Detecting the Use of Undisclosed Privileged Information

Research output: Contribution to journalArticlepeer-review

Abstract

Enforcing regulators worldwide are mandated to discourage the use of privileged undisclosed information in security trading, as it may distort the true relative prices of securities and may lead to a market failure. This may be an impossible task, given the huge number of indications of such suspicious activity. Deviations of the bid-ask spreads from the optimal spreads provide real-time indications that some market participants may possess undis-closed privileged information. If we relied only on the real-time information that the bid-ask spreads deviate from the theoretical optimal spreads, our 3-years sample would generate over 11M deviations. This quantity of indications is unmanageable, both for the enforcing regulator, and for the uninformed market participants. In this study, we construct a mechanism that will enable market participants to significantly reduce the number of suspected cases to a manageable number. We establish an ex-post calibrated likelihood index that estimates the like-lihood that profit-oriented traders desire to exploit undisclosed privileged information that they possess. Our mecha-nism is based on the hypothesis that at any point in time when the likelihood index that predicts how likely traders are to exploit privileged information is positively correlated with an abnormal increase in the bid-ask spreads, then, probabilistically, these transactions should be suspected cases of illegal price-distorting trading.

Original languageEnglish
Pages (from-to)162-173
Number of pages12
JournalReview of Economics and Finance
Volume18
DOIs
StatePublished - 2020

Bibliographical note

Publisher Copyright:
© 2020 Better Advances Press. All rights reserved.

Keywords

  • Bid-ask spread
  • Insider trading
  • Privileged information

Fingerprint

Dive into the research topics of 'Detecting the Use of Undisclosed Privileged Information'. Together they form a unique fingerprint.

Cite this