Empirical Regularities in Stock Market Crashes

From edegan.com
Jump to navigation Jump to search
Academic Paper
Title Empirical Regularities in Stock Market Crashes
Author Ed Egan
Status Working paper
© edegan.com, 2016

This paper develops results discovered in my analysis of the 2020 Stock Market Crash, which spawned a series of Op. Ed. submissions, as well as blog post. Version 1.5 was posted on SSRN (https://ssrn.com/abstract=3679630) on September 7, 2020. The first posted version was 1.2, which was posted on August 23, 2020.


From version 1.3:

I define a stock market crash as the period from an index's prior peak until its recovery. Then measures of its scale are very highly correlated. These correlations suggest that crashes belong to well-defined categories and become increasingly predictable as they progress. Furthermore, being in a crash is then the default state of U.S. stock markets.

Note that version 1.5 has a longer abstract.


The TeX files, pdfs, and general development files are in:


Key files include:

  • StockMarketAnalsisV1-5-SSRN.tex The source of the version posted at https://ssrn.com/abstract=3679630
  • CrashesV2.xlsx The main excel file for building tables and figures
  • Analysis.sql The SQL file for loading and processing the source data
  • Quick.do (and Time.do) The statistical analysis code

The data files are in:


The dbase is stockmarket.

Submission History

The paper was submitted as follows:

  • Economics Letters: Submitted on Aug 23, 2020 ($65 fee). Desk rejected by Joao F. Gomes on Aug 31. ("...while of some interest...")
  • Economics Bulletin: Submitted on Sept 7, 2020 (No fee). Under review at last check. http://www.accessecon.com/pubs/eb/

The next obvious choices are: