Your broker already knows your entry price, your exit price, and what you made or lost. A trading journal that records only those three things is a duplicate database with worse formatting. It will never tell you why a month went badly, because the numbers it stores are outcomes — and for an overnight hold, the outcome is decided by one gap you did not control.
The fix is not more discipline. It is recording different fields. Below is the behavior that breaks most journals, what it costs, and the nine columns that replace it.
The behavior: grading the decision by the result
Open almost any trader’s spreadsheet and the emotional weight sits in one column: profit and loss. Green rows get read as “I was right.” Red rows get read as “I was wrong.” Everything else in the row is treated as decoration.
This is a documented cognitive failure, not a personal weakness. In Baron and Hershey’s 1988 study of outcome bias, subjects rated identical decisions differently once they learned whether the result had been good or bad — and kept doing so even after stating that outcomes should not affect the judgment. Knowing about the bias did not switch it off.
Overnight holds make this worse than usual. You place one decision, you sleep, and a single gap resolves it. There is no intraday sequence to review, no series of adjustments to grade. If your journal records only the result, you have logged the coin, not the call.
The cost: you cannot see drift until it is expensive
The practical damage is slow. A journal that grades outcomes cannot distinguish a sound process running through a bad stretch from a process quietly degrading — more trades, looser rules, later exits. By the time the account balance makes it obvious, months have passed.
The size of that drift is well documented. Barber and Odean’s study of 66,465 households at a large discount broker from 1991 to 1996 found the average household earned 16.4% annually against a 17.9% market return — while the most active traders in the sample earned 11.4%. Same market, same period. The gap was behavior, and it accumulated invisibly.
A journal that only stores results can tell you that you lost. It cannot tell you whether you followed your own plan while doing it.
The fix: nine fields for an overnight trading journal
Every field below is written before the position exists, except the three marked “after.” That ordering is the whole point — a trading journal filled in after the fact is a memoir.
| # | Field | Why it earns its column |
|---|---|---|
| 1 | Session type | Normal night, half day, or the night before a scheduled release. Gap behavior is not the same across these. |
| 2 | Thesis, one sentence | If it takes three sentences, you do not have one. |
| 3 | Signal and its strength | Whatever you use — a pressure reading, breadth, your own read. Record the number, not “looked good.” |
| 4 | Invalidation level | The price that proves the thesis wrong, written before the close. |
| 5 | Size as % of account | Percent, not share count. Share counts hide risk creep. |
| 6 | Worst-case gap, in dollars | What a 1% adverse open costs this position. Write the actual number. |
| 7 | Planned exit time | The clock time you committed to, decided while the market was closed. |
| 8 | Actual exit time (after) | The single most diagnostic field in the sheet. |
| 9 | Deviation note (after) | Anything you did that was not in fields 1–7, and the reason you gave yourself at the time. |
Why fields 7, 8 and 9 do the real work
Planned exit versus actual exit converts a vague feeling into a countable event. Ten rows where those two columns match is a process. Ten rows where they diverge is a different strategy than the one you tested, and you now have the count to prove it.
The deviation note matters because reasons are perishable. “Futures looked weak so I closed early” is obvious at 9:41 a.m. and unrecoverable three weeks later. Written down, those notes cluster fast — most traders find two or three recurring triggers, not twenty. The overnight-specific ones tend to involve checking futures at hours when nothing actionable can be done, which is a cost we have written about separately.
Field 6 exists because risk arrives as a gap, not as a stop. If writing the dollar figure makes you uncomfortable, the position is too large — the same conclusion the position sizing math for overnight holds reaches from the other direction.
The review cadence that makes it pay
- Nightly, sixty seconds. Fill fields 1–7 before the close. If you cannot, you are not ready to hold overnight.
- Weekly, deviations only. Read column 9 and nothing else. Count the entries. The count is the metric.
- Monthly, full read. Sort by planned-versus-actual exit rather than by P&L, and read the mismatches first.
- Quarterly, one question. Which recurring deviation cost the most, and what rule would have prevented it?
Sorting by P&L is the default in every spreadsheet and it is the wrong first sort. It leads you straight back to grading outcomes. Sorting by rule adherence surfaces the trades that were wrong before the market opened — the category that a closer read of your trade log is uniquely good at exposing.
Twenty rows is enough to see a pattern in the deviation column. It is nowhere near enough to judge whether your edge is real. Keep those two questions separate, and the trading journal stops being a chore and starts being the only instrument you own that measures you rather than the market.
This article is for educational and informational purposes only. It does not constitute financial advice, investment recommendations, or a solicitation to buy or sell any securities. Past performance and historical data do not guarantee future results. All trading involves risk, including the potential loss of principal. Always do your own research and consult with a qualified financial advisor before making trading decisions.
