Paper: arXiv 2404.15495

Authors: Marcin Wątorek, Paweł Szydło, Jarosław Kwapień, Stanisław Drożdż

Abstract

The non-fungible token (NFT) market emerges as a recent trading innovation leveraging blockchain technology, mirroring the dynamics of the cryptocurrency market. The current study is based on the capitalization changes and transaction volumes across a large number of token collections on the Ethereum platform. In order to deepen the understanding of the market dynamics, the collection-collection dependencies are examined by using the multivariate formalism of detrended correlation coefficient and correlation matrix. It appears that correlation strength is lower here than that observed in previously studied markets. Consequently, the eigenvalue spectra of the correlation matrix more closely follow the Marchenko-Pastur distribution, still, some departures indicating the existence of correlations remain. The comparison of results obtained from the correlation matrix built from the Pearson coefficients and, independently, from the detrended cross-correlation coefficients suggests that the global correlations in the NFT market arise from higher frequency fluctuations. Corresponding minimal spanning trees (MSTs) for capitalization variability exhibit a scale-free character while, for the number of transactions, they are somewhat more decentralized.

Complexity vs Empirical Score

  • Math Complexity: 7.5/10
  • Empirical Rigor: 8.5/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper employs advanced multivariate statistical techniques like detrended correlation coefficients, Random Matrix Theory (Marchenko-Pastur distribution), and Minimal Spanning Trees (MSTs), indicating high mathematical complexity. It is highly data-driven, utilizing a large set of real-world transaction data (90 collections over 500 days) and conducting rigorous statistical comparisons between Pearson and detrended coefficients.

Research Flowchart

  flowchart TD
  A["Research Goal:<br>Understand NFT Market Dynamics<br>& Collection Dependencies"] --> B["Data Source:<br>Ethereum NFT Collections<br>(Capitalization & Transaction Volumes)"]
  B --> C["Methodology 1:<br>Calculate Detrended<br>Correlation Coefficient"]
  B --> D["Methodology 2:<br>Construct Minimal<br>Spanning Trees MST"]
  C --> E["Compute Correlation Matrix &<br>Eigenvalue Spectra"]
  D --> F["Analyze Network Structure<br>Scale-free vs. Decentralized"]
  E --> G["Key Findings/Outcomes"]
  F --> G
  G --> H["Lower correlation strength<br>vs. traditional markets"]
  G --> I["Correlations driven by<br>high-frequency fluctuations"]
  G --> J["Marchenko-Pastur comparison<br>confirms limited correlations"]
  G --> K["Capitalization MST is<br>scale-free; Transaction MST<br>is decentralized"]

Evidence Card

What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card

Proposes

The paper examines collection‑collection dependencies in the NFT market using the multivariate formalism of detrended correlation coefficient and correlation matrix.

Source sentence
In order to deepen the understanding of the market dynamics, the collection-collection dependencies are examined by using the multivariate formalism of detrended correlation coefficient and correlation matrix.

Assumes 3

  • Each NFT collection must have on average at least two transactions involving at least one of its tokens to be included in the analysis. data
    Source sentence
    The collections were selected according to the criterion requiring that, on average, each collection had at least two transac-tions involving at least one of its tokens.
  • Capitalization increments are modeled as logarithmic differences of total capitalization expressed in USD. statistical
    Source sentence
    Each dataset was then converted into a bivariate time series representing (1) the collection capitalization logarithmic increments c∆t(t) = lnK(t +∆t)−lnK(t) expressed in USD, where K(t) is the to-tal capitalization at a moment t, and (2) the number N∆t of transactions in time unit ∆t.
  • Two sampling frequencies (1 h and 24 h) are sufficient to capture the dynamics of the most liquid and all collections respectively. model
    Source sentence
    Two sampling frequencies were chosen: ∆t = 1h, suitable for the 13 most liquid collections where less than 50% of hours had no trading activity (see the column %0∆t=1h in Tab. I), and ∆t = 24h, applicable to all col-lections.

Data

  • NFT collections tick‑by‑tick dataset · CryptoSlam! crypto-onchain public
    2022‑06‑09 to 2023‑10‑21 · 90 Ethereum NFT collections
    Source sentence
    DATA AND METHODS A. Data specification The dataset contains tick-by-tick data representing trans-actions (price and time) on 90 NFT collections from the Ethereum blockchain. It covers the period of T = 500 days that began on June 9, 2022 and ended on October 21, 2023.

