本 walkthrough 使用虛構大型消費股「公司 ABC」，示範三種披露如何回答三個不同問題。金額與日期為示意。
「未來兩至四季，機構對 ABC 的持股股數是否上升（而非僅因股價上漲），且內部人交易相對過去兩年是否異常？」
假設管理人 XYZ 對 ABC 股數季增 8%，同期組合對標普權重下降；市值增 12% 但股數增 8%，顯示部分為加碼而非僅估值上升。
季末後兩週，CFO 以 10b5-1 計劃代碼 S 出售 15,000 股；董事以代碼 P 公開市場買入 2,000 股。
通訊聲稱 ABC 將在 90 個交易日跑贏板塊 ETF。先寫清：相對 XLY、自發布日 T 起 90 日、二元命中定義。
13F：機構股數緩步增加。內部人：混合；計劃出售主導頭條但公開買入規模小。預測：在固定規則評分前不納入結論。
For Company ABC, keep six fields on one page: research question; 13F share-count delta with as-of and filing dates; Form 4 table (filer, role, code, size, plan flag); forecast claim sentence; synthesis paragraph; falsifiers. If a field is empty, say “not yet observed” rather than inventing color.
Empty fields are progress. Tip culture hates blanks; education needs them. Blanks tell you which filing family you still need to open.
Ending A: next 13F shows flat share counts—institutional accumulation story weakens. Ending B: clustered non-plan open-market sales appear—insider context darkens without proving a crash. Ending C: the pundit’s 90-day claim misses versus XLY—remove that source from the thesis weight. Practice writing endings before prices arrive.
The point of alternate endings is process rehearsal, not prediction theater. You are training the muscle that updates notes when filings change.
After this walkthrough, read the workflow guide for combining disclosures, the Form 4 codes guide for code literacy, and the forecast journal template to capture media claims cleanly. Methodology explains how InsightMeter handles data and scoring at a high level.
Assign invented but consistent dates: quarter-end 31 December; 13F accepted mid-February; Form 4 transactions in early February with filings two business days later; newsletter claim published 1 March with a 90-trading-day horizon. Writing the timeline prevents the mashup error even inside a teaching story.
Have students highlight which facts were knowable on which dates. The exercise builds lag muscles faster than another definition list.
Prompt 1: rewrite the newsletter claim three ways with different benchmarks and explain how hit/miss could change. Prompt 2: replace the CFO plan sale with a non-plan sale of equal size and rewrite the insider paragraph. Prompt 3: invent an amendment that cuts the 13F share count and update synthesis.
Good answers show process changes, not price predictions. If a student ends with “so buy ABC,” the lesson failed—restart from falsifiers.
好的筆記按披露類型分開記錄、標明滯後並列出反證條件。真實公司可套用我們的工作流程指南。
整合 13F、內部人與預測訊號 · Building a personal forecast journal
The fictional ABC note is a scaffold. Replace ABC with a real issuer only after you can keep timestamps separate without help. If mashups return, return to this walkthrough.
Teachers and study groups may copy the binder fields freely for non-commercial education. Do not present the fictional numbers as real market facts.
Next practice: complete one real binder using only EDGAR primary sources and our glossary for vocabulary. Then decide whether any optional dashboard view is even necessary for your question.
依主題學習: Form 4／內部人素養 · Form 8-K 與基本面 · Form D／豁免發行 · 13F／機構快照