Trading Journal: What to Track So It Actually Helps

Almost everyone starts a trading journal and almost everyone quits within three weeks. Not from laziness, from design. A template with twenty four columns and an "emotional state" box is impossible to fill in live and answers no useful question when you read it back.

A journal that works has eight fields, takes thirty seconds per trade and gets reviewed once a week. Everything else is decoration. Here is exactly what to log, what to ignore, and what questions to ask the data once you have fifty trades in it.

Why journals get abandoned

Three reasons, always the same ones.

It asks for too much. If logging one trade takes three minutes, a six trade day costs eighteen minutes after the close, when you are tired and done. By the third bad session you stop, and those are exactly the sessions worth recording.

It answers no question. Writing "felt impatient" cannot be summed, sorted or compared. A journal is a database, not a diary. Any field you cannot filter or average is decoration.

Nobody reads it back. That is the worst one. Two hundred beautifully filled rows nobody has opened since March are worth exactly zero. The value is not in writing, it is in the Saturday review.

Rule of thumb: if a field will not change a decision, do not log it.

The eight fields that matter

These are the ones. Nothing more to start.

Field	Example	What it is for
Entry date and time (ET)	08/18 9:42 AM	Finding which part of the day you win in and which you lose in
Instrument and size	MNQ, 2	Seeing whether size climbs exactly when things go wrong
Direction	Long	Plenty of traders carry a clear bias and never notice it
Setup (closed tag)	Sweep-VWAP	The single most useful metric: which setup pays
Initial risk in points	18 pts	Converts everything into R multiples so trades compare
Result in R	+1.8R	A dollar result is not comparable across account sizes
Max adverse excursion (MAE)	7 pts	Tells you whether stops are too wide or too tight
Was it in the plan? (yes/no)	No	Isolates the real cost of improvising

Two notes on that table.

Setup has to be a closed tag. Four or five fixed names picked from a dropdown. The moment you allow free text you end up with "support bounce", "bounce support" and "supp. bounce" as three separate categories and nothing groups.

Everything in R, not dollars. Risk 18 points, make 32, that is +1.8R. It does not matter whether it was 1 contract or 4, or whether the account is 50K or 150K. Working in risk multiples is what lets you compare Tuesday with Thursday. Converting points to dollars on each micro is covered in MNQ, MES and MGC tick value.

Where the data comes from without typing it

NinjaTrader 8 already stores half of it. In the Control Center, Trade Performance tab, you get every execution with time, instrument, quantity and result, and it exports to CSV. That covers six of the eight fields with zero typing. The two you add by hand are the setup tag and the plan yes/no. Thirty seconds.

The one tag that saves you the most money

Of the eight fields, one is worth more than the other seven combined: was this trade in the plan, yes or no?

You mark it in the moment, not later. Yes means the setup was written down before the chart opened and every condition was met. No means anything else: I chased it, I took it to make back the last one, I had been watching for two hours doing nothing.

Once you have fifty trades, filter that column and total the two halves separately. The answer is uncomfortable and almost always the same: in-plan trades are net positive, out-of-plan trades eat the entire profit. That number, in dollars, is what improvising costs you. No motivational speech lands the same way.

And once measured, it becomes a mechanical rule: if the month is running above 20 percent out-of-plan trades, size gets cut in half until it comes back down. Rules, not willpower. We build out that mechanism in how to stop overtrading and revenge trading.

What not to track

Things you can delete from your current template:

• Emotional state on a 1 to 10 scale. Not measurable and nobody is honest with themselves at 10 PM. If you want to capture some of that, use objective data: hours of sleep, or simply the trade count for the day, which correlates with bad days better than any self assessment.
• Long narratives. A paragraph per trade never gets read again. If you need a note, one line, ten words maximum.
• Daily dollar targets. Writing them down pushes you to force trades to hit the number. The target is clean execution, and the plan column already measures that.
• Indicators used. With a plan they are always the same. If they change weekly, the journal is not your problem.
• Market predictions. Writing "I think it goes up today" adds nothing your entry direction does not already show.

Screenshots: when and how

Screenshots help, but not all of them. Saving an image of every trade produces four hundred files you will never open.

The rule that works: capture two kinds of trade only. The ones that went perfectly by the plan (to recognize the pattern sooner) and the ones that were an execution disaster (to recognize the mistake sooner). Ordinary trades that win a little or lose a little teach nothing visually.

How to make them useful: capture after the trade is done, with entry, stop and exit visible, and enough bars of context to see what led into it. A screenshot of the entry moment with none of the aftermath teaches nothing.

One step further: replay the full session in Market Replay on the weekend and pause right at your entries. Seeing what happened five minutes before your click, without knowing the outcome, teaches more than twenty screenshots. Covered in NinjaTrader 8 Market Replay.

The twenty minute weekly review

Saturday morning, coffee, twenty minutes. Market closed, no rush. Four questions, in this order.

1. How many trades were out of plan and what did they cost? First because it weighs the most.
2. Which setup produced the most R this week and which one drained it? Sort by tag and total. Four weeks of data already shows a direction.
3. What time of day am I losing? Group by entry hour. It is common to discover that everything made between 9:30 and 11:00 AM ET gets given back after 2:00 PM.
4. What single rule changes next week? One. Written down. For example: "no trades after 2:00 PM" or "three trades a day maximum".

One rule a week. Changing five things at once makes it impossible to know which one worked, and that is the trap almost everyone falls into the first time they take journaling seriously.

Metrics that actually say something

With those eight fields you can calculate all of this in a spreadsheet. None of it needs paid software.

• Expectancy in R. Sum of all R results divided by trade count. If it comes out at +0.15R, you make 0.15 times your risk per trade on average. That single number summarizes your trading.
• Win rate by setup. Not overall, by tag. A setup hitting 40 percent at 3R average beats one hitting 70 percent at 0.6R.
• MAE on winners. Look at how far your eventual winners went against you first. If none exceeded 6 points and your stop sits at 20, you are risking triple what you need.
• Result by trade count per day. Group days by how many trades you took. Typical pattern: one to three trade days are net positive, seven or more are net negative.

Not worth calculating early: Sharpe ratio, recovery factor, correlations. Those are portfolio metrics and with eighty intraday trades they mean nothing.

Where to keep it: spreadsheet or app

Three options and one clear recommendation.

Spreadsheet. Google Sheets or Excel, eight columns, one pivot table. Free, yours, filters in two clicks and does not depend on anyone keeping a service alive. For 90 percent of traders it is the right answer, especially in year one.

NinjaTrader export. The Trade Performance tab exports to CSV. The workflow that sticks is exporting Friday, pasting into the sheet, and hand filling only the tag and plan columns.

Journaling apps. Some connect to your broker and import automatically. They are well built and save time, but run 20 to 40 dollars a month and hand you fifty charts of which you will use three. They make sense once you already have six months of discipline and data entry is genuinely the bottleneck. Not before.

Whichever you pick, one warning: a journal is not a finished product. Start with eight columns and add a ninth only when one specific question goes unanswered three weeks in a row.

The journal does not make you better because you write it. The Saturday review does, when you see in dollars what the out-of-plan trades cost and walk away with one new rule. Start this week with the eight columns and change nothing else until you have fifty trades.

trading journal what to track

Educational content, not investment advice. Trading futures involves substantial risk of loss and is not suitable for every investor. Any reference to Algomyth results is hypothetical, simulated over historical data, and shown in full on the performance page; past performance does not guarantee future results.