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

Download:

Current browse context:

q-fin.CP

Change to browse by:

References & Citations

Bookmark

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

Quantitative Finance > Computational Finance

Title: Weak error estimates for rough volatility models

Abstract: We consider a class of stochastic processes with rough stochastic volatility, examples of which include the rough Bergomi and rough Stein-Stein model, that have gained considerable importance in quantitative finance. A basic question for such (non-Markovian) models concerns efficient numerical schemes. While strong rates are well understood (order $H$), we tackle here the intricate question of weak rates.
Our main result asserts that the weak rate, for a reasonably large class of test function, is essentially of order $\min \{ 3H+\tfrac12, 1 \}$ where $H \in (0,1/2]$ is the Hurst parameter of the fractional Brownian motion that underlies the rough volatility process. Interestingly, the phase transitation at $H=1/6$ is related to the correlation between the two driving factors, and thus gives additional meaning to a quantity already of central importance in stochastic volatility modelling. Our results are complemented by a lower bound which show that the obtained weak rate is indeed optimal.
Comments: 36 pages
Subjects: Computational Finance (q-fin.CP); Probability (math.PR)
MSC classes: 60L90 (Primary) 60G22, 91G20 (Secondary)
Cite as: arXiv:2212.01591 [q-fin.CP]
  (or arXiv:2212.01591v1 [q-fin.CP] for this version)

Submission history

From: Thomas Wagenhofer [view email]
[v1] Sat, 3 Dec 2022 10:46:43 GMT (40kb,D)

Link back to: arXiv, form interface, contact.