Insiders’ Profits, Costs of Trading, and Market Efficiency
Seyhun connects insider returns to firm size and trading costs, showing why an insider’s documented advantage is not the same as an outsider’s implementable result.
Seyhun examines roughly 60,000 reported purchases and sales from 1975 through 1981. The paper asks four different questions: what insiders earned, what outsiders could have earned after observing the Official Summary, which firm and trade characteristics shaped the results, and what informed trading cost other market participants. Keeping those questions separate is the paper’s lasting contribution.
The reported patterns are asymmetric and conditional. Purchases were associated with positive subsequent abnormal returns and sales with negative returns in the historical sample, but smaller firms accounted for much of the magnitude. More important for a public-data site, outsiders who imitated disclosed activity generally did not retain abnormal profits after realistic transaction costs. Reporting delay, bid–ask spreads, and portfolio construction stand between an insider’s transaction and an outsider’s result.
This is why Insider Atlas does not translate a filing into a claim about a security’s future price. It reports what occurred, identifies the filing, and supplies context about direction, company size when available, and whether the transaction belongs to a multi-insider event. Seyhun also cautions against carrying results from older, less liquid markets into the present without re-estimation. Post-SOX disclosure is faster, but faster publication does not remove costs, confounding news, or the concentration of historical results in small firms.
Abstract excerpt
“This study investigates the anomalous findings of the previous insider trading studies that any investor can earn abnormal profits by reading the Official Summary.”
Connections in Insider Atlas
Contrast an opportunistic BGDE purchase cluster with the DELL sales cluster, then inspect the underlying DELL company history. The comparison illustrates direction and context; it is not a return study.