Keeping an Investment Decision Journal

Investors remember the past imperfectly.

After an investment succeeds, it is easy to think:

I knew this would happen.

After an investment fails, it is easy to think:

There was no way I could have known.

Both conclusions may be wrong.

Memory changes as outcomes become known.

An Investment Decision Journal helps preserve what the investor actually believed at the time of the decision.

It records:

• assumptions,
• evidence,
• expectations,
• uncertainty,
• valuation,
• risks,
• and decisions

before hindsight rewrites the story.

Why Keep a Decision Journal?

Investing is a repeated decision-making process.

Over many years, the investor will:

• research companies,
• reject opportunities,
• buy investments,
• add to positions,
• reduce positions,
• sell,
• and hold through uncertainty.

Without a written record, it becomes difficult to determine whether those decisions were made well.

The final stock-price outcome is not enough.

Outcome Is Not the Same as Decision Quality

Suppose an investor buys a highly speculative company without understanding:

• the business,
• financial strength,
• valuation,
• or risk.

The stock rises:

100%.

Was the decision good?

The outcome was favorable.

The process may still have been poor.

A Good Decision Can Have a Bad Outcome

Now suppose another investor:

• understands the business,
• performs careful research,
• buys with a margin of safety,
• sizes the position conservatively,
• and identifies the major risks.

An unforeseeable event causes a loss.

The outcome was poor.

The decision process may still have been sound.

Judge Process and Outcome Separately

A Decision Journal allows the investor to ask two different questions:

Process Question

Was the decision reasonable based on the evidence available at the time?

Outcome Question

What actually happened afterward?

Keeping these separate is essential for learning.

Hindsight Bias

Hindsight bias makes past events appear more predictable after they occur.

After a company succeeds, investors may remember the positive evidence more clearly.

After failure, the warning signs can suddenly appear:

obvious.

A journal preserves uncertainty as it actually existed.

Record Before the Outcome

The most useful journal entry is written before the result is known.

For example:

I estimate a 60% probability that the company can sustain double-digit owner-earnings growth for the next five years. The largest uncertainty is customer retention as competition increases.

Later, the investor can compare that expectation with reality.

The Journal Is Not a Diary

An Investment Decision Journal is not primarily a record of emotions or daily market observations.

It is a structured record of investment reasoning.

Useful entries focus on:

• what you believe,
• why you believe it,
• what evidence supports it,
• what could prove it wrong,
• and what decision follows.

Emotions can be recorded when they affect the decision.

But they should not dominate the journal.

What Decisions Should Be Recorded?

Important entries may include:

• Initial Research
• Pass
• Watchlist
• Buy
• Add
• Hold
• Reduce
• Sell
• Thesis Revision
• Valuation Revision
• Position-Size Change

The journal should capture meaningful changes in reasoning.

Record Rejections Too

Investors often record purchases but forget rejected companies.

That loses valuable information.

Suppose you reject a company because:

Valuation is too demanding at $80.

Two years later the stock trades at:

$45

while the business has strengthened.

The original rejection reason may no longer apply.

A rejection journal creates future research opportunities.

Record Why You Passed

Possible reasons include:

• business too difficult to understand,
• valuation unattractive,
• weak financial strength,
• insufficient evidence,
• management concerns,
• or better opportunities elsewhere.

A clear reason makes later review possible.

The Initial Research Entry

When serious research begins, record:

• why the company attracted attention,
• what you currently understand,
• what remains unknown,
• and what questions need answers.

This creates the starting point of the research history.

Discovery Source

Record how the company entered the research pipeline.

Examples include:

• Screener
• Discovery Intelligence
• Watchlist
• Industry research
• Theme research
• Company comparison
• Existing holding
• External research idea

This can later reveal which discovery methods produce useful candidates.

Research Question

Every research entry should have a question.

Examples include:

Does this company have a durable switching-cost moat?

Is the current low valuation caused by temporary weakness or structural decline?

Can the company reinvest at high returns for another decade?

A question makes research purposeful.

The Buy Entry

Before purchasing, record the reasoning that justifies committing capital.

