# Recurrence analysis of the NASDAQ crash of April 2000

**Authors:** Annalisa Fabretti, Marcel Ausloos  
**arXiv:** [physics/0505170](https://arxiv.org/abs/physics/0505170)  
**Date:** 2005-05-24  
**Summary page:** https://www.summarizepaper.com/en/paper/physics--0505170v1/

## Key points

- Study by Annalisa Fabretti and Marcel Ausloos on NASDAQ crash of April 2000
- Application of Recurrence Plot (RP) and Recurrence Quantification Analysis (RQA) in analyzing financial indices
- RP and RQA adept at handling non-linear dynamical systems and non-stationarity, capturing changes in system states over time
- Capability of RP and RQA to detect critical regimes within financial indices before market bubble bursts
- Use of techniques on daily closing prices of NASDAQ index from January 1998 to November 2003 pinpointed initial time of NASDAQ bubble to be October 19, 1999
- Recurrence analysis as a tool for early detection and estimation of market bubbles
- Contribution to econophysics field by showcasing advanced analytical methods for understanding complex financial systems
- Results detailed in forthcoming publication "Practical Fruits of Econophysics" edited by H. Takayasu and published by Springer in 2006

## Summary

In their study titled "Recurrence analysis of the NASDAQ crash of April 2000," authors Annalisa Fabretti and Marcel Ausloos delve into the application of Recurrence Plot (RP) and Recurrence Quantification Analysis (RQA) in analyzing the dynamics of financial indices. Specifically focusing on the NASDAQ market, these signal numerical analysis methodologies are adept at handling non-linear dynamical systems and non-stationarity. They effectively capture changes in the states of a system over time. The researchers demonstrate that RP and RQA have the capability to detect critical regimes within financial indices, similar to a phase transition, before a market bubble bursts. By utilizing these techniques on daily closing prices of the NASDAQ index from January 1998 to November 2003, they were able to pinpoint the initial time of the NASDAQ bubble to be October 19, 1999. This finding highlights the potential of recurrence analysis as a tool for early detection and estimation of market bubbles. This research contributes to the field of econophysics by showcasing how advanced analytical methods can offer a deeper understanding of complex financial systems and their underlying dynamics. The study's results will be detailed in a forthcoming publication titled "Practical Fruits of Econophysics," edited by H. Takayasu and published by Springer in 2006. Overall, Fabretti and Ausloos' work underscores the importance of employing innovative approaches like recurrence analysis in studying financial markets. It offers a fresh perspective on predicting market fluctuations and identifying critical turning points with potential economic implications.