Blockholder networks, information exchange and M&A performance

The challenge of measuring information exchange

Institutional investors are the majority owners of many listed firms around the world, with ownership reaching levels as high as 70% in economies such as the United States (Dasgupta et al, 2021). Anecdotal evidence suggests that information exchange occurs among a firm’s largest shareholders. Yet research on how these investors exchange information through their co-shareholding networks remains scarce. The fundamental empirical challenge is that private information flows are unobservable to researchers.

To address this, we proxy for blockholder information exchange using more than 1.3 billion trip records from the New York City Taxi and Limousine Commission. By matching taxi pickup and drop-off coordinates with the headquarters addresses of Manhattan-based institutional investors, we identify rides between pairs of blockholders and test whether those connected through joint shareholdings interact more frequently.

Evidence from taxi rides

We define 2 investors as connected if both hold at least 5% of the same firm’s market capitalisation. Consistent with information exchange occurring through co-shareholding networks, connected blockholder pairs exhibit more than 20% higher taxi ride frequency between their headquarters compared to non-connected pairs.

This result holds in Poisson regressions controlling for investor characteristics, portfolio concentration, and the number of co-located investors sharing the same building. The relationship is driven by contemporaneous blockholding ties rather than historical connections, and the pattern is concentrated during business hours and weekdays, consistent with professional rather than social interactions. Since we compare ride frequency within the same investor pair with and without a blockholding connection, differences in investor characteristics cannot explain the results.

Network centrality and acquisition performance

Having established that co-shareholdings facilitate information exchange, we examine whether this mechanism affects corporate decisions. We focus on mergers and acquisitions because these transactions are inherently plagued by information asymmetries, making access to private information particularly valuable (Capron and Shen, 2007).

Using co-blockholdings to establish investor ties, we construct the broader network of US institutional blockholders and compute centrality measures that capture both direct and indirect connections (Crane et al, 2019). Eigenvector centrality, our primary measure, places greater weight on connections to well-connected investors, reflecting the intuition that information access depends on affiliates’ own network positions.

Analysing 15,223 acquisitions by US public firms over the 1980-2022 period, we find that acquirers held by more centrally positioned institutional investors earn significantly higher cumulative abnormal announcement returns. A one-standard-deviation increase in eigenvector centrality is associated with announcement returns approximately 15% higher relative to the sample mean. This effect persists after controlling for clique ownership, common ownership and a comprehensive set of deal and firm characteristics.

The role of information opacity

If network centrality captures information advantages, the effect should be strongest when private information is most valuable. We test this prediction by exploiting heterogeneity across target, deal and investor characteristics.

Consistent with an information channel, the positive relationship between centrality and deal quality is concentrated in acquisitions of private targets, small targets and targets with low analyst coverage or low annual report readability. For transparent public targets with extensive analyst following, network centrality provides no significant benefit. Chow tests confirm that coefficients differ significantly across these subgroups.

We also find that acquirers with more central shareholders are more likely to pay with cash, face less bidding competition, engage in rumoured rather than announced deals, and complete transactions faster. These patterns reinforce the interpretation that network centrality confers information advantages in target identification and valuation.

Advisory versus monitoring channel

Central investors could enhance deal quality through 2 channels: preventing value-destroying acquisitions (monitoring) or identifying high-quality targets that other acquirers miss (advisory). To distinguish between these, we examine whether centrality predicts the most successful deals, the least successful deals, or both.

We find that network centrality significantly increases the probability of deals in the top performance quartile and decile but shows no relationship with bottom-tercile or bottom-quartile deals. This asymmetry supports an advisory channel where central investors help portfolio firms identify ‘hidden gems’ rather than a monitoring channel focused on preventing empire-building.

Which investors matter?

The effect of network centrality on deal quality should be stronger among investors with a comparative advantage in exploiting information (Chen et al, 2007). Consistent with this, we find that the positive effects are driven entirely by long-term and independent investors, such as university endowments and independent investment advisors (Gaspar et al, 2005; Harford et al, 2018). Centrality among short-term traders and grey investors (banks and insurance companies with business relationships) shows no positive effect.

Identification

To address endogeneity concerns, we exploit plausibly exogenous variation in network centrality arising from mergers between financial institutions. When 2 institutional investors merge, the combined entity inherits the network positions of both predecessors, creating discrete shifts in centrality for their portfolio firms that are unrelated to firm-specific characteristics or acquisition strategies.

We implement a stacked difference-in-differences design around these financial institution mergers. Firms with higher pre-merger ownership by the acquired institution experienced increases in centrality post-merger due to the acquiring institution’s more central network position. Our estimates confirm that treated firms exhibited superior acquisition performance following the merger, supporting a causal interpretation.

Conclusion

Our findings suggest that co-shareholding networks facilitate information flows that translate into superior advice and enhanced M&A performance. Unlike studies of social networks among executives and directors, which can foster entrenchment, the network of financial co-shareholdings appears to provide independent advice that benefits portfolio firms. These results highlight an underappreciated channel through which institutional ownership concentration may improve corporate governance.