Paper: arXiv 2304.09939
Authors: Jevgeni Tarassov, Nicolas Houlié
Abstract
In this study, we investigate the BTC price time-series (17 August 2010-27 June 2021) and show that the 2017 pricing episode is not unique. We describe at least ten new events, which occurred since 2010-2011 and span more than five orders of price magnitudes ($US 1-$US 60k). We find that those events have a similar duration of approx. 50-100 days. Although we are not able to predict times of a price peak, we however succeed to approximate the BTC price evolution using a function that is similar to a Fibonacci sequence. Finally, we complete a comparison with other types of financial instruments (equities, currencies, gold) which suggests that BTC may be classified as an illiquid asset.
Complexity vs Empirical Score
- Math Complexity: 3.0/10
- Empirical Rigor: 5.5/10
- Quadrant: Street Traders — practical and empirical, lighter on theory
Why this score: The paper provides an empirical exploratory analysis of Bitcoin price cycles, comparing them to traditional assets and utilizing heuristic pattern fitting like Fibonacci-like sequences. While visually and empirically interesting across a long time series, it lacks advanced mathematical modeling or rigorous statistical testing for predictive power.
Research Flowchart
flowchart TD
A[Research Goal: Analyze BTC Price Time-Series & Bubble Dynamics] --> B[(Data: BTC Prices & Asset Data<br>Aug 2010 - Jun 2021)]
B --> C[Time-Series & Historical Analysis]
C --> D[Identify 10+ New Price Events<br>50-100 Day Duration, $1-$60k]
C --> E[Model Price Evolution via<br>Fibonacci-like Sequence]
C --> F[Comparative Asset Analysis<br>Equities, Currencies, Gold]
D & E & F --> G[Outcome: BTC Classified as<br>Illiquid Asset with Recurring Dynamics]