Build a dated screening research note: a worked template
A useful note records the result date, the inputs you actually have and the checks that remain unresolved. The example below follows Arrie's published universe, Stage-2, base-quality, pattern and proximity descriptions, while keeping a personal worksheet separate from product functionality. [R][U]
Use one consistent result date
Write the screen/result date separately from the date you take notes. If you do not know an input's date, record it as unknown; do not combine values from different dates and present them as one observed run. This is a proposed documentation practice, not a claim about an Arrie data-export feature.
A fictional completed note
Illustration only. “Candidate A” is not a real ticker and every value below is invented. No screening engine was run for this example.
| Note field | Fictional entry | What it allows you to say |
|---|---|---|
| Result date / note date | 20 November 2026 / 21 November 2026 | The two dates are distinct, rather than an implied live observation |
| Universe inputs | Main board; HK$35bn capitalisation; HK$4 last price; HK$20m 20-day average turnover | These supplied inputs meet the three published universe gates [U] |
| Stage-2 score | 9/10, supplied for the exercise | Meets the stated score gate; individual score components are not independently calculated [R] |
| Base inputs | ATR% 1.1% against own comparison median 1.3%; depth 18%; trigger HK$4.80, 50-day moving average HK$4 | 1.1<1.3; 18≤20; 4.80≤1.25×4=5.00; these supplied comparisons meet the written base limits [R] |
| Pattern evidence | No swing or breakout-volume record supplied | Pattern qualification remains unresolved [R] |
| Proximity evidence | No product-calculated proximity value or definition supplied | Do not assert compliance with the documented 13% limit [R] |
| Follow-up | Obtain dated pattern/proximity evidence and relevant definitions | A research task, not an order or a conclusion that A qualifies |
The completed rows do not cancel the missing rows. This worksheet cannot establish that Candidate A belongs in the final screen.
Keep calculations reproducible
For the trigger-to-moving-average comparison above, the published ceiling is 1.25 × HK$4 = HK$5. The illustrative trigger of HK$4.80 is below that ceiling. This arithmetic only checks supplied numbers; it does not validate how the trigger or moving average was generated.
For any comparison, retain the input, unit, period and source. The public rule page describes a proximity limit but does not specify its full calculation in the brief paragraph; do not invent a denominator to fill the missing field. [R]
Finish with an explicit unresolved section
A useful closing line is: “Universe, supplied score and supplied base comparisons checked against the public text; pattern and proximity not established.” It says more than a generic “looks good” without pretending the whole process was verified.
Use the pattern comparison for the published volume tests and the four-stage overview for context. This personal research-note template is educational, not an Arrie feature or investment recommendation.
Sources and review scope
Rules checked against current public publisher documentation on 22 November 2026. All worked examples use invented inputs; arithmetic was checked. This review does not independently verify production execution, current stock qualification, model accuracy or trading outcomes. AI-assisted writing, translation and editorial review.