How to read research dates and missing data in ThesisMemo

Use the date that answers your question

A company page can contain several valid dates because it brings together different kinds of information. A research review date, a filing date, a financial reporting period and a price observation date describe different things. Treating the newest of these as the date of the whole page can create a misleading impression of freshness.

Open the free Palantir preview and identify the dates attached to its assessment and price trend. You are not trying to memorize today’s values. You are learning where the labels are so that you can repeat the check whenever you return.

Separate review time from business time

The Reviewed label describes when the company assessment was reviewed. A later review can retain the same interpretation when the collected evidence does not justify a material change. It does not necessarily represent a new company announcement or a new financial reporting period.

A filing date describes when a disclosure was filed. The period covered by its financial statements can end earlier. If you investigate a financial claim, open the source and record both the reporting period and publication information. This makes it possible to compare like with like when the next disclosure arrives.

For example, comparing one quarter with a full year would not answer a question about quarterly growth. The same caution applies to a company total versus a segment, or reported earnings versus an adjusted measure. A matching label and compatible basis are prerequisites for a meaningful comparison.

Read price-trend dates independently

The price-trend panel describes observed price behavior over its stated window and shows the dates applicable to that observation. It is not a forecast of the next move, and its direction does not establish which news caused the past move.

If you are relating a company event to market behavior, check when the event became public and what price period you are comparing. A broad trend can span several announcements and unrelated market developments. Avoid attributing the whole trend to the most memorable headline without additional evidence.

Interpret missing information as a limit

A missing assessment, pending trend or unavailable field should remain unknown until evidence becomes available. It does not mean zero revenue, no risk, no news or a neutral investment outlook. ThesisMemo can retain an earlier dated assessment when a refresh does not complete, and the page may provide a notice explaining that condition.

When you encounter a gap, first check the visible label. Then follow the relevant source link or issuer disclosure if your question depends on the missing information. If you cannot verify it, leave it out of your calculation or explicitly mark the conclusion as unresolved.

Do not fill the gap with a value from a different period merely because it is available elsewhere. That can turn an obvious missing field into a less visible analytical error. If you find an apparent inconsistency, collect the ticker, page URL, affected field and supporting source before reporting it through the contact details on Data & Methodology.

A worked example: four dates, one research question

Imagine a hypothetical company reports a quarter ending June 30 in a document published in August. A research assessment is reviewed in September, while its price panel uses a more recent set of market closes. All four dates can be valid. The June period belongs to the financial result; the August date belongs to the disclosure; the September review date belongs to the assessment; the market-close date belongs to the price observation.

Suppose your question is whether the latest quarter improved compared with the prior year. The reporting periods are the starting point, not the assessment date. If your question is whether the assessment includes a newly published announcement, compare publication and review dates and inspect the cited sources. If your question concerns a recent price move, examine the price window and the timing of relevant public information.

Now suppose one field is unavailable. Keep that gap in the record. Replacing it with an older figure could make a table look complete while making the comparison less reliable. If you use older context intentionally, label it clearly and explain which question it can still help answer.

This example contains no actual company figures. Its purpose is to show why freshness is a property of a specific observation, not a single label that applies automatically to every item on a company page.

Build a date ledger for one material observation

Create a small external ledger before combining several panels. Give the observation a descriptive label and record the period it measures, the date the source became public, the assessment review date and the date you personally checked it. Leave a field blank when it is unknown, with a short explanation. This is a suggested note format, not a claim that the product provides an exportable ledger.

For a hypothetical quarterly result, a June reporting period and an August publication date can both be correct. Your September reading date adds a third event: when you reviewed the document. Moving all three labels to September would make the record look fresh while erasing the timing needed to understand it. Preserve the original dates even when you later update your interpretation.

Use the ledger to answer a specific question. To compare operating performance, align the reporting periods and definitions. To assess whether a summary could incorporate an announcement, inspect the publication date and the actual cited material. To document your own workflow, use the date you checked it. One date cannot safely do all three jobs.

Test how a missing value affects your conclusion

Suppose an invented example reports revenue of $100 million, while the corresponding cost figure is unavailable. You cannot calculate a profit margin by assuming the missing cost is zero. You can still record the revenue observation, but the margin question remains unanswered. The example illustrates a calculation boundary and does not describe data for any company in ThesisMemo.

If you want to explore possible outcomes, keep any scenario separate from reported results. You might ask what the margin would be under a clearly labeled cost assumption, but that answer would describe your scenario rather than an observed fact. Do not copy it back into a table of actual results. A neat-looking complete table can hide more uncertainty than an honest blank.

Before sharing your conclusion, identify whether the missing item could reverse it. If the answer depends heavily on the gap, stop at the supported observation and name the evidence still required. If you are only using an older figure as background, state its period beside it. This approach lets you keep useful context without disguising an unavailable current measurement as a verified one.

Apply a short check before using a figure

Before adding a figure to your notes, ask: which company or segment, which period, which unit, which accounting basis, and which source? Then ask whether the research interpretation was reviewed before or after the relevant disclosure. This small habit is useful even when every field appears populated.

ThesisMemo’s Data & Methodology page explains intended update intervals, source types and limits. Read it alongside this guide when a label is unclear. Scheduled checks depend on available sources and successful processing; they are not a promise that every panel will contain new content at a fixed moment.

The aim is to preserve the meaning of the evidence as you move between a source, a summary and your own conclusion. A carefully labeled older observation can still be useful, provided it is not presented as something it does not measure.