We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Quantitative Finance > Computational Finance

Title: Variational Autoencoders: A Hands-Off Approach to Volatility

Abstract: A volatility surface is an important tool for pricing and hedging derivatives. The surface shows the volatility that is implied by the market price of an option on an asset as a function of the option's strike price and maturity. Often, market data is incomplete and it is necessary to estimate missing points on partially observed surfaces. In this paper, we show how variational autoencoders can be used for this task. The first step is to derive latent variables that can be used to construct synthetic volatility surfaces that are indistinguishable from those observed historically. The second step is to determine the synthetic surface generated by our latent variables that fits available data as closely as possible. As a dividend of our first step, the synthetic surfaces produced can also be used in stress testing, in market simulators for developing quantitative investment strategies, and for the valuation of exotic options. We illustrate our procedure and demonstrate its power using foreign exchange market data.
Subjects: Computational Finance (q-fin.CP); Mathematical Finance (q-fin.MF)
Cite as: arXiv:2102.03945 [q-fin.CP]
  (or arXiv:2102.03945v1 [q-fin.CP] for this version)

Submission history

From: Nicholas Fung [view email]
[v1] Sun, 7 Feb 2021 23:31:33 GMT (1304kb,D)

Link back to: arXiv, form interface, contact.