Memory does not store your reasoning; it reconstructs it, and it reconstructs it to fit whatever happened. The only honest record is the one made before the outcome existed.
9 min read · Level 4
The usual case for a decision journal is that it helps you learn from your mistakes, and that undersells it in a way that also makes it easier to skip. Learning from mistakes is a slow, uncertain benefit that arrives years later, which is exactly the kind of benefit people discount to nothing. The immediate benefit is different: writing the reasoning down changes the reasoning while it is still changeable. A conviction that is clear in the head and incoherent on the page is extremely common, and it is not a writing problem - the incoherence was always there, and the page is simply the first surface hard enough to show it. That is why the four lines are worth writing even for someone who never opens the journal again. There is a cost worth naming honestly, though, because the practice is often sold as free. A written thesis is also a commitment, and commitments are sticky: an investor who has articulated their reasoning at length can become harder to move than one who never did, which is the opposite of the intended effect. The protection is the line about what would change your mind, written first and treated as binding - and an investor who writes everything except that line has built themselves a machine for defending positions.
A decision journal is a written record of why you made a decision, made at the time you made it, and never edited afterwards.
It is not a diary and it is not a record of trades. A brokerage statement already knows what you bought, when, and at what price. What no system anywhere holds is the thing that actually mattered: what you believed at the moment you committed, what you expected to happen, and what would have told you that you were wrong.
The entry needs to contain claims rather than feelings. 'I like this business' cannot be checked against anything later. 'I expect earnings to grow at least 20% a year for three years, and I am paying 18 times current earnings for that' is a falsifiable claim: in two years it will be either true or false, and which one it turns out to be is not a matter of interpretation.
The discipline is entirely in the timing. An entry written afterwards is not a record. It is a reconstruction, and it will be a flattering one.
Because hindsight bias is not a weakness you can decide to be free of. It is how memory works.
Once you know what happened, the version of events in which it was foreseeable becomes vastly easier to recall than the version in which it was genuinely uncertain, and the second version quietly stops being available. This is not lying to yourself; you are not aware of it happening, and confidence in the reconstructed memory is typically higher than confidence in the original belief was.
The practical consequence is that an investor without written entries cannot learn from experience in any reliable way. Every outcome confirms that they understood the situation, because their memory of understanding it was assembled after the fact from the outcome itself. Twenty years of that produces twenty years of confirmation and very little knowledge.
And because writing it down changes the decision, not just the record of it. A reason that will not survive being written in a sentence is often not a reason.
Educational, not personalised financial advice. Nothing in the Investor Academy is a recommendation to buy, sell or hold anything — what is right for you depends on your own circumstances and the rules where you live.
March. You buy a holding at 72.00 a share, and you write four lines before you do:
I am paying 18 times current earnings of 4.00 a share. I expect earnings to grow at least 20% a year for the next three years. What would tell me I am wrong: two consecutive reported periods with growth below 10%. What I have not examined: whether 18 times is a reasonable price to pay for that growth.
Two years later the shares are at 51.84 and you are down 28.00%.
Now read the entry against what happened. Earnings are 5.76 a share - 4.00 grown at 20% a year for two years is 4.00 x 1.20 x 1.20 = 5.76, which is 44.00% growth over the period and precisely what you forecast. The business did exactly what you said it would do. The price fell because the market now pays 9 times earnings rather than 18: 9 x 5.76 = 51.84, against 18 x 4.00 = 72.00, which is the entire 28.00% decline.
So the entry tells you something you could not otherwise have known. Your forecast was right, to the decimal. Your loss came from the one thing your own fourth line admitted you had not examined. That is a specific, correctable weakness with a name.
Without the entry, what would you have? Almost certainly a memory along the lines of 'I always thought it was expensive' - a memory that is now available precisely because the multiple fell, and which would have taught you the opposite lesson: to be more sceptical of growth forecasts, when your growth forecast was the only part of this you got exactly right.
The specific wrong move is writing the entry at the end of the week, or the month, or 'when I have time'.
An entry written three days later is contaminated by three days of price movement and three days of other people's opinions. It is not obviously worse - it reads perfectly well, it may even be more articulate - which is precisely what makes it dangerous, because it will be filed and consulted as though it were a contemporaneous record.
It is tempting because the moment of deciding is busy and the writing feels like admin. It is also tempting because at the moment of deciding the reasoning feels so vivid that recording it seems unnecessary. That vividness is the illusion; it is exactly as strong for reasoning you will have completely forgotten in six weeks.
The second mistake is writing a mood instead of a claim. 'Strong company, good management, exciting market' cannot be wrong about anything, so it can never teach you anything. Every sentence in an entry should be one that a future fact could contradict.
The entry is part of the decision, not a record of it. If there is no time to write four lines, there was no time to make the decision either - which is itself information worth noticing.
A number, a timeframe, a comparison. 'Earnings above X by date Y' can be checked; 'a great business' cannot. The test is simple: name the fact that would prove the sentence wrong. If no such fact exists, rewrite the sentence.
This line is the most valuable one in the entry and the hardest to write later, because once evidence arrives it is nearly impossible to say honestly what threshold you would have set in advance. Written first, it becomes a commitment. Written afterwards, it becomes a rationalisation.
Every decision has parts you did not have time or ability to check. Writing them down costs nothing and converts an unknown unknown into a known one - and when something goes wrong, this line is frequently where the answer is.
For any decision you would be unhappy to be wrong about, in either direction: committing money, declining something, changing an allocation, or deliberately doing nothing when a case for acting existed.
Not worth the effort for small, routine, reversible actions. A journal that covers every transaction becomes a chore, is abandoned within a month, and then covers nothing at all.
Ask an experienced investor why they made a particular decision years ago and you will usually get a clear, coherent, confident account. Ask someone who kept written entries to compare the account they would have given with what they actually wrote, and the two frequently diverge - not on the facts of the transaction, which are documented, but on the reasoning, the expectation and the degree of doubt. The remembered version is characteristically tidier: fewer considerations, more certainty, and a closer fit to how things turned out.
Treat any confident recollection of your own past reasoning as a reconstruction until a contemporaneous document says otherwise. The absence of doubt in the memory is evidence about the memory, not about the original decision.
Create a free account to read the rest and apply it to your own investing.