Decoding Bitcoin's Volatility: A Random Walk Analysis and Bubble Detection Using the DS LPPLS Model

Articles in Press

Document Type : Original Article

Authors

Malihe heydari ¹

Mohammad Nabi Shahiki Tash ²

Mohammad Reza Ghasemi ³

¹ economy

² Economics Department, University of Sistan and Baluchestan

³ Faculty of Economics, Isfahan University, and Ashraf-Isfahani Building and Isfahan Payame Noor University

https://doi.org/10.22067/mfe.2025.91430.1488

Abstract

The Bitcoin market, characterized by its high volatility, consistently attracts the attention of investors and researchers. However, two primary challenges persist within this market: assessing its efficiency and identifying periods of price bubbles. These bubbles can lead to erroneous financial decisions. Despite extensive research, ambiguities surrounding non-random behavior and the formation of price bubbles in Bitcoin remain. This study investigates the efficiency of the Bitcoin market and identifies potential bubbles using the DS LPPLS (Discrete Singularity Log-Periodic Power Law Singularity) model, alongside confidence and sentiment indices. Daily closing prices of Bitcoin from the beginning of 2025 to the end of 2024 were utilized for this analysis. Through the examination of price charts and relevant indicators, it was determined that the Bitcoin market generally does not exhibit weak-form efficiency, and its behavior does not fully align with a purely random walk model. The confidence and sentiment indices, which reflect investor emotions towards the market, experienced significant fluctuations and showed a high correlation with price changes. Sudden and simultaneous increases in positive sentiment indicators and price suggest the formation of price bubbles during certain periods. The findings of this research indicate that the DS LPPLS model, in conjunction with confidence and sentiment indices, serves as a robust tool for analyzing Bitcoin market behavior and detecting price bubbles. Nevertheless, for more precise results, further in-depth investigation into factors influencing the market and the application of more complex models are warranted.

Keywords

Bitcoin

market efficiency

random walk model

bubble

DS LPPLS model

Subjects

• Financial Economics

---