Scholarlink

<p>This study examines the clustering behavior and market efficiency of China’s Growth Enterprise Market Index (GEMI) and STAR 50 Index (STARI) using a Neyman–Scott point process framework. Index movements are modeled as spatial point patterns, capturing second-order dependence (correlation-stationarity) through cluster-based structures, with parameters estimated via Palm likelihood. The inhomogeneous L-function and permutation tests are employed to detect clustering beyond first-order heterogeneity. A sequence of nested models was constructed to isolate baseline, trend, and stochastic effects, with macroeconomic and market covariates incorporated through the trend component and point interactions captured via multiple kernels. Results reveal clear deviations from complete spatial randomness: first-order heterogeneity and second-order dependence coexist, indicating structured and interdependent trading behavior. A critical transition point is identified at <InlineEquation ID="IEq1"> <InlineMediaObject> <ImageObject Color="BlackWhite" FileRef="10614_2025_11165_Article_IEq1.gif" Format="GIF" Height="19" Rendition="HTML" Resolution="72" Type="Linedraw" Width="220" /> </InlineMediaObject> <EquationSource Format="TEX">\((\text {GEMI}, \text {STARI}) = (2670, 1261)\)</EquationSource> </InlineEquation>, beyond which the market is more susceptible to regime shifts. This threshold, together with a strong baseline intensity, suggests self-organized criticality and nonlinear feedback mechanisms. Clustering intensity is further influenced by cross-market interactions, including the SZSE Component Index, STAR Market trading volume, and inter-exchange linkages, reflecting systemic interdependence and partial segmentation between the two markets. Overall, short-range clustering and multi-center aggregation persist beyond first-order heterogeneity, challenging the weak-form Efficient Market Hypothesis. The significance of covariates highlights the role of cross-market dynamics and behavioral amplification in shaping index evolution. The spatial point process framework provides a unified statistical foundation for analyzing spatial and temporal dependencies in financial markets, offering insights into the microstructure and bounded efficiency of emerging segments.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Trading Clusters across China’s Growth Enterprise and Science-Technology Innovation Markets: A Neyman–Scott Point Process Framework

  • 期刊论文

  • 发表日期：

    2025年11月19日

查看全文
  • Quan Long,
  • Haoyun Yan

摘要

This study examines the clustering behavior and market efficiency of China’s Growth Enterprise Market Index (GEMI) and STAR 50 Index (STARI) using a Neyman–Scott point process framework. Index movements are modeled as spatial point patterns, capturing second-order dependence (correlation-stationarity) through cluster-based structures, with parameters estimated via Palm likelihood. The inhomogeneous L-function and permutation tests are employed to detect clustering beyond first-order heterogeneity. A sequence of nested models was constructed to isolate baseline, trend, and stochastic effects, with macroeconomic and market covariates incorporated through the trend component and point interactions captured via multiple kernels. Results reveal clear deviations from complete spatial randomness: first-order heterogeneity and second-order dependence coexist, indicating structured and interdependent trading behavior. A critical transition point is identified at \((\text {GEMI}, \text {STARI}) = (2670, 1261)\) , beyond which the market is more susceptible to regime shifts. This threshold, together with a strong baseline intensity, suggests self-organized criticality and nonlinear feedback mechanisms. Clustering intensity is further influenced by cross-market interactions, including the SZSE Component Index, STAR Market trading volume, and inter-exchange linkages, reflecting systemic interdependence and partial segmentation between the two markets. Overall, short-range clustering and multi-center aggregation persist beyond first-order heterogeneity, challenging the weak-form Efficient Market Hypothesis. The significance of covariates highlights the role of cross-market dynamics and behavioral amplification in shaping index evolution. The spatial point process framework provides a unified statistical foundation for analyzing spatial and temporal dependencies in financial markets, offering insights into the microstructure and bounded efficiency of emerging segments.