUs and Radon transforms

Abstract: The Carr-Madan spanning formula which underlies the calculation of the VIX gives a decomposition of an arbitrary payoff function as a sum of calls and puts. This formula can be generalized to multi-asset option using Radon transforms, and the corresponding inverse problem can be approximated numerically using one hidden-layer feedforward ReLU neural networks (NN). In the first part of this talk, the NN approach is used to identify static hedges (“spanning portfolios”) for a selection of multi-asset options, such as best-of calls or dispersion puts, using basket payoffs that are easier to trade and price. In the second part of this talk, perfect static hedges based on (suitably regularized) Radon transform solutions are discussed. (Joint work with Stéphane Crépey, Univ. Paris-Cité, and his PhD students Nisrine Madhar and Hoang-Dung Nguyen, Univ. Paris-Cité and Natixis)

Bio: Sébastien Bossu received his Ph.D. in quantitative finance from Université Paris-Saclay in 2022 under the guidance of Peter Carr, Stéphane Crépey and Andrew Papanicolaou. He has published 3 articles in top academic journals as well as two textbooks and several industry papers in mathematical finance, including the popular read “Just What You Need to Know About Variance Swaps” which circulated widely. Sébastien previously served as part-time faculty at Fordham University, Pace University, Johns Hopkins Carey Business School, NYU Courant and Boston University Questrom School of Business, and most recently as full-time Visiting Professor of Finance at WPI Business School. He is a regular speaker at regional, national and international conferences in his field. Sébastien is also a graduate from The University of Chicago, HEC Paris, Columbia University and Sorbonne Université (fmr. Pierre-et-Marie-Curie).

Prior to his Ph.D., Sébastien was principal at his startup investment and consulting company in New York City since 2011, and worked at JPMorgan and Dresdner Kleinwort (now Commerzbank) in London between 2003 and 2008. As a young Analyst at JPMorgan, he discovered the correlation proxy formula and was the first author to establish the connection between dispersion trading and correlation swaps.

Click for more  

UNC Charlotte Homepage

Campus Links

  • Alerts
  • Jobs
  • Make a Gift
  • Maps / Directions
  • Accessibility

Resources

  • Alumni & Friends
  • Faculty & Staff
  • Prospective Students
  • Community
  • Current Students
  • Parents and Family

Stay In Touch

facebook instagram flickr linkedin twitter youtube maps

The University of North Carolina at Charlotte
9201 University City Blvd, Charlotte, NC 28223-0001
704-687-8622

© 2017 UNC Charlotte | All Rights Reserved
Contact Us | Terms of Use | University Policies
Skip to toolbar
  • Log In