What Is a Trading Journal?
A trading journal records the reasoning behind each trade — entry, stop, target and the read that produced them — so that decisions can be reviewed separately from outcomes. Its value depends on recording enough trades, and on changing one thing at a time between them.
The advice to keep one is universal and the reason given for it is usually wrong. It is not about discipline. It is about having enough records to answer a question.
How it works
Four fields, all written before the outcome is known:
Entry — the price. Stop — the price that says the read was wrong. Target — where you are done. The reason — one sentence, in plain language.
That last one is the whole exercise. The three prices are recorded by your broker anyway; the reason is not recorded anywhere, and it is the only field that can be reviewed.
The result is not the data
A good decision can lose and a bad one can win, and over a small number of trades that is the usual case rather than the exception.
So a journal organised by outcome teaches you almost nothing. It tells you which trades made money, which you already knew, and invites you to reverse-engineer a reason — which is the right-hand-edge problem this site ends every page on, applied to your own history.
A journal organised by decision can be reviewed. Did the read follow the rule? Was the stop where the rule says? Was the size right? Those have answers independent of what price did next.
How many you need
Five trades is noise. A rule with an even chance of working produces five wins in a row about three times in a hundred, which is often enough that it will happen to somebody, repeatedly, and feel like evidence.
25 mechanical entries on the shared history gave 13 stopped, 10 at target, 2 still open.
How small is worth stating as arithmetic. With 25 trades, an observed win rate is uncertain by roughly plus or minus 20 percentage points at two standard errors. So a rule whose true rate is 50% will show anywhere from about 30% to 70% across 25 trades, purely by chance.
Which means 25 trades cannot separate a 45% rule from a 55% one, and that gap is the difference between a losing strategy and a good one. Reaching plus or minus 5 points needs about 400.
That is a number you can actually work with — and it is still a small sample. The honest floor is higher than most people’s journals ever reach, which is the real reason journals get abandoned: they are asked to answer questions before they contain enough to answer anything.
Change one thing
Same entries, same targets, one variable moved: the stop went from 0.5 average true range (ATR) to 2 ATR, and stop-outs went from 24 of 25 to 7.
One change, an enormous difference. If you had changed the entry rule at the same time, nothing in the record could tell you which change did it.
That is what a journal is for, and it is why the discipline matters more than the format: a record of many things changing at once is a diary, not data.
Record the conditions
The same rule met very different markets. ATR ran from 0.55 to 1.65 across those 25 trades — a threefold difference in how much the market was moving.
A run of losses in the volatile stretch and a run of wins in the quiet one is not evidence about the rule. Without the conditions recorded, it looks like it is.
Two fields cover most of it: what volatility was doing, and whether the market was trending or ranging — from market structure, decided before entry like everything else.
A worked example
Before entry, write four lines. Three prices and one sentence.
If the sentence will not come, do not take the trade. That is the single highest-value use of the whole exercise.
Add the conditions. Two words: trending or ranging, quiet or wild.
Then leave it alone until you have enough. Reviewing after five trades produces confident conclusions from nothing.
And when you do review, change one thing.
The original data
Across our study of 24,971 trading videos, 34 cover trading journals. The median one gets 21,369 views, and only 62% fail to pass 50,000 — a strong result on a very small field.
The corpus carries no description text at all for any of those 34, so this page makes no claim about how the topic is written.
What is worth noting is the contrast in field size. 34 videos on recording your decisions, against 1,465 on day trading and 592 on Fibonacci.
When it fails
It becomes a diary
Recording how a trade felt produces a document about your moods. That may have its own value and it is not a journal in the sense this page means — it cannot answer a question about a rule.
The trades you did not take are missing
A journal of taken trades is a filtered sample, and the filter is your judgment — which is the thing under examination.
If you skip the setups that look bad, the record cannot tell you whether skipping them helped, because the skipped ones have no outcome in it.
It is reviewed after wins
Reviewing when things are going well finds reasons things are going well. A fixed review schedule removes the selection, and it is the least popular sentence on this page.
The rule changed while you were recording
A journal of a rule you kept adjusting is a record of many rules with one trade each. Fix the rule, gather the trades, then change it — the discipline the algorithmic trading page describes, applied by hand.
Related
Algorithmic trading is this exercise taken to its conclusion: writing the rule down until nothing is left unspecified.
Risk per trade supplies the size field, which is the one most journals omit.
And entry and exit is what the three price fields are actually recording.
The entry in my journal that has been worth the most is the reason field, and it is the one I most often could not fill in. If I cannot write down why I am taking a trade in one sentence before I take it, that has told me something more useful than anything the trade goes on to do. I do not track feelings in it. I track what I decided and what I decided it on.
— Michael Whitman, from this video
This page is educational, not financial advice. Test every idea on your own charts before risking money.