The framing that wastes the argument
"Manual journaling makes you engage with each trade" versus "manual journaling is friction you will abandon". Both sides are right, which is a reliable sign the question is wrong.
The useful split is not by method. It is by what kind of information a field holds.
What import does better than you
Not just faster — more accurately, and this part is rarely conceded:
- Prices and times. Memory rounds; the file does not. Your recollection of the entry as "around 1.0850" is worse data than the 1.08487 in the export.
- Costs. Commission and swap per trade, which almost nobody enters by hand and which change every downstream statistic. See net P&L.
- Completeness. Manual journals miss trades, and they miss them non-randomly: the ones you skip are the impulsive ones you would rather not write down, which are exactly the ones the record needs.
- Partial closes. A position scaled out three times is a bookkeeping exercise by hand and automatic in an import.
That last point about non-random omission is the strongest argument against manual-only journals, and it is structural rather than a matter of discipline. The trades a manual journal loses are the trades that would have told you the most.
What import structurally cannot hold
Why. No file records the reason. If it was not written before the outcome, it does not exist — and reconstructing it afterwards produces a flattering story every time.
What you expected. Related but distinct: the level you were watching, what would have invalidated the idea.
Whether it matched your criteria. The file shows a trade. It cannot show whether it was your setup or something you talked yourself into.
Your state. Not in any export.
Anything about execution — slippage, spread at fill, requotes. Not in the file either, and not something you can enter by hand: see what your statement hides.
The arrangement that works
Import the facts. Write one line of intent.
The line is written before the outcome is known — that is the whole requirement, and it is what makes it evidence rather than narration. One sentence naming the setup and the reason takes ten seconds at entry and cannot be reconstructed later at any price.
Everything else the machine does: instruments, sizes, times, prices, costs, partial closes, and the stop and target where the platform carries them.
Where each fails
Manual-only fails on completeness and accuracy, and it fails hardest exactly where it matters — the impulsive trades go unrecorded, so the journal describes a more disciplined trader than the one keeping it.
Import-only produces a perfect record of what happened with no trace of what was meant. It can tell you that your busy days cost you money; it can never tell you whether you knew you were doing it. That is why every analyzer on this site stops at describing behaviour and hands the interpretation back — the file has no other half.
A practical note on rebuilding history
If you are starting a journal now, import your history first. Months of accurate facts arrive in one step, and they are enough for every question about behaviour and sample size.
Do not try to reconstruct intent for past trades. Whatever you write today about a trade from March is fiction, and mixing it with the real entries you start writing tomorrow contaminates the only field that was ever worth having.