Decomposing cryptocurrency high-frequency price dynamics into recurring and noisy components

ArXiv ID: 2306.17095 “View on arXiv”

Authors: Unknown

Abstract

This paper investigates the temporal patterns of activity in the cryptocurrency market with a focus on Bitcoin, Ethereum, Dogecoin, and WINkLink from January 2020 to December 2022. Market activity measures - logarithmic returns, volume, and transaction number, sampled every 10 seconds, were divided into intraday and intraweek periods and then further decomposed into recurring and noise components via correlation matrix formalism. The key findings include the distinctive market behavior from traditional stock markets due to the nonexistence of trade opening and closing. This was manifest in three enhanced-activity phases aligning with Asian, European, and U.S. trading sessions. An intriguing pattern of activity surge in 15-minute intervals, particularly at full hours, was also noticed, implying the potential role of algorithmic trading. Most notably, recurring bursts of activity in bitcoin and ether were identified to coincide with the release times of significant U.S. macroeconomic reports such as Nonfarm payrolls, Consumer Price Index data, and Federal Reserve statements. The most correlated daily patterns of activity occurred in 2022, possibly reflecting the documented correlations with U.S. stock indices in the same period. Factors that are external to the inner market dynamics are found to be responsible for the repeatable components of the market dynamics, while the internal factors appear to be substantially random, which manifests itself in a good agreement between the empirical eigenvalue distributions in their bulk and the random matrix theory predictions expressed by the Marchenko-Pastur distribution. The findings reported support the growing integration of cryptocurrencies into the global financial markets.

Keywords: Cryptocurrency market microstructure, Temporal pattern analysis, Correlation matrix formalism, Random matrix theory, Cryptocurrency

Complexity vs Empirical Score

• Math Complexity: 6.5/10
• Empirical Rigor: 7.0/10
• Quadrant: Holy Grail
• Why: The paper employs random matrix theory (Marchenko-Pastur distribution) and correlation matrix formalism for decomposition, indicating significant mathematical density. It is backed by high-frequency 10-second data across three years for multiple cryptocurrencies, analyzing specific macro events and trading sessions, demonstrating strong data and implementation focus.

  flowchart TD
    A["Research Goal<br>Identify recurring vs. noise components<br>in crypto high-frequency dynamics"] --> B["Data & Inputs<br>Bitcoin, Ethereum, Dogecoin, WINkLink<br>Jan 2020 - Dec 2022<br>10-second sampling<br>Returns, Volume, Transaction #"]
    B --> C["Methodology: Correlation Matrix Formalism"]
    C --> D["Computational Process<br>1. Aggregate data into intraday/intraweek periods<br>2. Compute correlation matrices<br>3. Apply Random Matrix Theory<br>4. Extract eigenvalue distributions"]
    D --> E["Key Findings & Outcomes"]
    E --> F1["Distinct 24/7 Market Structure<br>3 Enhanced-activity phases (Asian, EU, US)"]
    E --> F2["Algorithmic Trading Signal<br>Activity surges at 15-min & full-hour intervals"]
    E --> F3["Macro-Driven Recurrence<br>Bursts aligned with U.S. economic reports<br>(Nonfarm Payrolls, CPI, FOMC)"]
    E --> F4["Internal Dynamics ≈ Random<br>Eigenvalue bulk matches Marchenko-Pastur distribution<br>External factors drive recurring components"]