Costs

Not applicable The paper reports no trading or execution result.

Finds 5

  • Correlation strength in the NFT market is lower than that observed in previously studied markets. statistical
    Source sentence
    It appears that correlation strength is lower here than that observed in previously studied markets.
  • The eigenvalue spectra of the correlation matrix more closely follow the Marchenko‑Pastur distribution, with some departures indicating existing correlations. statistical
    Source sentence
    Consequently, the eigenvalue spectra of the correlation matrix more closely follow the Marchenko-Pastur distribution, still, some departures indicating the existence of correlations remain.
  • Global correlations in the NFT market arise from higher‑frequency fluctuations, as shown by comparing Pearson and detrended cross‑correlation matrices. statistical
    Source sentence
    The comparison of results obtained from the correlation matrix built from the Pearson coefficients and, independently, from the detrended cross-correlation coefficients suggests that the global correlations in the NFT market arise from higher frequency fluctuations.
  • Minimal spanning trees for capitalization variability exhibit a scale‑free character, while those for transaction numbers are more decentralized. statistical
    Source sentence
    Corresponding minimal spanning trees (MSTs) for capitalization variability exhibit a scale-free character while, for the number of transactions, they are somewhat more decentralized.
  • The number of eigenvalues exceeds the Marchenko‑Pastur bounds, indicating statistically significant correlations, with the largest eigenvalue well outside the random bulk. statistical
    Source sentence
    In the Pearson-coefficient approach (upper panels of Fig. 6(a)), the number of eigenvalues exceeds the bounds of the Marchenko-Pastur (M-P) region, which contains the bulk of eigenvalues corresponding to random fluctuations. Particularly note-worthy is the largest eigenvalue λ1, which stands well outside this region.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 41 related papers). Coverage is partial, so this is neither replication nor refutation.

Cited by 5 papers in the archive (1 by independent authors)

Code

No repository by the authors found in the paper.

Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.

Cited References (78)