A useful Buy entry should include:

1. Business
2. Investment Thesis
3. Supporting Evidence
4. Contradictory Evidence
5. Important Assumptions
6. Valuation
7. Bear, Base, and Bull Cases
8. Thesis Breakers
9. Position Size
10. Expected Holding Logic
11. Major Unknowns
12. Reasons Not to Buy

This creates a snapshot of the decision.

Record the Business in Plain Language

Write:

What does the company do, and how does it make money?

If the explanation requires pages of jargon, understanding may still be incomplete.

The journal should preserve your own understanding rather than copied company language.

Record the Thesis

The thesis should explain:

• why value can grow,
• why attractive economics may persist,
• what management must do,
• what valuation is being paid,
• and what could invalidate the argument.

Keep it concise enough to review later.

Record Supporting Evidence

List the strongest evidence supporting the thesis.

For example:

• retention above 95%,
• ROIC above 20% for ten years,
• net cash balance sheet,
• consistent free-cash-flow conversion,
• and rational capital allocation.

Specific evidence is more useful than vague optimism.

Record Contradictory Evidence

Every serious entry should include:

What evidence argues against me?

Possible examples include:

• slowing growth,
• customer concentration,
• rising competition,
• falling incremental ROIC,
• or expensive valuation.

A journal should preserve conflict rather than present a sales pitch.

Record the Strongest Bear Argument

Write the strongest reasonable argument against the investment.

For example:

The apparent moat may be weaker than historical retention suggests because new technology is reducing switching costs.

If you cannot articulate a serious opposing case, confirmation bias may be influencing the research.

Record Assumptions

Many investment conclusions depend on assumptions.

Examples include:

• revenue growth remains above 8%,
• margins remain near 20%,
• customer retention remains above 90%,
• debt declines,
• or reinvestment continues earning high returns.

Write these explicitly.

Hidden assumptions cannot be monitored.

Separate Facts From Assumptions

This distinction is important.

Fact

Revenue grew 12% last year.

Assumption

Revenue can grow approximately 10% annually for the next five years.

Conclusion

The business may compound intrinsic value at an attractive rate.

Do not treat all three as equivalent.

Record Expected Ranges

Avoid unnecessary precision.

Instead of:

Revenue will grow exactly 11.7%.

record:

Base-case revenue growth: approximately 8% to 12%.

Ranges preserve uncertainty while keeping expectations testable.

Record the Valuation

Write:

• current market price,
• bear value,
• base value,
• bull value,
• and major assumptions.

If possible, record the date.

Later, you can see whether value changed because:

• the business changed,
• assumptions changed,
• or only the stock price changed.

Record Margin of Safety

Write why the current price provides—or does not provide—sufficient room for error.

Do not record only:

Undervalued.

Explain:

Estimated base value is $100 to $115 while market price is $72, but uncertainty is elevated because the growth runway is difficult to estimate.

This is much more useful later.

Record the Bear Case

The Bear Case should describe a plausible unfavorable outcome.

Ask:

• What goes wrong?
• What happens to growth?
• What happens to margins?
• What happens to financial strength?
• What might the business be worth?

The journal preserves what downside you believed you were accepting.

Record the Base Case

The Base Case should represent the outcome most strongly supported by current evidence.

Do not make it the outcome you hope will occur.

Write the assumptions clearly enough to compare them with future results.

Record the Bull Case

The Bull Case should describe a favorable but plausible outcome.

It should not require:

• perfect execution,
• unlimited growth,
• or permanently extreme valuation.

A realistic bull case helps define upside without turning optimism into fantasy.

Record the Stress Case

For investments with meaningful financial or operating risk, record a Stress Case.

Ask:

Can the company survive unusually severe conditions?

This may include:

• recession,
• customer loss,
• refinancing difficulty,
• or severe margin compression.

Survival matters before upside.

Record Thesis Breakers

Before buying, write the evidence that would materially weaken or invalidate the thesis.

Examples include:

• retention below 85%,
• leverage above a defined level,
• sustained negative free cash flow,
• major moat erosion,
• or repeated value-destructive acquisitions.

These conditions become monitoring anchors.

Record Position Size

Write:

Position size: 5%

Then explain why.

For example:

High business quality and strong financial strength support a meaningful position, but valuation uncertainty limits initial size.

This allows later review of sizing judgment.

