Frequently asked questions.

An independent research publication that brings together company financial statements, tracked fund holdings, insider activity, buybacks, earnings dates and IPO records. The website provides the research library; the Substack letter develops ideas from that work.

The website’s current research pages and linked JSON files are public and require no account. Subscription options for the letter are shown on Substack. A paid API and MCP connector are not currently offered.

No. The current statement extract covers 531 companies, and the fund directory tracks 20 managers. Coverage differs by dataset and period. The directories and dataset dates show what is actually available.

The site publishes saved snapshots when its data pipeline and website are refreshed. Source records have different release schedules, and an optional source failure may leave an older snapshot in place. Check each page’s as-of and filing dates; these are not streaming prices or live portfolio positions.

The company may not have filed that line for the period in machine-readable form, or a value may have been withheld because quarterly totals did not reconcile to the year. Missing values are not estimated. A blank is not a zero.

Sites can use different fiscal periods, restatement dates, split bases, concepts or adjusted earnings definitions. The Filings Report uses the latest filed value available in its extract, marks derived values and identifies the source concepts. Compare the period, filing date, units and method before comparing the numbers.

The value was calculated from filed figures rather than read as a discrete reported figure. Common examples are fourth-quarter flow values calculated from the year less nine months, and quarterly cash flows separated from year-to-date totals. These values are marked; diluted EPS is not derived by subtraction.

They are reported holdings at a quarter end, published later in a filing. They do not reveal subsequent trades, cash, shorts or the complete portfolio. Managers can also have different latest reporting periods. Each page shows the report and filing dates.

No. It names a set of published screens based on portfolio weights and changes in reported shares. The criteria describe the filing data; they do not establish a manager’s motive or predict a stock’s future return.

The purchase and sale summaries count transaction codes P and S. Awards, exercises, gifts, tax withholding and other transaction types remain in a company’s history but do not count as open-market buying or selling. Reported owners can include institutions and large shareholders as well as executives and directors.

The screen uses exactly 90 calendar dates including the snapshot date. Company summaries label a 91-day window as three months. Both use transaction dates. Filters, missing transaction prices and a difference in snapshot dates can also affect totals.

The chart uses split-adjusted historical prices. Insider transaction prices and dollar values remain as reported in the filing. A split can therefore put the two prices on different bases. Chart activity is assigned to the transaction date, not the filing date.

Upcoming dates can be provider expectations or estimates, as labelled on the page. They can change. A registration filing is not a promise that an IPO will price, and a final prospectus is a different record from an expected calendar entry.

Yes. The Datasets page links to public JSON snapshots and example company and fund files. Keep the reporting period, filing date, extraction date and units with any saved analysis. Public availability is not a blanket license to redistribute third-party data.

Yuriy Matso supplies the ideas, editorial direction and messaging. AI assists the research behind them, including the Substack letter, and helps with data analysis and development. Published table values come from source datasets and documented calculations rather than a language model’s estimate.

Email info@thetrading.tools with the page URL, company or fund, relevant period and source filing if available. A reproducible discrepancy is easier to investigate than an isolated number without its context.

The publication is independent of the SEC and is not affiliated with or endorsed by it. The site and letter provide research for informational purposes, not personalized investment advice or a recommendation to buy or sell a security.