Information Propagation Across Investor Types: Transfer Entropy Networks in the Korean Equity Market

Abstract

Whether heterogeneous investor flows transmit private information across stocks or merely reflect coordinated responses to public signals remains an open question in market microstructure. We construct Transfer Entropy (TE) networks from investor-type flows -- foreign, institutional, and individual -- for \numNStocks{} Korean equities over \numNDates{} trading days (January 2020 to February 2025), and evaluate their economic content through interaction information (II), conditional TE, mutual information (MI), Kelly criterion bounds, and Fama-MacBeth regressions. Three findings emerge. First, TE networks are sparse and structurally heterogeneous: foreign investors maintain few but strong links (\numEdgesFor{} edges, mean TE = \numMeanTEFor{}), while individual investors form many but weak links (\numEdgesInd{} edges, mean TE = \numMeanTEInd{}). Second, cross-investor information is redundant rather than synergistic, no investor type directionally dominates another, and MI between signals and returns is zero at the daily horizon. Third, network centrality adds negligible alpha in cross-sectional regressions, with only one of six signal-centrality interactions reaching marginal significance. These results indicate that the observed propagation structure captures shared information processing rather than private signal cascades, consistent with daily-frequency market efficiency.

Figures (10)

• Figure 1: Directed Transfer Entropy network graphs for foreign (left), institutional (center), and individual (right) investor types. Nodes represent the 100 stocks; edges indicate statistically significant TE at FDR-corrected $\alpha = 0.05$ using 200 block-permutation surrogates. Edge width is proportional to TE magnitude. Foreign networks exhibit few but strong connections, while individual networks are denser but weaker.
• Figure 2: Pairwise Transfer Entropy heatmaps for each investor type. Rows represent source stocks and columns represent target stocks. Color intensity indicates TE magnitude. The sparse structure confirms that significant information propagation is confined to a small subset of stock pairs across all investor types.
• Figure 3: Distributions of network centrality measures (out-degree, in-degree, betweenness, closeness) across stocks for each investor type. All distributions are right-skewed, with the foreign network exhibiting the highest concentration of centrality among a small number of hub stocks.
• Figure 4: Distribution of per-stock interaction information (II) values for each cross-investor pair. Negative values indicate redundancy; positive values indicate synergy. The distributions are centered slightly below zero, with nearly all stocks exhibiting redundant cross-investor information about returns.
• Figure 5: Conditional Transfer Entropy (CTE) and directionality index for each investor-pair comparison. Points show the estimated directionality index $D$; error bars show 95% bootstrap confidence intervals. All estimates are zero, indicating no directional information dominance among investor types.