Record Why the Position Is Not Larger

This is a powerful question.

Possible answers include:

• valuation uncertainty,
• customer concentration,
• short operating history,
• financial risk,
• or portfolio correlation.

The answer reveals what limits conviction.

Record Why the Position Is Not Smaller

Likewise ask:

What evidence justifies making this position meaningful?

This prevents position sizing from becoming arbitrary.

Record the Opportunity Cost

Write the main alternatives considered.

For example:

Alternative A offers similar quality but materially higher valuation. Alternative B is cheaper but financially weaker. Cash remains a reasonable alternative.

This shows why the chosen investment won the capital-allocation decision.

Record Emotional State

Emotion matters when it may influence judgment.

Examples include:

• fear after a market decline,
• excitement after strong earnings,
• FOMO after rapid appreciation,
• reluctance to realize a loss,
• or attachment to a long-held company.

A simple note can be enough.

For example:

I feel urgency because the stock rose 20% this week. I am delaying the decision until valuation is updated.

Record the Decision

End the entry clearly.

For example:

Decision: Initiate a 4% position.

or:

Decision: Pass because financial risk is too high despite attractive valuation.

or:

Decision: Watchlist until evidence of margin stabilization appears.

Ambiguous journal entries are harder to learn from.

Record What Would Change the Decision

After recording the decision, write:

What evidence would cause me to act differently?

For example:

I would add if retention remains above 92% and valuation stays below the base-case range.

or:

I would reduce if leverage rises above the thesis threshold.

This creates accountability.

The Add Entry

When adding to a position, do not simply write:

Bought more because the stock fell.

Record:

• what changed in price,
• what changed in value,
• what changed in evidence,
• what changed in conviction,
• and what changed in portfolio weight.

Adding is a new capital-allocation decision.

Averaging Down Journal Entry

If adding after a decline, explicitly answer:

Why did the stock fall?

Did intrinsic value fall too?

Is the thesis stronger, unchanged, or weaker?

Would I buy this amount if I had never owned the stock before?

These questions reduce anchoring.

Adding After a Price Increase

The journal is equally useful when adding at a higher price.

Perhaps:

• evidence strengthened,
• uncertainty declined,
• and intrinsic value increased faster than price.

Record that reasoning.

A higher purchase price is not automatically worse.

The Hold Entry

Holding should also be recorded when a meaningful review occurs.

For example:

Decision: Hold. Thesis remains intact. Intrinsic value increased modestly, current valuation is reasonable, and position size remains appropriate.

This turns holding into an explicit decision rather than inertia.

The Reduce Entry

When reducing, record why exposure is being lowered.

Possible reasons include:

• valuation,
• concentration,
• rising uncertainty,
• weakening thesis,
• or opportunity cost.

Then record why the entire position is not being sold.

That distinction clarifies the reasoning.

The Sell Entry

A Sell entry should record:

• thesis status,
• current value estimate,
• reason for sale,
• evidence,
• opportunity cost,
• and any relevant tax or portfolio considerations.

Do not simply write:

Sold because stock fell.

or:

Sold to take profits.

Those statements reveal little about economic reasoning.

The Thesis Revision Entry

When important evidence changes, create a new journal entry rather than silently rewriting the original thesis.

Record:

• previous thesis status,
• new evidence,
• interpretation,
• revised assumptions,
• revised valuation,
• and new thesis status.

This preserves the history of the investment.

Preserve the Original Thesis

Suppose the original thesis said:

Customer switching costs should support retention above 90%.

Two years later retention falls to:

78%.

Do not edit the old thesis to say:

Retention was never central to the investment.

Preserve what you actually believed.

Then record what changed.

Thesis States

A journal can use consistent thesis states such as:

• Strengthening
• Intact
• On Watch
• Weakening
• Broken

This creates a simple history of conviction.

The label should always be accompanied by evidence.

Example Thesis Revision

A journal entry might say:

Status: Intact → On Watch

Trigger: Customer retention declined from 94% to 88% while competitors introduced easier migration tools.

Interpretation: The switching-cost thesis remains plausible, but evidence has weakened.

Next evidence required: Retention and renewal pricing over the next two reporting periods.

