Langevin modelling of high-frequency Hang-Seng index data

Lei Han TANG*

*Corresponding author for this work

Research output: Contribution to journalConference articlepeer-review

2 Citations (Scopus)

Abstract

Accurate statistical characterization of financial time series, such as compound stock indices, foreign currency exchange rates, etc., is fundamental to investment risk management, pricing of derivative products and financial decision making. Traditionally, such data were analyzed and modeled from a purely statistics point of view, with little concern on the specifics of financial markets. Increasingly, however, attention has been paid to the underlying economic forces and the collective behavior of investors. Here we summarize a novel approach to the statistical modeling of a major stock index (the Hang Seng index). Based on mathematical results previously derived in the fluid turbulence literature, we show that a Langevin equation with a variable noise amplitude correctly reproduces the ubiquitous fat tails in the probability distribution of intra-day price moves. The form of the Langevin equation suggests that, despite the extremely complex nature of financial concerns and investment strategies at the individual's level, there exist simple universal rules governing the high-frequency price move in a stock market.

Original languageEnglish
Pages (from-to)272-277
Number of pages6
JournalPhysica A: Statistical Mechanics and its Applications
Volume324
Issue number1-2
DOIs
Publication statusPublished - 1 Jun 2003
EventProceedings of the International Econophysics Conference - Bali, Indonesia
Duration: 29 Aug 200231 Aug 2002

Scopus Subject Areas

  • Statistics and Probability
  • Condensed Matter Physics

User-Defined Keywords

  • Langevin equation
  • Non-Gaussian statistics
  • Time series analysis

Fingerprint

Dive into the research topics of 'Langevin modelling of high-frequency Hang-Seng index data'. Together they form a unique fingerprint.

Cite this