Random matrix approach to the dynamics of stock inventory variations

Wei Xing Zhou*, Guo Hua Mu, János Kertész

*Corresponding author for this work

    Research output: Contribution to journalArticleScientificpeer-review

    13 Citations (Scopus)
    165 Downloads (Pure)


    It is well accepted that investors can be classified into groups owing to distinct trading strategies, which forms the basic assumption of many agent-based models for financial markets when agents are not zero-intelligent. However, empirical tests of these assumptions are still very rare due to the lack of order flow data. Here we adopt the order flow data of Chinese stocks to tackle this problem by investigating the dynamics of inventory variations for individual and institutional investors that contain rich information about the trading behavior of investors and have a crucial influence on price fluctuations. We find that the distributions of cross-correlation coefficient C i j have power-law forms in the bulk that are followed by exponential tails, and there are more positive coefficients than negative ones. In addition, it is more likely that two individuals or two institutions have a stronger inventory variation correlation than one individual and one institution. We find that the largest and the second largest eigenvalues (λ 1 and λ 2) of the correlation matrix cannot be explained by random matrix theory and the projections of investors' inventory variations on the first eigenvector u(λ 1) are linearly correlated with stock returns, where individual investors play a dominating role. The investors are classified into three categories based on the cross-correlation coefficients C V R between inventory variations and stock returns. A strong Granger causality is unveiled from stock returns to inventory variations, which means that a large proportion of individuals hold the reversing trading strategy and a small part of individuals hold the trending strategy. Our empirical findings have scientific significance in the understanding of investors' trading behavior and in the construction of agent-based models for emerging stock markets.

    Original languageEnglish
    Article number093025
    JournalNew Journal of Physics
    Publication statusPublished - Sept 2012
    MoE publication typeA1 Journal article-refereed


    Dive into the research topics of 'Random matrix approach to the dynamics of stock inventory variations'. Together they form a unique fingerprint.

    Cite this