This is far more useful than:

Getting worried.

Record What Remains Intact

A negative development should not automatically erase everything else.

Record what remains supportive.

For example:

• financial strength remains high,
• free cash flow remains strong,
• and management allocation remains rational.

This helps prevent overreaction.

Record What Became More Uncertain

Sometimes the evidence does not clearly prove deterioration.

Instead, uncertainty increases.

Write that explicitly.

For example:

Moat confidence reduced from High to Moderate because customer behavior has become less predictable.

Uncertainty is a legitimate analytical outcome.

The Valuation Revision Entry

Intrinsic value changes when important assumptions change.

Record:

• previous value range,
• new value range,
• assumptions that changed,
• and evidence supporting the revision.

This prevents anchoring to old valuations.

Example Valuation Revision

Suppose the original value range was:

$90 to $110.

Growth evidence weakens.

The new range becomes:

$75 to $95.

Record why.

Do not preserve the old $110 value simply because the stock now trades at $70.

Separate Price Changes From Value Changes

Every valuation review should distinguish:

Market Price

What changed in the quotation?

Intrinsic Value

What changed in the economics?

A stock can fall while value remains stable.

A stock can rise while value falls.

The journal helps preserve the distinction.

The Monitoring Entry

Periodic monitoring entries should focus on thesis variables.

Instead of recording every quarterly number, ask:

What changed that matters to the thesis?

Possible areas include:

• retention,
• margins,
• debt,
• incremental ROIC,
• growth runway,
• management behavior,
• or valuation.

The journal should capture signal rather than noise.

Quarterly Review Entry

A quarterly review might record:

Thesis Status

Intact.

Important Positive Evidence

Retention stable and free cash flow ahead of expectations.

Important Negative Evidence

Gross margin declined modestly.

Interpretation

Margin decline appears related to temporary investment rather than pricing weakness.

Valuation

Base value unchanged.

Decision

Hold.

This creates a concise analytical history.

Annual Deep Review

An annual entry can be more comprehensive.

Revisit:

• business quality,
• moat,
• management,
• financial strength,
• growth,
• valuation,
• risk,
• and thesis breakers.

Then compare the company with the original thesis.

Compare Expectations With Reality

The journal becomes powerful when old expectations are reviewed.

Suppose you wrote:

Expected five-year revenue growth: 8% to 12%.

Three years later actual growth averages:

5%.

Ask:

• Why was the expectation wrong?
• Did the market mature faster?
• Did competition increase?
• Was the original evidence weak?

This turns forecasting error into learning.

Review Assumptions Individually

Do not judge the thesis only by whether the stock rose.

Review assumptions such as:

• growth,
• margins,
• retention,
• ROIC,
• debt,
• and capital allocation.

Determine which were:

• correct,
• too optimistic,
• too pessimistic,
• or irrelevant.

Calibration

Over time, the journal can reveal whether your confidence matches reality.

Suppose you repeatedly describe investments as:

High Confidence

but many critical assumptions fail.

Your confidence may be poorly calibrated.

The journal makes this visible.

Probability Calibration

If you use probabilities, review them over many decisions.

Suppose you classify events as:

80% likely.

Over a large enough sample, events with that confidence should occur roughly often enough to justify the label.

The goal is not mathematical perfection.

It is better awareness of uncertainty.

Avoid False Precision

Do not feel obligated to assign exact probabilities when the evidence does not support them.

It may be more useful to write:

• Low Confidence
• Moderate Confidence
• High Confidence

with an explanation.

The journal should clarify uncertainty, not disguise it.

Record Surprises

When something materially unexpected happens, write:

Why did this surprise me?

Possible answers include:

• missing evidence,
• poor industry understanding,
• management unpredictability,
• excessive confidence,
• or genuinely unforeseeable events.

This helps distinguish bad process from bad luck.

Good Luck and Bad Luck

A journal should recognize luck.

A good outcome can result partly from:

• favorable macro conditions,
• unexpected acquisition interest,
• or multiple expansion.

A bad outcome can result partly from:

• unusual external shocks.

Learning requires separating skill from luck where possible.

Process Attribution

When reviewing an investment, classify what drove the result.

