13F Earnings Watchlist: Build One Without Guessing

What is a 13F earnings watchlist?

A 13F earnings watchlist is a research queue that starts with dated company events, then adds institutional ownership context from SEC Form 13F. The event tells you what is about to happen. The 13F layer tells you who had disclosed long exposure before the event.

That separation matters. The SEC Form 13F FAQ describes Form 13F as a quarterly report for institutional investment managers that meet the Section 13(f) threshold. The Form 13F instructions say the report is filed within 45 days after the end of the relevant calendar quarter. That makes it useful context, not a live tape.

Use the watchlist when an earnings date creates a clear question: which public stocks are about to produce new primary records, and how much pre-event ownership was visible in the last filed quarter?

What goes in the watchlist?

The table should be small enough to inspect. A useful first version has nine fields: event date, company, ticker, event source, latest 13F quarter, disclosed holders, disclosed long value, source filing link, and the specific question to answer after the event.

The source of the event should be an issuer investor-relations page, a filed 8-K, or another primary record whenever possible. Today’s bank-earnings map used the JPMorgan Chase earnings-call notice. The Tesla setup used the company’s July 2 Form 8-K and attached release.

How do you source the filing side?

Start from the legal record. The SEC’s EDGAR search is the human place to inspect filings, while a resolved API or MCP layer is the practical way to repeat the work across many tickers.

For a manual check, open one stock page and inspect the latest disclosed holders. For example, Arkolith’s JPM ownership page and Tesla ownership page show the current institutional ownership context for two earnings names in today’s newsroom slate. For a full workflow, start from the 13F data layer, then use the quickstart to query the same data from code.

The filing side should always carry the quarter and source boundary. A Q1 2026 position can be valuable before a July earnings call because it shows reported pre-event exposure. It cannot prove a July trade.

How do you rank earnings names?

Rank names by the question you can answer, not by the largest dollar number. The cleanest scoring model is simple:

A mega-cap will often win on ownership breadth. It may lose on surprise. A smaller company may have fewer holders but a cleaner question, such as a single new product, regulatory deadline, or credit-loss disclosure.

What should an AI agent do with it?

An agent should not write an earnings take from memory. It should fetch the event source, fetch the ownership data, state the lag, and then list the open questions for the post-event read.

The workflow is compact. Ask the agent to resolve the ticker, fetch the latest holders, name the source quarter, and cite the event record. The stock-owner workflow is the direct Build path, and /connect gives the agent a key for REST or MCP access.

For a plain REST starting point, use the 13F holdings API guide. For the filing mechanics, keep how to read a 13F filing, what securities are 13F reportable, and institutional ownership explained in the same cluster.

What can go wrong?

The most common error is turning a lagged filing into a live claim. A manager reported a position for the prior quarter. That does not mean the manager still owns it on the earnings date.

The second error is reading passive ownership as active conviction. Large index managers dominate many large-cap cap tables, so a high holder count may say more about index membership than about a fresh thesis.

The third error is mixing event facts with commentary. An issuer notice proves the event date. A Form 8-K proves what the company filed. Social posts and news reports may help discovery, but the watchlist should cite primary records for every event field.

What is the first useful version?

Start with five names. Add the event date, primary source, latest 13F quarter, holder count, disclosed long value, and one post-event question. Then run the same query every morning during earnings season.

That creates a repeatable research habit. Before the event, you know where institutional exposure was visible. After the event, you know which new filing, release, call transcript, or presentation changed the record.

The watchlist is not a trade list. It is a provenance tool. It helps analysts and agents move from "earnings are today" to "these are the source-backed ownership questions worth checking next."

This article explains public filings and data infrastructure. It is not investment advice, legal advice, tax advice, or a recommendation to trade any security.

Frequently asked questions about 13F earnings watchlists

Is a 13F earnings watchlist a trading signal?

No. It is a source-linked research queue. It tells you which earnings events overlap with disclosed institutional ownership, not what to buy or sell.

How stale is 13F data during earnings season?

Form 13F is quarterly and generally filed within 45 days after quarter end. During July earnings season, the broad public 13F view is usually still the Q1 filing map until Q2 filings arrive.

Should I rank names by 13F value?

Use value as one input, not the ranking rule. A smaller disclosed value can still carry a better research question if the event is clearer or the ownership is more concentrated.

Can an AI agent build the watchlist automatically?

Yes, if it has source-linked event inputs and a 13F data tool. The agent should return the event source, ownership quarter, holder count, source filing context, and the explicit lag boundary.

What is the best first query?

Ask: "Which funds disclosed positions in this earnings stock last quarter, and what filing proves each row?" That single query tests resolution, provenance, and whether the workflow is useful.