Freeze a research information cutoff without rewriting history
A research cutoff defines the information set used for a particular note. Record both what was publicly available by that time and what you actually inspected. Later evidence should appear as a later update, not be quietly inserted into the earlier basis.
W3C provenance offers conceptual background for documenting activity and derivation. SEC search can help locate publication records relevant to a cutoff. The snapshot below is a manual research convention, not a guarantee that every source available at that time has been found.
Define the cutoff precisely enough for the question
Write the date, time, and time zone if the question requires intraday ordering. If only a date is justified, retain date-level precision. Keep preparation time separate: a note written today can reconstruct a historical information set, but it must disclose that reconstruction rather than claiming contemporaneous inspection.
Create three categories: available and inspected by cutoff, available but not inspected by cutoff, and released after cutoff. A fourth pending category can hold uncertain publication timing. This distinction prevents “could have known” from becoming “did check” in a retrospective account.
Hypothetical snapshot
Assume fictional Quarry Systems has an issuer notice on Monday, a detailed filing on Thursday, and a correction the following week. Your research cutoff is Wednesday. Dates and events are invented.
| Source | Status at Wednesday cutoff | Treatment |
|---|---|---|
| Monday notice, read Tuesday | Available and inspected | Part of original evidence basis |
| Another Monday attachment, unread | Available, not inspected | Disclose missing inspection |
| Thursday filing | Later publication | Keep outside original snapshot |
| Following-week correction | Later update | Attach to current view separately |
A current reviewer can use the correction to understand what changed. The historical snapshot should still show the earlier basis. It should not claim the original researcher had read the correction before it existed.
Preserve the snapshot and add an update layer
Save the source manifest, note version, and unresolved questions associated with the cutoff. Give the snapshot its own identifier or dated filename. When new evidence appears, create an update entry referencing the original rather than overwriting its source list.
If you reconstruct a past cutoff after the fact, say which publication metadata supports inclusion and which access history is unavailable. Do not fabricate a precise earlier inspection time. An honest reconstruction can be useful while remaining distinct from a contemporaneously preserved record.
Verify inclusion one source at a time
For every material source, identify the publication record and compare it with the cutoff. Then check your actual inspection record where available. A file’s local modification date may reflect copying or conversion, so it should not automatically serve as publication evidence.
Inspect corrections and language companions separately. A later corrected document may be easiest to download today, but the snapshot may require the earlier version. If that version cannot be retrieved, record the limitation rather than presenting a later file as though it were unchanged historical evidence.
Limits and common mistakes
A cutoff snapshot cannot prove what every market participant knew, determine private knowledge, or guarantee complete historical availability. It defines a bounded research basis. State whether the task concerns public availability, actual inspection, or both.
Common mistakes include combining release and access dates, using later information to sharpen an earlier claim, and backdating a revised note. Another is letting a current source manifest silently replace the historical one. Preserve both roles so a reader can evaluate the earlier reasoning and the current evidence independently.
A final record check
A later correction can improve today’s understanding while leaving the historical snapshot unchanged. Store both the original basis and the current clarification, connected by a dated update. If the earlier version is unavailable, do not recreate its exact wording from memory. Record the reconstruction limit and distinguish it from the source’s documented publication timing.
FAQ
Can I correct a factual error in an old note?
Yes. Preserve the original or a clear revision history, label the correction date, and identify the new evidence. Accuracy and historical traceability can coexist.
Does publicly available mean inspected?
No. Availability describes the source record; inspection describes your action. Keep both fields when evaluating what the research actually relied on.