New Investment Screener

New Investment Screener

Read-only. reference/READ-ONLY-POLICY.md (hard rule) and reference/RESEARCH-STANDARDS.md (peer framework, mandatory news on finalists, disclosure) apply. Steps below name capabilities — resolve each against reference/BROKER-CAPABILITIES.md for the broker(s) in BROKERS.md. The fundamentals/technical screeners and market movers are Groww-only; if Groww isn't active, say so and build the starting universe from WebSearch instead of dropping the screen.

Steps

1. Clarify the brief if it's vague: theme/sector, criteria (value/growth/quality/dividend/momentum), instrument type, and roughly how much they're deploying. State the criteria numerically before screening ("P/E < 20, ROE > 15%, 3Y revenue CAGR > 12%" — not "reasonably valued growth") so the screen and the shortlist can both be checked against the same bar. Don't screen blind.

2. Check what would actually help. Equity holdings capability (every active broker) plus PORTFOLIO-PLAN.md's target allocation and fund list (Groww returns no fund data; INDmoney's holdings/net-worth capability may, if active — unverified) — is there a real underweight this screen should target, or is the user exploring?

3. Screen. Fundamentals-screener capability against the stated criteria; symbol search / market-movers capability for an open-ended starting universe. Technical setups (breakouts, momentum, oversold) can't be screened for — Groww's technical screener is down — so screen fundamentally or by mover list first, then read the setups off the technical-indicators capability over the shortlist, 10 names per call. ETFs: symbol search filtered to ETF, plus web-sourced fees and tracking error, since Groww's ETF screener errors (see Tool availability).

4. Filter out bad fits. Drop anything that would breach the plan's concentration limits, sits on its exclusions list, or that the decision log shows the user already declined. Size the shortlist to deployable capital and the preferred deployment style, not to a round number. Flag heavy overlap with existing holdings — the same exposure in a different wrapper.

5. Rank the shortlist (3–7 ideas) with reasoning per pick, not "screener said so". Pull single-name fundamentals for the finalists, then run each through the full multi-timeframe technical framework in RESEARCH-STANDARDS.md (trend/momentum/volatility/volume, not one indicator) before it earns a spot — a name that screens well fundamentally but is in a broken technical structure gets that noted, not silently dropped or silently ranked top.

6. News on the finalists (mandatory). One WebSearch per shortlisted name for anything that would change the pick — results just out, a downgrade. Finalists only, not the screened universe.

7. Present: shortlist table (name, the metrics matching the brief, one-line rationale, material news), then a short "why not others" if useful, then the disclosure block. These are ideas for the user to evaluate and place themselves. Formal version: reference/templates/new-investment-screener.md — worth it for several candidates, not for a two-name shortlist.