A market-scenario journal: test an idea without hindsight
Save the available facts, hypothesis and invalidation condition before you see the next part of the chart.
A useful entry lets you reconstruct what was known at the time. It separates observation from interpretation. One successful example does not establish a strategy’s edge, and a journal does not guarantee profits.
Freeze the information available
Record creation time with timezone, symbol, interval, provider and the last completed candle’s time. Save the input file or its checksum. List observed facts separately from the hypothesis you want to test.
For economic observations, also retain the release/version and both observation periods. A later revision can change both values: see how to compare data versions without mixing releases. Compare economic-data versions →
Planning the check in advance borrows a transparency principle from scientific preregistration. This local template is not a formal research registration or a validated trading method. Record hypothesis changes in a new dated row rather than overwriting the original.
One entirely invented scenario
Use the candles from the OHLC article. At 13:05 UTC, only completed intervals through 13:00 are available, including the last 12:00–13:00 candle. Facts: overall range 99–110 and last close 106.
Observation hypothesis: a future completed hourly candle will close above 110, prompting a review of the scenario. Invalidation: a completed close below 99. These are educational criteria, not buy or sell orders. In a separate invented continuation, a close of 98 invalidates the hypothesis.
Our own teaching example; all figures are hypothetical.
| Field | Entry at 13:05 UTC | Later review |
|---|---|---|
| Available facts | Through 13:00: H=110, L=99, C=106 | New completed candle C=98 |
| Rule | Close >110; invalidate below 99 | 98 <99: hypothesis invalidated |
| Action | Observe and retain the original entry | Append the outcome; preserve the past |
Review without rewriting the original
Later append review time, new facts, status and reasoning. If a condition never occurs, do not label the absence of a signal an error. If invalidation occurs, do not move the threshold retrospectively to make the example look successful.
The CSV below is a plain local table without macros or formulas; this website receives none of its contents. Open it in your spreadsheet and replace the example row. Then follow the first-calculation guide: compare inputs and Pro/Light main and alternative scenarios, and record your own conclusion separately.
If you change the original threshold from 99 to 97 after seeing the outcome, is it the same advance test?
No. It is a revised hypothesis. Keep the old entry and create a new one with the change time.
Sources
Sources checked: 2026-10-20
This material and its translations were prepared with AI assistance. Educational calculations were checked programmatically. Sources explain concepts; examples and files were created for this article.
Related reading
Next step
The first-calculation guide covers Pro/Light differences, requirements and checking results. A program scenario still requires evaluation.
First calculation in Pro and Light →