Most investors track what they bought. Almost none track why they bought it. A decision journal is a written record of your investment decisions — including the reasoning, the information you had at the time, and the outcome you expected — kept so you can audit your own thinking after the fact. It is the primary tool for separating skill from luck.
The problem it solves: outcome bias
When an investment works out, we remember the reasoning as sharp. When it doesn't, we remember the thesis as obviously flawed — in hindsight. This is outcome bias: letting the result regrade the process retroactively. It is nearly impossible to correct without a written record of what you actually thought before you knew the outcome.
A decision journal breaks that loop. It locks your reasoning in place at decision time so the future you can read it without revision.
What goes in an entry
A useful entry captures four things:
| Field | What to write |
|---|---|
| Decision | What you're doing — buy, sell, hold, position size. |
| Thesis | The 2–3 facts that make this a good bet. |
| Falsifier | What evidence would prove the thesis wrong. |
| Expected outcome | The rough return range you expect and over what horizon. |
The falsifier is the most important and most skipped field. Naming it in advance forces you to hold the thesis to a specific standard instead of moving the goalposts later.
How to run a review
Reviews are more valuable than entries. Schedule a review when each thesis resolves — or quarterly for long positions:
- Read the original entry before looking at what happened.
- Grade the process, not the outcome: was the reasoning sound given what you knew at the time?
- Note what you got right, what you got wrong, and what information you were missing.
- Flag recurring errors (overconfidence on a specific sector, ignoring rate sensitivity, anchoring
on entry price, etc.).
A pattern appearing three or more times across entries is a systematic bias worth correcting. One bad call is noise. Three bad calls with the same flaw in the same field is signal.
A worked example
Suppose you bought a mid-cap software stock. Your original entry read:
Thesis: recurring revenue growing 25% YoY, net revenue retention above 120%, valuation at 8× ARR is below sector median. Falsifier: growth decelerates below 15% or churn rises. Expected outcome: 30–50% return over 18 months.
The stock then falls 40% after a disappointing earnings report. Without a journal, you might remember your thesis as vague and revise it in your head. With the journal you can see exactly what happened: revenue growth slowed to 12%, which hit the falsifier you named. The process was sound; the thesis failed. That is useful data.
Now contrast a situation where growth held but the stock fell on a broad macro rotation. In that case the thesis held and the outcome was noise — the market, not the business, moved against you. A journal keeps those two cases from blurring together after the fact.
Why investors specifically benefit
Most portfolios hold 10–30 positions with decisions spread over years. Memory degrades fast enough that after six months, most investors cannot accurately reconstruct their original reasoning. A journal is the only retrieval system that works at that time horizon.
It also creates a lightweight accountability layer: a decision you would not commit to paper is probably not ready to execute.
Try this
Before your next buy or sell, write three sentences: your thesis, your falsifier, and what outcome you expect. Store it in JustJot.ai under a tag like decision-journal — the semantic search means you can ask "what did I think about rate sensitivity in 2025?" and surface the relevant entries even if you did not use those exact words when you wrote them. That retrieval is where the value compounds.