Measuring private information in a specialist market

Research output: Contribution to journalArticlepeer-review


Since the reduced forms of the popular measures of asymmetric information in the price formation process are not nested within larger models we cannot evaluate their fit using standard statistical tools. Furthermore, pairwise correlations amongst the measures are small. We benchmark these measures cross-sectionally to realized specialist loss rates (using alternatively volume and number of trades) in the TORQ data. While five of the six measures are significantly correlated with this benchmark, this is only because they are correlated with the specialist participation rate. We infer that the measures do not measure private information in order flow, even in the setting for which they are designed.

Original languageEnglish (US)
Pages (from-to)92-119
Number of pages28
JournalJournal of Empirical Finance
StatePublished - Jan 1 2015


  • Measuring adverse selection
  • Realized specialist loss rate
  • Specialist market
  • Specialist participation rate

ASJC Scopus subject areas

  • Finance
  • Economics and Econometrics


Dive into the research topics of 'Measuring private information in a specialist market'. Together they form a unique fingerprint.

Cite this