  1. Bitcoin: A peer-to-peer electronic cash system (2009) — link
  2. NFT Wash Trading : Quantifying suspicious behaviour in NFT markets (2022) — link
  3. Suspicious trading in nonfungible tokens ( NFTs ) (2024) — link
  4. A review of the key challenges of non-fungible tokens (2023) — link
  5. Beyond the bubble: Will NFTs and digital proof of ownership empower creative industry entrepreneurs? (2022) — link
  6. Decomposing cryptocurrency high-frequency price dynamics into recurring and noisy components — → 2306.17095
  7. Characteristics of price related fluctuations in non-fungible token (NFT) market — → 2310.19747
  8. CryptomSlam
and 70 more
  1. Physical approach to complex systems (2012) — link
  2. Random matrix approach to cross correlations in financial data (2002) — link
  3. The bulk of the stock market correlation matrix is not pure noise (2006) — link
  4. miladyMusk
  5. Detrended fluctuation analysis made flexible to detect range of cross-correlated fluctuations (2015)
  6. Random Matrices (2004)
  7. The Generalised Product Moment Distribution in Samples from a Normal Multivariate Population (1928)
  8. DISTRIBUTION OF EIGENVALUES FOR SOME SETS OF RANDOM MATRICES (1967) — link
  9. Financial return distributions: Past, present, and COVID-19 (2021)
  10. Multiscale characteristics of the emerging global cryptocurrency market (2021)
  11. Fast unfolding of communities in large networks (2008) — link
  12. A Systematic Review of Blockchain Applications (2022) — link
  13. Overview of Colored Coins (2012) — link
  14. quantmNFT
  15. Bubbles in Ethereum (2022) — link
  16. Understanding digital bubbles amidst the COVID-19 pandemic: Evidence from DeFi and NFTs (2022) — link
  17. Bubbles in NFT markets: correlated with cryptocurrencies or sentiment indexes? (2023)
  18. Don’t miss out on NFTs?! A sentiment-based analysis of the early NFT market (2023)
  19. Exploring the diffusion of digital fashion and influencers’ social roles in the Metaverse : an analysis of Twitter hashtag networks (2024)
  20. Exploiting Unfair Advantages: Investigating Opportunistic Trading in the NFT Market — → 2310.06844
  21. Competition of noise and collectivity in global cryptocurrency trading: R oute to a self-contained market (2020)
  22. Noise Dressing of Financial Correlation Matrices (1999) — link
  23. Universal and Nonuniversal Properties of Cross Correlations in Financial Time Series (1999) — link
  24. World currency exchange rate cross-correlations (2007)
  25. Scale free effects in world currency exchange network (2008)
  26. Correlation, hierarchies, and networks in financial markets (2010)
  27. Efficiency of communities and financial markets during the 2020 pandemic (2021)
  28. On financial market correlation structures and diversification benefits across and within equity sectors (2022)
  29. Cross-correlation dynamics and community structures of cryptocurrencies (2020)
  30. Volatility and returns connectedness in cryptocurrency markets: Insights from graph-based methods (2023) — link
  31. Collective correlations, dynamics, and behavioural inconsistencies of the cryptocurrency market over time (2022)
  32. Community detection in cryptocurrencies with potential applications to portfolio diversification (2023)
  33. Quantifying the dynamics of financial correlations (2001)
  34. Detecting correlations and triangular arbitrage opportunities in the Forex by means of multifractal detrended cross-correlations analysis (2019)
  35. Information and hierarchical structure in financial markets (1999) — link
  36. Hierarchical organization in complex networks (2003) — link
  37. Degree stability of a minimum spanning tree of price return and volatility (2003) — link
  38. Structural and topological phase transitions on the German Stock Exchange (2013) — link
  39. Collective behavior of cryptocurrency price changes (2018) — link
  40. Core–periphery organization of the cryptocurrency market inferred by the modularity operator (2020) — link
  41. Inferring interactions in multispecies communities: The cryptocurrency market case (2023)
  42. Statistical physics in foreign exchange currency and stock markets (2000)
  43. Empirical properties of asset returns: stylized facts and statistical issues (2001)
  44. Scaling of the distribution of fluctuations of financial market indices (1999)
  45. On the origin of power-law fluctuations in stock prices (2004)
  46. Note on regression and inheritance in the case of two parents (1895)
  47. Mosaic organization of DNA nucleotides (1994)
  48. Multifractal detrended fluctuation analysis of nonstationary time series (2002)
  49. Multifractal analysis of financial markets: A review (2019) — link
  50. DCCA cross-correlation coefficient: Quantifying level of cross-correlation (2011) — link
  51. Effect of detrending on multifractal characteristics (2013) — link
  52. Multifractality of self-affine fractals (1991)
  53. Quantitative features of multifractal subtleties in time series (2009) — link
  54. Genuine multifractality in time series is due to temporal correlations (2023)
  55. Multifractal cross-correlations between the world oil and other financial markets in 2012–2017 (2019)
  56. Bitcoin market route to maturity? Evidence from return fluctuations, temporal correlations and multiscaling effects (2018)
  57. Answering the Skeptics: Yes, Standard Volatility Models do Provide Accurate Forecasts (1998)
  58. Forecasting financial market volatility: Sample frequency vis-a-vis forecast horizon (1999)
  59. Intraday periodicity, long memory volatility, and macroeconomic announcement effects in the US Treasury bond market (2000)
  60. Exploiting Unfair Advantages: Investigating Opportunistic Trading in the NFT Market — → 2310.06844
  61. Detecting a currency’s dominance or dependence using foreign exchange network trees (2005)
  62. On the shortest spanning subtree of a graph and the traveling salesman problem (1956)
  63. Shortest Connection Networks And Some Generalizations (1957) — link
  64. Detecting and interpreting distortions in hierarchical organization of complex time series (2015) — link
  65. The components of empirical multifractality in financial returns (2009)
  66. Econophysics and sociophysics: Their milestones & challenges (2019) — link
  67. Quantitative approach to multifractality induced by correlations and broad distribution of data (2018) — link
  68. Fractal properties, information theory, and market efficiency — → 2306.13371
  69. Blockchain technologies and art: Opportunities and open challenges (2023) — link
  70. Analysis of inter-transaction time fluctuations in the cryptocurrency market (2022)

5 of 78 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.