Part of the Insider Trading Tracker guide.
Sea Limited: a C-suite sale cluster is not the inverse of a buy cluster
COO Ye Gang and two presidents reported overlapping Sea (SE) sales in late August 2026, several in the Strong band, all indirect - a worked example of why high-score selling is a reading-priority flag, not a short thesis.
What happened
From August 25 to 28, 2026, Sea Limited (SE) showed a stack of Form 4 sales from COO Ye Gang and from presidents Zhao Feng (President of Garena) and Feng Zhimin. Filings landed August 27 and August 31. Several rows scored in the 80s, including 89 on a Ye Gang sale.
Ye Gang's reported lots in this window run to the millions of dollars across multiple prints near $117 to $121. Zhao Feng's rows are smaller (about 2,000 shares per print). Feng Zhimin's lots sit in between. Every export row is coded as a sale. Every one is indirect.
Why it matters
Three different senior officers of one issuer, same direction, inside a seven-day window: that satisfies the coordination structure the Greediness Score treats as a cluster. Direction matters. This is a sale cluster, not a buy cluster, and the research implication is not symmetrical.
Insiders sell for taxes, diversification, liquidity, and prearranged 10b5-1 plans. They rarely buy with cash for those reasons. Ranking a sale cluster next to a CEO open-market buy as if they were mirror images is how a scored feed gets misread as a signal service.
Key signals
- Titles: COO and two presidents - high role weight.
- Overlap: trade dates August 25 to 28, 2026 - inside a seven-day window.
- Ownership: Indirect on every export row in this window. Read the footnotes for the vehicle.
- Direction: all sales. Check the 10b5-1 checkbox and plan adoption dates before calling any lot discretionary.
The score is doing its job: it surfaced a structurally notable C-suite tape. The human reading has to apply the buys-versus-sales explainer and the 10b5-1 explainer on top of that ranking.
How to read it on EDGAR
Open each Form 4. Confirm transaction code S versus F/M/A, whether the 10b5-1(c) box is checked, the plan adoption date if any, the indirect-ownership footnote, and whether the lots are weighted-average prices. ADS versus ordinary-share lines can also split what looks like one sale.
Until those fields show discretionary, unplanned, personal-account selling, describe this as a coordinated C-suite sale event on the public tape - not as evidence that three officers share a negative thesis.
Final takeaway
Use the SE prints as a cautionary twin of a Strong-band buy cluster: the titles and overlap are real, the bearish-conviction story is not something a Form 4 can prove. Verify codes, plans, and indirect vehicles, then decide whether the tape even belongs in a research queue. A Strong Greediness Score is not a recommendation to short SE. See the rows at /app?band=strong&q=SE.
Sources
Related articles
- Why insider buys carry more information than insider sales
Insiders sell for taxes, diversification, 10b5-1 calendars, and liquidity. They usually buy with cash for one reason. Here is why the Form 4 tape is asymmetric, and how Greedy Insider uses that without pretending sales are worthless.
- What is a 10b5-1 trading plan? How prearranged insider trades work
A 10b5-1 plan lets an insider schedule trades in advance and claim an affirmative defense against insider trading claims. Here is how the plans work, what the Form 4 checkbox tells you, and why a planned sale reads differently from a cash open-market buy.
- Direct vs indirect ownership on Form 4: whose money is actually in the trade?
A Form 4 can report shares held directly by the insider or indirectly through a trust, spouse, or entity. Here is how to read that field, why direct cash buys usually carry more information, and what indirect ownership does not hide.
Educational analysis only. Nothing on this page is investment advice. The Greediness Score describes filing structure and does not predict returns.