Rethinking reversals

Timothy C. Johnson

Research output: Contribution to journalArticlepeer-review


High-frequency reversals are an economically important characteristic of the returns to tradeable claims to the market portfolio. This paper demonstrates that short-horizon negative autocorrelation can arise in a tractable model of agents with tournament-type preferences. Intuitively, investors act as if they are averse to missing out on a trend, causing the risk premium to move strongly counter to realized returns. The model features fully rationalizing agents, complete markets, and no exogenous transaction demand. Plausible parameterizations can match the autocorrelation in the data. Supporting evidence on novel first and second moment implications is presented.

Original languageEnglish (US)
Pages (from-to)211-228
Number of pages18
JournalJournal of Financial Economics
Issue number2
StatePublished - May 1 2016


  • Disagreement
  • G11
  • G12
  • G23
  • Peer effects
  • Stock market autocorrelation

ASJC Scopus subject areas

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


Dive into the research topics of 'Rethinking reversals'. Together they form a unique fingerprint.

Cite this