Investment decision journal: record what was knowable
An investment decision journal records what was knowable when you made a research judgment. Preserve the question, evidence available, alternatives considered and reasons for the chosen next step. Add later outcomes in a separate section so they cannot rewrite the original reasoning.
Record a decision, not a diary
A useful entry has a specific decision boundary: continue investigating a claim, defer judgment because evidence is missing, or revise an existing thesis. You do not need to record a securities transaction for the journal to be useful. The purpose is to make the reasoning reviewable.
Use source publication dates as well as reporting periods. A 10-K describes a completed fiscal period, but it becomes available later. A later filing cannot be placed into an earlier decision’s evidence set simply because it covers that earlier year.
Hypothetical entry: delaying a conclusion
Assume a retailer reports strong holiday sales in January, while detailed inventory and returns information will arrive later. The decision is whether the sales release alone supports a claim of healthier unit economics.
| Journal field | Example entry |
|---|---|
| Recorded at | January 20, hypothetical research date |
| Question | Did holiday growth improve economics or use heavy promotions? |
| Available evidence | Sales release; no detailed return or inventory bridge |
| Main alternative | Discounting increased volume without improving contribution |
| Decision | Defer the unit-economics conclusion |
| Next evidence | Margin, returns and inventory disclosures for the same season |
| What could prove this cautious decision unhelpful? | Earlier evidence already available but overlooked |
The last row prevents the journal from validating caution automatically. Waiting can be a poor process if decisive information was available and ignored. Equally, a favorable later outcome does not prove that the original evidence justified certainty.
Preserve an evidence snapshot
Save links, document titles, publication dates and relevant excerpts within permitted use. Record unresolved definitions and calculations. Distinguish information you actually considered from information you could theoretically have found. Both matter in a process review, but they answer different questions.
When results arrive, append: observed outcome; comparability issues; mechanism assessment; and one change to the research process. Keep price performance separate from the question being tested. A rising share price does not show that holiday promotions were profitable.
Common mistakes include writing only after the outcome, recording vague expectations, and judging every decision by whether the subsequent price rose. Another is treating an uncertainty label as permanent permission not to decide. Set the next evidence date and the consequence of continued non-disclosure.
This journal is not an auditable investment performance record unless it includes the required transaction and valuation information. It is a reasoning artifact. Protect sensitive personal or company information when sharing entries, and include only what a reviewer needs to understand the judgment.
A further check on the example
For the retailer entry, preserve the exact question when the full report arrives. If later evidence shows higher margin but a longer return window, the unit-economics conclusion may still be incomplete. Append that result instead of marking the original cautious decision correct merely because some profit measure improved. Also record whether you returned on schedule. A journal that stores uncertainty but never revisits it cannot distinguish disciplined deferral from abandoning a difficult research question.
FAQ
Should I edit a mistaken original entry?
Correct factual errors with a dated note while preserving the original. Silent edits destroy the distinction between the initial judgment and the later repair.
How much detail is enough?
Enough for another reader to reproduce the key inference and identify what was unknown. A short, specific entry is better than a long narrative without a decision boundary.