Related Experiment Video
Updated: Jul 11, 2026
05:31
Imaging and Quantification of the Area of Fast-Moving Microbubbles Using a High-Speed Camera and Image Analysis
Published on: September 5, 2020
Dissection of Bitcoin's multiscale bubble history from January 2012 to February 2018
J C Gerlach1, G Demos1, D Sornette1,2
1Department of Management, Technology and Economics ETH Zürich, Zürich, Switzerland.
Royal Society Open Science
|August 17, 2019
Summary
This study analyzes Bitcoin price bubbles from 2012-2018, identifying 13 peaks and using the Log-Periodic Power-Law Singularity (LPPLS) model to predict crash risks. The findings offer insights into potential future Bitcoin market crashes.
Area of Science:
- Quantitative finance
- Computational economics
- Time series analysis
Background:
- Bitcoin price dynamics exhibit complex patterns, including speculative bubbles.
- Understanding these bubbles is crucial for assessing market risks and predicting crashes.
Purpose of the Study:
- To conduct a detailed bubble analysis of Bitcoin price dynamics (2012-2018).
- To identify and classify market peaks (bubbles) and predict potential crash risks.
- To investigate the socio-economic drivers behind Bitcoin price movements.
Main Methods:
- Developed a robust automatic peak detection method for price time series.
- Applied the Lagrange Regularization Method to identify market regime shifts.
- Utilized the Log-Periodic Power-Law Singularity (LPPLS) model and LPPLS Confidence Indicators for risk analysis.
- Employed a clustering method to group predicted bubble end times for scenario generation.
Main Results:
- Identified three major and 10 smaller Bitcoin price peaks (bubbles) between 2012 and 2018.
- Quantified risks associated with long and short bubbles using the LPPLS model.
- Generated plausible scenarios for Bitcoin price evolution based on predicted crash times.
- The predictive scheme successfully provided warnings of imminent crash risks.
Conclusions:
- The study provides a robust framework for analyzing and predicting cryptocurrency bubbles.
- The LPPLS model and associated confidence indicators are effective tools for assessing market risk.
- The findings offer valuable insights for investors and regulators concerned with Bitcoin market stability.
