Paper: arXiv 2404.00028

Authors: Peng Liu

Abstract

The correlation-based financial networks are studied intensively. However, previous studies ignored the importance of the anti-correlation. This paper is the first to consider the anti-correlation and positive correlation separately, and accordingly construct the weighted temporal anti-correlation and positive correlation networks among stocks listed in the Shanghai and Shenzhen stock exchanges. For both types of networks during the first 24 years of this century, fundamental topological measurements are analyzed systematically. This paper unveils some essential differences in these topological measurements between the anti-correlation and positive correlation networks. It also observes an asymmetry effect between the stock market decline and rise. The methodology proposed in this paper has the potential to reveal significant differences in the topological structure and dynamics of a complex financial system, stock behavior, investment portfolios, and risk management, offering insights that are not visible when all correlations are considered together. More importantly, this paper proposes a new direction for studying complex systems: the anti-correlation network. It is well worth reexamining previous relevant studies using this new methodology.

Complexity vs Empirical Score

  • Math Complexity: 7.5/10
  • Empirical Rigor: 6.0/10
  • Quadrant: Holy Grail — high math complexity, high empirical rigor

Why this score: The paper employs advanced network theory concepts (correlation matrices, topological measurements like assortativity and clustering) and formal definitions, indicating high math complexity. It uses a substantial 24-year dataset of China A-shares with robustness checks, but lacks explicit backtesting or trading strategy implementation, placing it in the high-math, high-rigor quadrant.

Research Flowchart

  flowchart TD
  A["Research Goal:<br>Isolate and analyze<br>Anti-correlation in<br>China A-share Market"] --> B["Data Input:<br>24 Years of Stock Price Data<br>Shanghai & Shenzhen Exchanges"]
  B --> C["Methodology: Separate Networks<br>Construct Weighted Temporal Networks:<br>Positive Correlation vs Anti-correlation"]
  C --> D["Computational Analysis:<br>Systematic Topological Measurements<br>(e.g., Degree Centrality, Clustering, Path Length)"]
  D --> E{"Key Findings & Outcomes"}
  E --> F["Structural Differences:<br>Essential distinctions between<br>Positive and Anti-correlation networks"]
  E --> G["Market Asymmetry:<br>Asymmetry effect between<br>market decline and rise"]
  E --> H["New Direction:<br>Proposed methodology for<br>Complex Systems & Risk Management"]

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 proposes a methodology that separates anti‑correlation and positive‑correlation to construct weighted temporal anti‑correlation and positive‑correlation networks among stocks.

Source sentence
This paper is the first to consider the anti-correlation and positive correlation separately, and accordingly construct the weighted temporal anti-correlation and positive correlation networks among stocks listed in the Shanghai and Shenzhen stock exchanges.

Assumes 1

  • Assume we have N stocks. data
    Source sentence
    Assume we have N stocks.

Data

  • Stocks listed on the Shanghai and Shenzhen stock exchanges market-prices unknown
    Jan 1 2000 to Dec 31 2023 · 5,329 stocks
    Source sentence
    To pay attention to the anti-correlation, this paper is the first to consider the anti-correlation and positive corre-lation separately, and accordingly construct the weighted anti-correlation and positive correlation networks among the 5,329 stocks listed in the Shanghai and Shenzhen stock exchanges. To investigate the temporal evolution characteristics of the networks, this paper uses the tech-nique of sliding

Costs

Not applicable The paper reports no trading or execution result.

