Practical stock researchModule 4, lesson 17 of 23
Reviewing mistakes and updating conclusions
How to run a post-mortem on an investment, separate bad luck from bad analysis, recognise common error patterns, and keep a research log that improves future work.
In this lesson
- Run a post-mortem on an investment using the original written thesis.
- Separate bad luck from errors of information, analysis and process.
- Recognise common patterns in research mistakes.
- Keep a research log that turns lessons into changes in process.
Builds on
- Monitoring an investment thesisPost-mortems compare outcomes with the thesis and monitoring records.
On this page
In simple terms
Every investor makes mistakes. What separates those who improve from those who do not is honest review. A post-mortem looks back at an investment, compares what happened with what you expected, and asks why, so that the next piece of research is better.
Use the written record
Memory rewrites itself to fit outcomes, as the decisions and outcomes lesson explains. Review against what you actually wrote: the thesis, key assumptions, scenarios, kill criteria and monitoring notes, as earlier lessons described. Without a written record, a post-mortem becomes storytelling.
Questions to ask
- What did I expect? Restate the thesis and assumptions.
- What happened? Which assumptions held and which failed?
- Why? Classify the cause.
- Did I act on the evidence in time? Were kill criteria met and ignored?
- What will I change?
Classify the cause
| Cause | Example | Fix |
|---|---|---|
| Bad information | Missed a large related-party dealing disclosed in the notes | Read notes fully; add to checklist |
| Bad analysis | Treated a temporary cost fall as lasting pricing power | Test competing explanations, as the explanations lesson describes |
| Broken process | Bought before finishing research; ignored a kill criterion | Pre-commitment rules; waiting period |
| Bad luck | An unforeseeable event damaged a sound thesis | Usually none, beyond reviewing position size |
The categories matter because each needs a different fix. Treating bad luck as bad analysis makes you abandon good methods; treating bad analysis as bad luck stops you learning.
Worked example: A made-up post-mortem
An investor bought a share expecting margins to rise from 14% to 17% as a new plant filled. Two years later, margins were 12% and the share had fallen 35%.
| Question | Answer from the written record |
|---|---|
| What did I expect? | Margin of 17% by year 2, volume growth of 8% a year |
| What happened? | Volumes grew 9%, but margins fell to 12% |
| Why? | Raw material costs rose sharply; the company could not raise prices |
| Was it knowable? | Partly; peers had warned about input costs, and I had not tested pricing power |
| What changes? | Add a checklist item: evidence of price increases passed on in past cost spikes |
The volume assumption was right; the margin assumption failed because of untested pricing power. That is bad analysis, not bad luck, and it has a specific fix.
Review winners too
A share that rose is not proof of good research. Ask whether it rose for the reasons in your thesis. If it rose because of an event you never considered, or despite your assumptions failing, the gain was luck, and repeating the process may not work next time.
Common error patterns
Over several reviews, patterns usually appear. Common ones include:
- Overpaying for growth that did not come, as the valuation bubble case study illustrates.
- Underestimating debt and the risks it brings in bad years.
- Trusting management claims without evidence, as the evaluating management lesson warns.
- Ignoring red flags in companies you liked.
- Anchoring on the purchase price or early estimates.
- Position sizes too large for the uncertainty.
Your own patterns matter more than any list.
A research log
Keep a simple log of every research decision and review:
| Date | Company | Decision | Key assumptions | Outcome | Cause | Change made |
|---|
The last column is the important one. Each lesson should become a specific change: a new checklist item, a tighter kill criterion, a lower position limit, or a rule such as "no purchase until the red-flag review is complete".
Why it matters
Without honest review, experience can make investors worse, by reinforcing lucky habits and repeating unexamined errors. With it, each mistake makes the process stronger. Over years, that improvement compounds.
Limitations
Separating luck from skill is hard with few decisions. Post-mortems take discipline, and it is tempting to review only the painful losses. Some lessons only become clear after many years.
Common mistakes
- Reviewing from memory instead of the written thesis.
- Blaming luck for every loss.
- Reviewing only losers.
- Learning lessons without changing the process.
Try it yourself
- Choose one past investment, profitable or not, with some written record.
- Answer the five questions above in writing.
- Classify the main cause and write one specific change to your process.
- Start a research log and add this review as the first entry.
Key takeaways
- Review every investment that is closed or whose thesis broke, using the original written thesis, not your memory.
- Sort the cause into bad information, bad analysis, a broken process or bad luck. Each needs a different fix.
- Review winners too. A gain that came from luck or broken rules teaches the wrong lesson if left unexamined.
- Look for patterns across reviews, such as overpaying for growth, ignoring debt or trusting management claims.
- Turn each lesson into a specific change, such as a new checklist item, kill criterion or position limit.
Check your understanding
Finished reading?
Your progress is saved only in this browser.
Terms used in this lesson
- Post-mortem
- A review after an investment ends or its thesis breaks, comparing what happened with the written thesis and classifying the causes.
- Outcome bias
- Judging a decision by how it turned out rather than by the information and reasoning available when it was made.
- Decision journal
- A written record of each decision, its reasons, what you expect and what would prove you wrong, made before the outcome is known.
Spotted a mistake or something unclear? Report it by email. Here is how corrections work.
This lesson is general education, not investment advice. Examples use made-up or clearly dated figures to explain ideas; they are not recommendations. Read the disclaimer.