Methodology

What the numbers on this site mean, how they are produced, and what they cannot support. Read this before quoting anything here.

Coverage is partial and not continuous. This corpus holds CORRESP filings from 2023Q1–2024Q2 (82–96% of each quarter's EDGAR total); 2025Q4 (16% of the 861 CORRESP filings EDGAR indexed that quarter). It holds nothing at all from 2024Q3, 2024Q4, 2025Q1, 2025Q2 or 2025Q3, and nothing filed after 2025-12-31. If an issue page shows no comment from one of those periods, the reason is that edgarwiki has no data for it — not that the staff raised nothing. Counts on this site are counts within this corpus and are not SEC-wide totals. Every quotation is verbatim and links to its filing; what is incomplete is coverage, not accuracy. Per-quarter figures: Methodology.

Sources

Everything comes from the SEC's EDGAR system. Two form types matter: UPLOAD, the staff's comment letter to a company, and CORRESP, the company's reply. This build is based on CORRESP, because a response letter conventionally reproduces each staff comment verbatim before answering it — putting the comment, the response, and the outcome in one document.

Extraction

Comments and responses are separated by deterministic parsing — regular expressions anchored on response headings and numbered comment openers. No language model produces any text, number, name or date on this site. Each stored comment keeps its character offsets into the cleaned source document, and a verifier re-reads every source and asserts that the text at those offsets still matches what is stored. That check currently passes on every quotation in the corpus.

Counting

Filing counts come from EDGAR's quarterly form.idx index files. They are not taken from EDGAR full-text search, which caps result totals at 10,000 and reports the cap as though it were a count — a trap that silently corrupts any long-run series built on it.

Coverage — which quarters are actually here

This corpus is not a complete or continuous run of EDGAR. Coverage is reported below as one row per calendar quarter, because reporting it as an earliest-to-latest date range would imply a continuity this corpus does not have. Filings held is a count of CORRESP filings in this corpus filed in that quarter. On EDGAR is the number of CORRESP rows in that quarter's form.idx; where that index file has not been fetched, the denominator is shown as not measured rather than guessed.

Quarter	Filings held	On EDGAR	Coverage	Status
2023Q1	2,164	2,329	92.9%	partial
2023Q2	2,391	2,638	90.6%	partial
2023Q3	2,701	2,913	92.7%	partial
2023Q4	2,346	2,443	96.0%	partial
2024Q1	2,313	2,494	92.7%	partial
2024Q2	2,635	3,214	82.0%	partial
2024Q3	0	not measured	0%	never ingested
2024Q4	0	not measured	0%	never ingested
2025Q1	0	not measured	0%	never ingested
2025Q2	0	not measured	0%	never ingested
2025Q3	0	not measured	0%	never ingested
2025Q4	138	861	16.0%	partial

A quarter with 0 filings held was never ingested. An issue page that shows no comment from such a quarter is showing an absence of data, not an absence of comments. Closing these gaps is open work, not a claim about what the staff did.

What is in the corpus right now

Measured on the corpus as it stands: 51,900 staff comments drawn from 6,525 substantive comment-response letters, out of 14,688 CORRESP filings examined. Those filings come from 7 calendar quarters — 2023Q1, 2023Q2, 2023Q3, 2023Q4, 2024Q1, 2024Q2, 2025Q4 — and from no others.

Letter kind	Filings	Share
substantive	6525	44.4%
acceleration_request	4878	33.2%
unclassified	1627	11.1%
short_cover_letter	1587	10.8%
transmittal_letter	42	0.3%
supplemental_materials	27	0.2%
withdrawal	2	0.0%

Most CORRESP filings are not comment responses at all — a large share are Rule 461 acceleration requests and similar procedural cover letters containing no staff comments. They are identified and excluded rather than counted, which is why the substantive figure is far below the raw form count.

Topic matching

Issue pages are assembled by matching regular expressions against the verbatim comment text. This is transparent and reproducible, and it is coarse: it will group a passing mention with a substantive objection, and a comment raising several issues appears under each. Span-justified classification is the next piece of work. Until it lands, treat topic counts as an index, not a statistic.

What these numbers cannot support

They are not SEC-wide totals. They describe this corpus, which does not yet cover every published letter. A share here is a share of what has been ingested.
They are not a time series. Coverage has holes, so a rise or fall in comments on an issue between two quarters here may be a change in what was ingested rather than a change in what the staff raised. Compare only quarters the coverage table shows as present, and read their coverage percentages first.
They are not a basis for annual projections. Extrapolating a full-year figure from a partial corpus produces a number that looks authoritative and is not.
They say nothing about any company. Receiving a comment is an ordinary part of SEC review, not a finding of wrongdoing.
They are unsuitable for trading. Comment letters are published roughly 20 business days after a review concludes.

Pages built 30 August 2026 from the corpus as it stood at that moment: 14,688 CORRESP filings, 51,900 comment/response exchanges, latest filing date 2025-12-31. Every figure above is computed from the corpus in the same run that writes this line, so the two cannot disagree — but the build date is the date these pages were generated, not the date the corpus was brought up to date. The latest filing date is that second thing.