Research areas

Financial econometrics, return predictability, mixed-frequency / MIDAS regression methods (developed jointly with Eric Ghysels and Pedro Santa-Clara), real estate predictability, long-horizon return regressions and overlapping-data inference. Author of the well-known Valkanov (2003) result that long-horizon overlapping regressions can produce inflated t-statistics due to severe serial correlation in the residuals. UCSD Rady School of Management.

Key papers

Recent work

(to be populated by future ingests — known active areas: MIDAS extensions, real estate predictability, factor models)

Collaborators

  • eric-ghysels — long-time co-author on MIDAS, mixed-frequency econometrics, and the Handbook chapter.
  • alberto-plazzi — Handbook chapter co-author and long-time CRE research collaborator.
  • walter-torous — Handbook chapter co-author and long-time CRE research collaborator.

My notes

The Plazzi-Torous-Valkanov (2010) paper on commercial real estate predictability — exploiting AR(1) structure on unobserved expected returns and rent growth — is a methodological precursor to the state-space approach used in the CRE asset pricing project, although the project takes the next step of embedding the dynamics in a fully no-arbitrage NK macro framework with regime switching.