Crypto trading journal — what to actually log beyond PnL

Most retail traders who keep a journal stop using it within three months. The reason isn't lack of discipline — it's that they're logging the wrong things. PnL, entry, exit, and a vague comment don't help you improve because they don't capture the variables that would let you find your real edges and your real leaks. Professional desks log entirely different fields. This article walks through the 12 that matter and the 6 that don't.

Why most journals stop being useful

A typical retail trading journal entry looks like this:

• Date: 2026-05-12
• Symbol: SOLUSDT
• Side: Long
• Entry: $148.20
• Exit: $151.80
• PnL: +$120
• Comment: "Took a long after BTC pump, exited at resistance"

Why is this useless three months later? Because the only structural information you have is "I made $120 on a SOL long after BTC pumped." You cannot answer any of the questions that would actually help you improve:

• How big was your conviction? Did you size this like a high-confidence setup or a maybe-trade?
• How long did you hold? Was the exit on a plan or on emotion?
• Did you check your stop after entry, or move it as the trade went against you?
• What setup pattern were you actually trading? "Long after BTC pump" is not a setup — it's a reason. The setup is the specific structural condition you saw.
• How much of your account did you risk? How does this trade compare to your other trades in the same setup?

None of these are captured. Your journal becomes a list of outcomes with no diagnostic information attached. That is exactly why three months in, the journal stops being interesting.

The 12 fields that correlate with improvement

1. Setup tag

The specific pattern you identified. Not "long" — "compression breakout on falling-funding setup" or "monitoring-tag fade on T2 coin" or "post-dump bounce on cascade with volume confirmation." Tag-based journaling lets you slice your trades by setup type six months later and see which ones actually pay.

2. Conviction score (1-5)

How sure were you when you entered? 1 = "I'm taking this for boredom." 5 = "I would size up if margin allowed." This is the single most valuable field for diagnosing leaks. If your 1-conviction trades have similar PnL to your 4-conviction trades, your conviction is uncalibrated and you have a meta-problem to fix before any single-trade improvement matters.

3. Initial risk in $

Not position size — risk. The dollar amount you would lose if your initial stop got hit. If you don't have a defined initial stop, log the level you would close at if the trade went obviously wrong.

4. Initial stop level

The exact price at which you planned to exit on the loss side. Critical for the next field.

5. Did you move the stop?

Yes/no, with timestamp if yes. The single biggest behavioural leak among retail traders is widening losing stops. If you log this consistently, you build a dataset on your own discipline that no amount of self-perception would give you.

6. Time in trade

Minutes from entry to exit. Combined with setup tag, this tells you if you're holding your winners and cutting your losers (good) or cutting your winners and riding your losers (bad).

7. Exit reason

Categorised: "TP hit", "SL hit", "manual exit at level", "manual exit on emotion", "time-cap", "external event". Not free-text. Categorised so you can group later.

8. Maximum favourable excursion (MFE)

The best unrealized PnL you saw during the trade. If your trades regularly show MFE much higher than realised PnL, you are exiting winners too early. The opposite pattern (MFE close to realised) means you're holding to plan.

9. Maximum adverse excursion (MAE)

The worst unrealized PnL you saw during the trade. If your winning trades regularly have large MAE, you're getting lucky on entries — the moves are working in spite of poor entry timing. This won't keep paying.

10. Market regime at entry

One word: "trending up", "trending down", "chop", "volatile". Doesn't need to be precise. The point is to identify if your setup tag works only in certain regimes — common pattern for swing traders.

11. Number of concurrent open positions

You might be a great trader in isolation but a poor portfolio manager. If your 1-position trades win 60% and your 5-position-open trades win 40%, you have a concentration problem.

12. Mental state (1-5)

How focused were you when you took the trade? 1 = distracted, tired, frustrated. 5 = rested, clear, present. Almost every trader can identify a category of trades they took in poor mental state that they would not have taken with clear judgment. Logging this surfaces the pattern.

The 6 fields that don't help

Entry candle pattern

"Engulfing bullish" or "doji" or "hammer." These are too granular to aggregate and too fuzzy to validate against. Use setup tag (#1) instead.

Indicator readings at entry

RSI 32, MACD bullish cross, etc. If these were part of your setup tag, you've already captured them. If they weren't, logging them is just visual clutter.

"Confidence" as a percentage

"I was 70% confident." This sounds rigorous but it isn't — your perceived 70% on Tuesday isn't comparable to your 70% on Friday. Use 1-5 conviction (#2) instead, where the granularity is matched to your actual reliability.

News context

"BTC ETF approved" or "Fed minutes released." Useful for the headline trades but not as a daily field. For most setups the news context is the noise, not the signal.

Free-text comment

The thing that kills most journals. You write "BTC was pumping, I felt good about this one." Three months later you can't aggregate or filter on this. Use structured fields instead.

Win/loss as a feeling

"Good trade" or "bad trade." Subjective tag that drifts with mood. Use the structured outcome fields and let the data tell you which were good.

How to actually structure post-trade review

Daily review (5 minutes): scan today's trades. Note any where conviction-score was high but outcome was bad — these are the most-diagnostic trades.

Weekly review (30 minutes): aggregate by setup tag. Which setups paid this week? Which lost? Are any setups systematically negative-EV? If so, stop trading them.

Monthly review (90 minutes): aggregate by setup tag × regime × conviction. The cross-tab is where you find your real edges and real leaks. Conditional WR (e.g., "compression breakouts in trending-up regime, conviction 4+") often reveals patterns invisible at the daily level.

The honest math on journal value

Most retail traders who start a structured journal see no improvement for the first month. By month 3, they typically have enough data to identify one large leak (most often: oversized losing trades that violated stops). By month 6, the cross-tab analysis surfaces 2-3 real edges and 2-3 real leaks. Eliminating just the largest leak usually adds more EV than discovering any single new edge.

Cumulatively, traders who keep a structured journal for 12 months and act on the analysis improve their risk-adjusted returns by a measurable amount — typical reports are 30-80% Sharpe improvement in year 2 versus year 1. This is not because journaling teaches you new patterns; it teaches you about yourself.

The systematic alternative

The unavoidable truth: most retail traders won't keep a journal for 12 months even after reading articles like this one. The discipline cost is too high, and the payoff is too delayed.

If you want the benefits of journal-driven improvement without the friction, the alternative is to outsource the journal-required parts. Systematic strategies eliminate the conviction-calibration leak, the stop-moving leak, the mental-state leak, the concentration leak, all in one move — the strategy executes the rules. Your job becomes capital allocation, not trade-by-trade discretion.

Our Pro feed is built for exactly this: defined setups, defined sizing, defined exits, delivered to your Telegram and executable through your Binance API. You don't have to keep a journal of conviction scores because the strategy's conviction is the same on every signal of the same type. You don't have to log stop moves because the stops are absolute. The leaks the journal would have caught are designed out.