Project Details
Description
Do firms learn from other firms? Using the SEC EDGAR server logs in May of 2013, we document substantial cross-firm viewing activities. Through IP WHOIS lookup and name matching, we identify 1,962 unique nonfinancial public viewing firms with 197,764 searches in total. A typical viewing firm views 8 other firms, initiates 18 searches, and downloads 12 di↵erent fillings. We also discover significant heterogeneity in the sources of information acquisition. Firms view their peer firms, suppliers, customers, banks, and holding institutions. Across all viewings, 24.5% are peer-firm viewings, 11.7% are searching for financial firms, and the remaining are for other economically related firms.
Given the scope and the scale of cross-firm viewing that we document in our pilot study, we propose to process all logs and construct directed cross-firm information networks based on viewing links, di↵erentiating information outflow versus information inflows. The benefit of the network approach is that we can study the impact of information flow via higher-order connections (i.e., information flow from i to j through several direct viewing links). We then study the e↵ect of information flow on corporate decisions and asset pricing. Specifically, we ask three questions: First, do cross-firm viewings improve a firm’s performance and possibly stock returns. Second, does learning contribute to correlated corporate decisions? Third, does information noises also spread through cross- firm viewing links? Do we observe the propagation of idiosyncratic cash flow shocks? If so, do the central firms in the information outflow network bear higher market risk and earn higher expected returns?
The cross-firm viewings in our sample do not seem to be random. Instead, managers appear to seek relevant information from firms in di↵erent sectors. As such, we expect both the search intensity and search heterogeneity to contribute positively to corporate decision making and improves firm performance and equity return. Intuitively, Information flow should lead to correlated corporate decisions. Moreover, when learning is imperfect, it can also lead to noise and correlated idiosyncratic cash flows. Whether this shows up in the data has crucial implications for asset pricing since firm-specific risk of firms with high information outflow centrality would be more associated with market risk.
If established, the evidence would provide insight into corporate decision-making and determinants of a firm’s performance. It can also help us understand how idiosyncratic shocks lead to aggregate fluctuations and how information outflow centrality drives a firm’s beta.
Given the scope and the scale of cross-firm viewing that we document in our pilot study, we propose to process all logs and construct directed cross-firm information networks based on viewing links, di↵erentiating information outflow versus information inflows. The benefit of the network approach is that we can study the impact of information flow via higher-order connections (i.e., information flow from i to j through several direct viewing links). We then study the e↵ect of information flow on corporate decisions and asset pricing. Specifically, we ask three questions: First, do cross-firm viewings improve a firm’s performance and possibly stock returns. Second, does learning contribute to correlated corporate decisions? Third, does information noises also spread through cross- firm viewing links? Do we observe the propagation of idiosyncratic cash flow shocks? If so, do the central firms in the information outflow network bear higher market risk and earn higher expected returns?
The cross-firm viewings in our sample do not seem to be random. Instead, managers appear to seek relevant information from firms in di↵erent sectors. As such, we expect both the search intensity and search heterogeneity to contribute positively to corporate decision making and improves firm performance and equity return. Intuitively, Information flow should lead to correlated corporate decisions. Moreover, when learning is imperfect, it can also lead to noise and correlated idiosyncratic cash flows. Whether this shows up in the data has crucial implications for asset pricing since firm-specific risk of firms with high information outflow centrality would be more associated with market risk.
If established, the evidence would provide insight into corporate decision-making and determinants of a firm’s performance. It can also help us understand how idiosyncratic shocks lead to aggregate fluctuations and how information outflow centrality drives a firm’s beta.
| Status | Finished |
|---|---|
| Effective start/end date | 1/01/22 → 30/06/24 |
Fingerprint
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.