TradePulse Research Playbook

What this page is

TradePulse is a research and ops checklist for market intelligence with AI assist — not a trading terminal, not financial advice, and not a real-time SaaS demo (that UI ships later).

Use this playbook to structure what you look at, what you ask AI, and how you log decisions before you talk to your broker, compliance team, or investment committee.

Disclaimer: MyGearHut provides operational workflows and prompt templates only. Nothing here is investment, legal, or tax advice. Past patterns do not predict future results. Verify all data with primary sources. Consult licensed professionals for financial decisions.

Built for operators tracking sectors they sell into, founders writing strategy memos, and teams who need a shared log — not day-traders or signal-chasers.

The weekly research sprint (90 minutes)

Run this on a schedule. Consistency beats chasing "real-time" noise.

Watchlist template

Maintain in Notion, Sheets, or plain Markdown:

TICKER: [symbol]
COMPANY: [name]
SECTOR: [e.g. cloud infra, fintech]
THESIS TAG: [growth / value / event / hedge / watch-only]
WHY ON LIST: [one sentence]
INVALIDATION: [what would remove it]
NEXT CATALYST: [earnings date, product launch, reg decision — verify manually]
LAST REVIEW: [YYYY-MM-DD]
SOURCE LINKS: [IR page, SEC EDGAR, primary news — you paste URLs]

Macro scan prompt

Paste into your AI tool after you attach or summarize your own news notes (RSS, earnings calendar, Fed release summary — do not ask the model for live prices).

You are a research assistant helping me write an internal market context memo — NOT trading advice.

I will paste my source notes from this week. Produce:

1. EXECUTIVE SUMMARY (3 bullets, factual tone)
2. SECTOR HEAT MAP — rank my watchlist sectors as tailwind / neutral / headwind with one-line rationale each
3. RISKS TO WATCH (max 5, with "so what" for operators)
4. OPEN QUESTIONS I should verify with primary sources
5. CONFIDENCE TAG per section: [high | medium | low]

Rules:
- Do not invent stock prices, EPS, or dates.
- Label every inference as INFERENCE.
- If my notes are insufficient, say what's missing.

MY NOTES:
[PASTE YOUR CLIPPED HEADLINES / BULLETS]

Single-name research brief prompt

Use one ticker at a time. Attach 10-K excerpt, last earnings call summary, or your bullet notes — never rely on the model's memory for financials.

Internal research brief for [TICKER] — operator audience, not investors.

Sections:
A. Business model in plain English (≤100 words)
B. Bull case (3 bullets) — cite only from MY ATTACHED CONTEXT
C. Bear case (3 bullets) — same rule
D. Key metrics to track (define each metric; no values unless in my context)
E. Competitive moat / fragility (qualitative)
F. Questions for the next earnings call (5)
G. What would change my thesis (invalidation triggers)

Do not recommend buy/sell/hold. End with: "Verify all figures at [SEC EDGAR / company IR]."

MY CONTEXT:
[PASTE FILINGS SUMMARY, YOUR NOTES, OR EXCERPTS]

Sentiment pass (headline annotation)

Goal: classify narrative drift, not predict price.

Classify each headline I provide:

| Headline | Sentiment (pos/neu/neg) | Topic tag | Operator takeaway (1 line) | Primary source needed? (Y/N) |

Rules:
- Sentiment = tone toward the company/sector, not market direction prediction.
- Flag clickbait or single-analyst notes.
- Do not aggregate into a "score" — I will judge manually.

HEADLINES:
[PASTE 10–20 HEADLINES WITH SOURCE NAME, NO URL REQUIRED]

Screening worksheet (manual + AI assist)

AI does not replace a screener. Use this worksheet, then ask AI to explain names you already filtered.

MY SCREEN CRITERIA (I applied these manually in [TOOL NAME]):
- Market cap: [range]
- Sector: [list]
- Profitability: [Y/N]
- Revenue growth YoY: [threshold — from my export]
- Other: [custom]

For each ticker in MY RESULTS LIST, one paragraph:
- What they do
- Why they passed my screen
- One risk the screen doesn't capture

Do not add tickers not on my list.

MY RESULTS:
[PASTE TICKER LIST OR TABLE EXPORT]

Decision log (paste weekly)

DATE / REVIEWER:
CONTEXT (5 bullets): —
WATCHLIST: added / removed / thesis updated
DEEP DIVES: [tickers]
SENTIMENT THEMES (not predictions): —
ACTION ITEMS (research only): verify [metric] on EDGAR; follow up [catalyst date]
NEXT REVIEW: [date]

Data hygiene

Primary sources first (SEC, IR, central banks). Timestamp every claim. Discard AI numbers you didn't supply. Separate price narrative from fundamentals. Run external memos past compliance.

Skip: "best stocks to buy" prompts, sentiment-as-signal, research without invalidation triggers.

Status

TradePulse interactive UI (watchlist sync, templated exports, team logs) ships later. This playbook is the full workflow today — not a trading bot, not financial advice. Weekly sprint for watchlists; daily only during events you define in advance.