A note of techniques that mitigate floating-point errors in PIN estimation

Wen Chyan Ke*, Hueiling Chen, Hsiou Wei William Lin

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

4 Citations (Scopus)

Abstract

This study aims at the estimation of the probability of informed trading (PIN), which may fail for stocks with high levels of trading activities due to a computer's floating-point exception (FPE). In this paper, we discuss two solutions of adopting scaled trade counts and reformulating the likelihood to estimate PIN for actively traded stocks. This study shows that, although scaled data mitigates the impact of the FPE, the effectiveness of scaled data, however, appears to underperform when users adopt the unsuitable expression of the likelihood function. In contrast, the remedy of reformulating the likelihood is more stable.

Original languageEnglish
Pages (from-to)458-464
Number of pages7
JournalFinance Research Letters
Volume31
DOIs
Publication statusPublished - 2019 Dec

Keywords

  • Floating-point exception
  • Maximum likelihood
  • PIN
  • Scaled trade counts

ASJC Scopus subject areas

  • Finance

Fingerprint

Dive into the research topics of 'A note of techniques that mitigate floating-point errors in PIN estimation'. Together they form a unique fingerprint.

Cite this