Possible contributors include:

• business analysis,
• valuation,
• position sizing,
• management assessment,
• timing,
• behavioral discipline,
• or luck.

This can reveal where the process is strongest and weakest.

Review Winners Carefully

Successful investments can teach bad lessons.

Suppose you buy an overleveraged company without sufficient research.

A favorable cycle rescues it and the stock triples.

The dangerous conclusion is:

My analysis was excellent.

The journal may show that the result depended heavily on luck.

Review Losers Carefully

Likewise, a losing investment does not automatically prove poor process.

Ask:

• Was the risk identified?
• Was the probability reasonable?
• Was position size appropriate?
• Was the loss survivable?
• Did an unforeseeable event occur?

The purpose is accurate learning.

Review Missed Opportunities

A decision journal can also examine companies you passed on.

Suppose you rejected a company because:

Valuation appeared too high.

The stock later rises dramatically because intrinsic value compounds much faster than expected.

Ask:

Did I underestimate quality or growth?

The lesson may be about valuation assumptions rather than the decision to maintain price discipline.

Review Correct Rejections

Some of the best investment decisions are companies never purchased.

A journal can preserve evidence that:

• risk was unacceptable,
• valuation was extreme,
• or the business was outside the circle of competence.

Avoiding permanent loss deserves recognition.

The Journal Should Record Process Errors

Useful error categories might include:

• Business Understanding
• Financial Analysis
• Moat
• Management
• Growth
• Valuation
• Risk
• Evidence
• Position Sizing
• Psychology
• Monitoring

Over time, patterns may emerge.

Example: Repeated Valuation Errors

Suppose several losses come from:

paying too much for excellent businesses.

The process may need stronger valuation discipline.

The lesson is not:

Avoid excellent businesses.

It is:

Improve the relationship between quality and price.

Example: Repeated Moat Errors

Suppose several investments disappoint because competitive advantages were weaker than expected.

The checklist may need better questions about:

• retention,
• substitution,
• competitor economics,
• and switching behavior.

The journal improves the checklist.

Example: Repeated Behavioral Errors

Suppose the journal reveals:

• buying after rapid price increases,
• selling during panic,
• and averaging down without thesis review.

The problem may be behavioral rather than analytical.

That insight can lead to new decision rules.

Journal and Checklist Work Together

The Research Checklist creates consistency before decisions.

The Decision Journal preserves what happened during decisions.

Later review asks:

Which checklist questions failed to protect me?

Then the checklist can improve.

This creates a learning loop.

Journal and Thesis History Work Together

The thesis records the current investment argument.

The journal records how that argument changed.

Together they show:

• original reasoning,
• new evidence,
• revised conviction,
• and final outcome.

This history is valuable for learning.

Journal and Portfolio Review Work Together

A journal can reveal portfolio-level mistakes.

For example:

Several individually reasonable investments may all depend on:

• the same economic cycle,
• interest rates,
• or technology trend.

The error may be concentration rather than company selection.

Do Not Rewrite Old Entries

Old entries should remain immutable in substance.

You can add:

• comments,
• corrections,
• or later reflections.

But do not silently rewrite what you believed.

Otherwise hindsight returns.

Add a Reflection Instead

Suppose an old entry contains an assumption you now consider foolish.

Keep it.

Add:

Reflection — 2027-06-15: I underestimated customer concentration because I focused too heavily on revenue growth.

The mistake becomes visible and useful.

Date Every Entry

Every journal entry should record when the reasoning existed.

Investment conclusions without dates can become misleading because:

• price changes,
• evidence changes,
• and businesses evolve.

Time is part of the analytical context.

Preserve Evidence References

Where practical, connect important claims with their sources.

This makes it easier to review:

• what information existed,
• how it was interpreted,
• and whether later evidence contradicted it.

A journal should preserve reasoning provenance.

Keep Entries Concise Enough to Review

A journal that becomes too burdensome may stop being used.

Not every entry needs thousands of words.

The depth should match the importance of the decision.

A major purchase may deserve extensive documentation.

A routine Hold review may require only a concise update.

Consistency Matters More Than Literary Quality

The journal does not need elegant prose.

It needs:

• clear reasoning,
• explicit assumptions,
• visible evidence,
• and honest uncertainty.