Finds 5

  • Almost all anti‑correlation networks are scale‑free in terms of strength, while all positive‑correlation networks are not. statistical
    Source sentence
    From the above studies, this paper finds that almost all anti-correlation networks are scale-free in terms of strength, while all positive correlation networks are not.
  • Anti‑correlation networks exhibit significantly disassortative mixing, lower average local clustering coefficients, and larger average shortest path lengths compared to positive‑correlation networks. statistical
    Source sentence
    (2) The anti-correlation p-6 Anti-correlation network among China A-shares networks behave significantly disassortative mixing, while the assortativity coefficients of the positive correlation net-works are close to 0; the average local clustering coeffi-cients for the anti-correlation networks are significantly smaller than those for the positive correlation networks;
  • The probability of anti‑correlation (p(ρij<0)) reaches as high as 0.43, and 31% of time windows have p(ρij<0)>0.1. statistical
    Source sentence
    1 shows that the probability of anti-correlation p (ρij < 0) is as high as 0.43, and the time windows with p (ρij < 0) > 0.1 account for 31% of the total windows.
  • Tail shape parameters are positive and <1 for almost all anti‑correlation networks, whereas they are negative for all positive‑correlation networks. statistical
    Source sentence
    It shows that the tail shape param-eters for almost all anti-correlation networks are positive and smaller than 1, and those for all positive correlation networks are negative.
  • An asymmetry effect between stock‑market decline and rise is observed. other
    Source sentence
    This paper also observes an asym-metry effect between the stock market decline and rise.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 32 related papers). Coverage is partial, so this is neither replication nor refutation.

Cited by 1 paper in the archive

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 (38)

  1. Collective dynamics of `small-world’networks (1998) — link
  2. Network medicine: a network-based approach to human disease (2011) — link
  3. The multilayer nature of ecological networks (2017) — link
  4. The physics of financial networks (2021) — link
  5. Tail event driven networks of SIFIs (2019) — link
  6. Coupled propagation dynamics on complex networks: A brief review (2024) — link
  7. Mapping coupled time-series onto a complex network (2020) — link
  8. Explaining Financial Markets in Terms of Complex Systems (2014) — link
and 30 more
  1. Hierarchical structure in financial markets (1999) — link
  2. Topology of correlation-based minimal spanning trees in real and model markets (2003) — link
  3. Dynamic asset trees and Black Monday (2003) — link
  4. Dynamics of market correlations: Taxonomy and portfolio analysis (2003) — link
  5. Intraday volatility and network topological properties in the Korean stock market (2012) — link
  6. Structural and topological phase transitions on the German Stock Exchange (2013) — link
  7. Dynamic asset trees in the US stock market: Structure variation and market phenomena (2017) — link
  8. Dynamic topology and allometric scaling behavior on the Vietnamese stock market (2019) — link
  9. Dynamic network topology and market performance: A case of the Chinese stock market (2022) — link
  10. A tool for filtering information in complex systems (2005) — link
  11. Correlation based networks of equity returns sampled at different time horizons (2007) — link
  12. Community Analysis of Global Financial Markets (2016) — link
  13. Stock market as temporal network (2018) — link
  14. Clustering and information in correlation based financial networks (2004) — link
  15. Intensity and coherence of motifs in weighted complex networks (2005) — link
  16. A network analysis of the Chinese stock market (2009) — link
  17. Research on energy stock market associated network structure based on financial indicators (2018) — link
  18. Comparison between global financial crisis and local stock disaster on top of Chinese stock network (2018) — link
  19. Impacts of the COVID-19 pandemic on financial market connectedness (2021) — link
  20. The structure and resilience of financial market networks (2012) — link
  21. Network entropies of the Chinese financial market (2016) — link
  22. Research on the stock correlation networks and network entropies in the Chinese green financial market (2021) — link
  23. Research on the dynamic evolution and its influencing factors of stock correlation network in the Chinese new energy market (2022) — link
  24. Precision measurement of the return distribution property of the Chinese stock market index (2023) — link
  25. Node centrality in weighted networks: Generalizing degree and shortest paths (2010) — link
  26. Assortative Mixing in Networks (2002) — link
  27. Temporal and spatial evolution of the distribution related to the number of COVID-19 pandemic (2022) — link
  28. Heavy-tailed distributions of confirmed COVID-19 cases and deaths in spatiotemporal space (2023) — link
  29. Emergence of Scaling in Random Networks (1999) — link
  30. Error and attack tolerance of complex networks (2000) — link

0 of 38 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.