What to Track in a Trading Journal (And What to Ignore)

Almost every trading journal ever started gets abandoned inside a month, and it usually deserves to. The typical one is a spreadsheet of entries, exits and profit — a record of what the market did to you. Read it back after thirty trades and the honest answer to "what should I change?" is nothing, because a list of outcomes cannot tell you which decisions were good.

A journal earns its keep when it can answer a different question: not did this trade make money, but would I take this trade again. Those two things come apart far more often than most traders expect, and the gap between them is where the actual learning is.

Two questions, not one

Every trade you take has two independent results. There is the decision — did you take a setup you had actually defined, at a size you had actually chosen, with an exit you had actually planned? And there is the outcome — did it pay?

You control the first completely. You control the second not at all. Over a small number of trades the two barely correlate, which is exactly why an outcome-only journal misleads: it hands you a random sample of results and invites you to reverse-engineer lessons from noise.

A journal that records both, separately, lets you sort your trades into four groups instead of two. That is a much more useful pile of information, and it costs one extra checkbox per trade.

The fields worth keeping

A journal you actually fill in beats a thorough one you abandon. Six fields is enough to answer every question that matters, and short enough to complete in about twenty seconds while the trade is still fresh.

That is the whole thing. If you want it already built, the trading journal on this site takes exactly these fields, runs the arithmetic below for you, and stores everything in your own browser — no account, and nothing uploaded anywhere.

What to leave out

Most abandoned journals died of too many columns. These are the ones that look responsible and produce nothing:

• —Entry and exit prices. You already have these in your broker statement, and you will never once read them back.
• —Screenshots of every trade. Enormously time-consuming, almost never reviewed. Screenshot the trades that surprised you, not all of them.
• —Emotion ratings out of ten. "Anxiety: 6" is unfalsifiable and unactionable. "I moved my stop because I didn't want to be wrong" is worth more than any number.
• —Indicator readings at entry. If an indicator value genuinely changes your decision, it belongs in your setup definition, not in the log.
• —Running account balance. It moves for reasons that have nothing to do with the quality of the trade in front of you, and watching it is how position sizing starts drifting.

Record the result in R, not dollars

R is simply your result expressed as a multiple of what you risked. Risk £200 and lose it: that is −1R. Risk £200 and make £600: +3R. The unit is the same whether you are trading a £2,000 account or a £200,000 one.

This matters more than it sounds. Dollars make two identical decisions look completely different because you happened to size them differently, and they make a good month on a big account look like skill when it was size. R strips that out and leaves the only thing worth comparing: how much you made relative to what you put at risk.

It also makes the summary arithmetic trivial. Add up your R and divide by the number of trades, and you have your expectancy — the average amount you make per trade, in units of risk. A positive number over a real sample is the closest thing to proof that a strategy works. If you are not yet sizing consistently enough for R to mean anything, fix that first with the position size calculator.

The four boxes

Once you are recording plan adherence alongside the result, every trade lands in one of four boxes — and they are emphatically not equally good or equally bad.

The top-right box is the one most traders judge themselves on and the one that matters least. A good setup loses constantly — that is what a 40% win rate means. Treating those losses as errors is how people abandon working strategies.

The bottom-left box is the dangerous one, and an outcome-only journal cannot see it at all. A rule you broke that happened to pay is the single most costly event in a trading career, because the market just reinforced the exact behaviour that will eventually take a large piece of your account. It goes in the ledger as a win. It is not one.

The only box you can fix is the one where you broke your own rule and lost. Everything else is either working as intended or outside your control.

How long before the numbers mean anything

Shorter answer than most people want: longer than you think, and the reason is arithmetic rather than patience.

Take a strategy that wins 40% of the time at 1:3. Over a hundred trades that is forty wins at 3R and sixty losses at 1R — plus 120R against minus 60R, so +60R net. Genuinely, comfortably profitable. Now look at what it feels like along the way. At a 40% win rate you should expect one or two six-trade losing streaks in every hundred trades, purely from the order the wins happen to arrive in.

The sample size problem is just as blunt. After thirty trades, a strategy whose true win rate is 40% will show somewhere between roughly 23% and 58% in your journal — that is the 95% range, and it is wide enough to contain both "this is broken" and "this is excellent". After a hundred trades the band narrows to about 30% to 50%. Thirty trades tells you very little about your win rate. It tells you a great deal about your discipline, which is why the plan-adherence column is useful from trade one and the win rate is not.

What to actually do with it

A journal nobody reads is a diary. Once a week, ten minutes, four questions:

1. 1.What percentage of trades did I take by the book? This is the only number you can directly improve next week. If it is below about 80%, nothing else in the journal is worth analysing yet — you are not running the strategy you think you are testing.
2. 2.Which setup is carrying the account? Sort by total R. Edge is rarely spread evenly; most traders have one idea that works and two or three that quietly leak.
3. 3.Which setup is bleeding? The same list, read from the bottom. Cutting a negative-expectancy setup is usually a faster improvement than finding a new one.
4. 4.What happened in the broke-the-plan trades? Not the results — the circumstances. Time of day, what happened immediately before, whether it followed a loss. Rule-breaking is almost always situational, and the situation is the thing you can plan around.

Notice that only one of those four questions is about profit.

Start with the next trade

Don't backfill three months from your broker statement. You cannot reconstruct whether you followed your plan on a trade you took in June, and a column of guesses is worse than no column — it will read as data and it is fiction.

Start with your next trade, log six fields, and give it thirty entries before you draw a single conclusion about your win rate. The discipline number will be telling you something useful long before that.

The journal on this site is free and takes no account — it runs entirely in your browser, so nothing you write ever reaches us. If you would rather use a spreadsheet, use a spreadsheet. The columns matter considerably more than the tool.