Trading journal fields that actually get reviewed

Most journals fail because the fields capture what is easy to type, not what is useful to review. A row that asks for a mood score from 1 to 10 and a free-text paragraph fills up for two weeks and then quietly stops. The journals that survive are the ones where execution data is captured automatically, intent is one sentence, and reflection is two checkboxes. Everything else is friction that competes with the next trade.

The fields that matter

A journal earns its keep with roughly a dozen fields, no more. Group them by who fills them in: the system, you before the trade, you after the trade.

Auto-captured execution data. These come from the broker fill data and the price feed. You never type them.

• Entry price and timestamp
• Exit price and timestamp
• Instrument and contract month
• Position size and direction
• Fees and commissions
• Realized P&L (in money and in ticks)
• MFE, MAE, MCE (in money)
• Holding duration

Pre-trade intent. Filled in the moment the order goes live, in under thirty seconds. If it takes longer, you will skip it during a fast tape.

• Setup name from your playbook
• One-sentence thesis (why this, why now)
• Planned stop and planned target (price levels)
• Planned R-multiple

Post-trade reflection. Filled at close, also under thirty seconds. Two checkboxes and one line.

• Rule adherence: yes or no
• Deviation reason (only if no): one of a fixed list (chased, moved stop, sized up, fat-fingered, ignored news)
• Annotated screenshot
• One-line note (what would you do differently)

Twelve to fifteen fields per trade. Anything beyond this list is decoration.

What to leave out

Mood scores from 1 to 10 do not survive contact with a five-trade morning. You stop being honest about them by the third entry of the day, and the column becomes noise that pollutes any correlation you try to run on it later. If emotion matters, capture it as a binary rule deviation ("traded outside plan because tilted") rather than a number.

Generic tags like "good trade" and "bad trade" tell you what you already know from the P&L column. They are a vocabulary that hides the actual cause, which lives in the deviation reason or the setup name.

Multi-paragraph essays per trade are a different failure mode. They feel productive, they take fifteen minutes, and you stop after two weeks. The one-line note is enough. If a trade genuinely needs more context, it belongs in the weekly review as a case study, not in the row.

The principle is simple. Every field should either be auto-filled or take under ten seconds to fill. Fields that need a paragraph go in the review, not the row.

The three review cadences

A journal is a write/read system. Most traders only build the write side. The read side is where rule changes come from, and it runs on three cadences.

Per-trade, 60 seconds. Right after the close, while the chart is still on screen. Two clicks: rule adherence yes or no, deviation reason if no. One sentence. The screenshot is auto-captured by the platform or pasted in. That is the whole entry. Anything longer at this point and you will stop within a month.

Weekly, 30 minutes. Friday close or Sunday evening. Sort the week's trades by setup name. For each setup, look at the realized P&L next to MFE, and compute the realized / MFE ratio across the group (see Maximum Favorable Excursion for why this ratio is the cleanest read on exit quality). Flag one thing to change next week. One thing, not five. Write it at the top of the next week's plan.

Monthly, 2 hours. Full dataset, all closed trades. Compute hit rate, average R, and median realized / MFE per setup. Sort setups by expectancy. Kill the worst setup outright for the next month. Size up on the best. This is the cadence where the playbook actually changes shape, and it only works if the per-trade and weekly cadences have been feeding it clean rows.

Why automated capture matters

If you have to type MFE, MAE and MCE, you will not. The math involves scanning every bar between fill and close, on the right side of the position, and converting to money. Nobody does that by hand at 4pm on a Friday. Within two weeks the columns are empty, and the journal becomes a trade log again.

The split is structural. The system captures execution. You capture intent and reflection. The Tradeways journal computes the execution fields from 1-second OHLCV bars on close, so by the time you fill in rule adherence and the one-line note, MFE, MAE, MCE and realized / MFE are already on the row.

This split also protects honesty. Execution data cannot be edited after the fact, which means the weekly and monthly reviews look at the same numbers your broker has. Intent and reflection are yours to write, but they sit next to execution data that does not negotiate.

What kills journals

Three failure modes account for almost every dead journal.

Too many fields. Twenty-plus columns, half of them free text. Looks thorough on day one, sees no entries by day twenty. Cut to the twelve-to-fifteen field set and resist additions.

No review cadence. Rows accumulate, nobody reads them. The write side keeps running on momentum for a month or two, then stops because writing without reading has no payoff. Schedule the weekly and monthly reviews like a session, not like an aspiration.

No link from review to next week. The monthly produces interesting charts, the trader nods, nothing changes in the playbook. The loop is broken at the end. A review that does not produce one rule change written into next week's plan is a review that did not happen.

Fields are easy. The hard part is the loop. Keep the row short enough that you fill it, keep the cadence short enough that you read it, and keep the link explicit so that what you read changes what you trade.