Disentangling diffusion from jumps

Research output: Contribution to journalArticlepeer-review

191 Scopus citations


Realistic models for financial asset prices used in portfolio choice, option pricing or risk management include both a continuous Brownian and a jump components. This paper studies our ability to distinguish one from the other. I find that, surprisingly, it is possible to perfectly disentangle Brownian noise from jumps. This is true even if, unlike the usual Poisson jumps, the jump process exhibits an infinite number of small jumps in any finite time interval, which ought to be harder to distinguish from Brownian noise, itself made up of many small moves.

Original languageEnglish (US)
Pages (from-to)487-528
Number of pages42
JournalJournal of Financial Economics
Issue number3
StatePublished - Dec 2004

All Science Journal Classification (ASJC) codes

  • Accounting
  • Finance
  • Economics and Econometrics
  • Strategy and Management


  • Cauchy jumps
  • Diffusion
  • Lévy process
  • Maximum likelihood
  • Poisson jumps


Dive into the research topics of 'Disentangling diffusion from jumps'. Together they form a unique fingerprint.

Cite this