Write for your future self.

Create Standard Entry Types

A repeatable journal might use templates for:

Research

Why am I investigating?

Buy

Why am I committing capital?

Add

Why should exposure increase?

Hold

Why does current ownership remain rational?

Reduce

Why should exposure decline?

Sell

Why should ownership end?

Thesis Revision

What changed in the investment argument?

Review

What did I learn?

Templates reduce friction.

Common Mistakes

Writing the journal after the outcome

This invites hindsight bias.

Recording conclusions without evidence

Future review cannot test the reasoning.

Recording only purchases

Passes, Holds, Adds, Reductions, and Sells also matter.

Hiding contradictory evidence

The journal should challenge conviction.

Rewriting old entries

Preserve what you actually believed.

Judging decisions only by stock returns

Process and outcome are different.

Writing so much that journaling becomes unsustainable

The system must remain usable.

Ignoring emotions

Behavioral state can materially influence decisions.

Practical Exercise

Create a Decision Journal entry for one investment.

Use this structure:

Date

Record today's date.

Decision Type

Choose:

• Research
• Watch
• Buy
• Add
• Hold
• Reduce
• Sell

Business

Explain what the company does in two sentences.

Thesis

Write the investment thesis in one paragraph.

Supporting Evidence

List the five strongest pieces of evidence.

Contradictory Evidence

List the three strongest pieces of evidence against the thesis.

Assumptions

Write the most important assumptions.

Valuation

Record:

• current price,
• bear value,
• base value,
• bull value.

Thesis Breakers

Write at least three.

Position Size

Record the current or proposed portfolio weight and explain why.

Opportunity Cost

Name the best alternative use of the capital.

Emotional State

Record whether fear, greed, FOMO, anchoring, or loss aversion may be influencing you.

Decision

Write exactly what you are doing.

Change Conditions

Write what evidence would cause you to change the decision.

Save the entry.

Do not edit it later.

When new evidence appears, create a new dated entry.

The Buffett Perspective

Investment skill develops over long periods.

Memory alone is a poor teacher because successful outcomes can make weak reasoning look brilliant, while unlucky outcomes can make sound decisions look foolish.

A written record forces the investor to confront:

• what was actually known,
• what was assumed,
• what price was paid,
• what risks were accepted,
• and what reasoning supported the decision.

This encourages intellectual honesty.

The goal is not to avoid every mistake.

That is impossible.

The goal is to learn the correct lesson from each decision.

The RW Finance Perspective

RW Finance should make the Research Journal a durable record of investment reasoning.

Journal entries should be able to preserve:

• date,
• decision type,
• thesis,
• assumptions,
• supporting evidence,
• contradictory evidence,
• valuation,
• confidence,
• thesis breakers,
• position context,
• and later reflections.

The journal should work alongside:

• Company Research,
• Investment Thesis,
• Evidence,
• Watchlist,
• Portfolio,
• Valuation,
• and thesis monitoring.

Historical entries should remain preserved rather than silently overwritten.

Users should be able to see:

What did I believe then?

What evidence did I have?

What did I expect?

What actually happened?

What changed my mind?

What did I learn?

Over time, this creates something more valuable than a record of trades.

It creates a record of judgment.

Key Takeaways

• An Investment Decision Journal preserves what the investor actually believed before hindsight changes the memory of the decision.
• Decision quality and investment outcome should be evaluated separately.
• Important journal entries include Research, Pass, Watchlist, Buy, Add, Hold, Reduce, Sell, Thesis Revision, and Review.
• Facts, assumptions, conclusions, valuation, and uncertainty should remain distinct.
• Supporting and contradictory evidence should both be recorded.
• Bear, base, bull, and stress cases preserve the range of outcomes considered at the time.
• Thesis breakers and change conditions make future decisions more accountable.
• Position size and opportunity cost should be recorded because investment selection and capital allocation are connected.
• Old entries should not be silently rewritten; later reflections should be added separately.
• Reviewing winners, losers, rejected ideas, and missed opportunities can reveal different process weaknesses.
• The Decision Journal and Research Checklist form a learning loop that can improve the investment process.
• The purpose of the journal is not merely to record trades but to preserve reasoning so judgment can improve over time.