kanchi-dividend-sop

---
name: kanchi-dividend-sop
description: Use when screening, evaluating, or timing entries for US dividend stocks with a repeatable yield/safety/valuation/event process, producing verdict tables, one-page memos, and limit-order plans.
category: Portfolio
version: 1.0.0
tools: []
---

# Kanchi Dividend SOP

Deterministic 5-step Kanchi method adapted for US dividend investing. Prioritize safety and repeatability over aggressive yield chasing.

## When to Use

- Kanchi-style dividend stock selection adapted for US equities.
- A repeatable screening and pullback-entry process instead of ad-hoc picks.
- One-page underwriting memos with explicit invalidation conditions.
- A handoff package for monitoring and tax/account-location workflows.

## Inputs

Work from a ticker list: broker export, manual list, or candidates the user supplies grouped by bucket (core / satellite) and profile (income-now / balanced / growth-first).

## Workflow

### 1) Lock the mandate before screening
Collect and fix: objective (current income vs dividend growth), max positions and position-size cap, allowed instruments (stock only, or REIT/BDC/ETF), and account-type context (taxable vs IRA-like). Apply baseline thresholds unless the user overrides them.

### 2) Build the investable universe
Core bucket: long dividend-growth names (Dividend Aristocrats-style quality set). Satellite bucket: higher-yield sectors (utilities, telecom, REITs) kept in a separate risk bucket. Return the ticker list grouped by bucket before moving forward.

### 3) Step 1 — yield filter with trap flag
Step-1 yield is the **regular forward yield** = latest declared regular dividend × cadence-implied frequency ÷ price. Never use trailing-twelve-month yield — it lags the latest declared raise and silently bundles special dividends. Apply the profile floor to the regular yield only: income-now 4.0%, balanced 3.0%, growth-first 1.5%.

Trap and freshness controls:
- Special dividend detected → exclude it; report regular vs. TTM yield separately.
- Variable payout policy → FAIL (not a reliable income base).
- Dividend cut or suspension → FAIL.
- Payout freeze → HOLD-REVIEW (an income cash-cow exception decided only at final synthesis, and only if safety is otherwise clean).
- **Data Freshness Gate**: if the regular yield sits within ±0.20pp of the floor and the latest declared dividend isn't confirmed from an authoritative source, emit `STEP1-RECHECK` — never a hard FAIL.

### 4) Step 2 — growth and safety, sector-dispatched
Safety is sector-specific: a uniform GAAP/free-cash-flow payout triad mis-judges banks (FCF is meaningless for them) and regulated utilities (FCF is structurally negative there).
- Always compute the payout triad: GAAP-EPS payout, Adjusted-EPS payout, FCF payout. The safety verdict uses Adjusted-EPS + FCF for consumer names, or a sector-specific module for banks, utilities, and insurers.
- If the adjusted-EPS source is unavailable, cap the verdict at HOLD-REVIEW — fail-safe, never a silent PASS.
- GAAP-vs-Adjusted EPS divergence over 25% raises a one-off flag for Step 4.
- A merger completed within the last 4 quarters presumes GAAP EPS is distorted — force the adjusted path or HOLD-REVIEW.
- For regulated utilities, negative FCF alone is not an automatic FAIL — judge on FFO/debt, allowed ROE, rate-case posture, and equity-issuance risk.
- A mixed-but-not-broken trend classifies as HOLD-REVIEW, not a hard reject.

### 5) Step 3 — valuation, mapped to US sectors
- Financials: PER × PBR can remain the primary lens.
- REITs: use P/FFO or P/AFFO instead of plain P/E.
- Asset-light sectors: combine forward P/E, P/FCF, and the stock's own historical range.
Always report which valuation method was used for each ticker.

### 6) Step 4 — one-off event filter (backward-looking)
Reject or downgrade names whose recent profits lean on one-time effects: asset-sale gains, litigation settlements, tax-effect spikes, a margin spike unsupported by the sales trend, or repeated "one-time/non-recurring" adjustments. Record one-line evidence for every FAIL to keep the process auditable.

### 7) Step 4b — forward structural-event scan
Step 4 only looks backward; this step catches pending or recent structural events. For each surviving candidate, search for issuer and regulatory disclosures using a source hierarchy: issuer investor-relations page → SEC filing (8-K/10-Q/10-K/proxy/S-4) → exchange or company deck → reputable wire service → finance portals (secondary only).

- Only a major structural event caps the verdict to HOLD-REVIEW: a transaction over 10% of market cap, share issuance of 10–20%+, leverage rising 0.5x EBITDA or more, a control/listing/HQ change, a merger-of-equals, reverse merger, spin-off, large asset sale, a dividend/rating/leverage-policy change, sector-specific materiality events, or cumulative M&A over a rolling 24 months exceeding 15% of market cap. Minor bolt-on acquisitions are a caution note only.
- Pessimistic cap: if the scan comes back failed, skipped, or "no event found" on a name that triggered the Step-5 entry signal, cap it at HOLD-REVIEW and block the first buy tranche. Treat search being unavailable the same way — never a silent skip. A primary-source-confirmed "clean" result is stronger evidence than a search-only "no event found."

### 8) Step 5 — buy on weakness, with mechanical rules
Entry triggers: current yield above its 5-year average yield plus an alpha spread (default +0.5pp), or a target valuation multiple reached (P/E, P/FFO, or P/FCF).

Execution: split orders 40% → 30% → 30%.
- **Pre-order blockers**: if a candidate has any unresolved blocker from the earlier steps (variable/cut/suspended dividend, adjusted-EPS unavailable, GAAP/Adjusted divergence, bank credit concern, utility FFO/debt concern, event-scan failed or skipped, stale dividend data) or is otherwise blocked from Step 4b, the first tranche is blocked entirely or downsized to a ≤20% tracking tranche — never the full 40%.
- **Sector cluster risk**: when a large number of same-sector names pass together (e.g., several small banks sharing one macro beta), flag a portfolio-level cluster-risk warning.
- Require a one-sentence sanity check before each unblocked add: "thesis intact vs. structural break."

### 9) Produce standardized outputs
1. A screening table with an actionable verdict tier per ticker — CLEAN-PASS, PASS-CAUTION, CONDITIONAL-PASS, HOLD-REVIEW, STEP1-RECHECK, or FAIL — synthesized from Steps 1, 2, and 4b plus any blockers, with one-line evidence per row.
2. A one-page stock memo per ticker with a provenance block: price/dividend/payout/event sources, unresolved blockers, and evidence references.
3. A limit-order plan with split sizing, blocker gate, and invalidation condition.
4. A run context header (profile, yield floor, safety bias, universe source, excluded asset types) so results from one profile are never silently reused under a different one.

## Cadence

- Weekly (15 min): check for dividend and business-news changes only.
- Monthly (30 min): rerun screening and refresh order levels.
- Quarterly (60 min): deep safety review against the latest filings and earnings.

## Handoff

Hand outputs to a dividend-monitoring workflow for ongoing anomaly detection, and to a tax/account-location workflow for placement and classification planning.

## Guardrails

- Never issue a buy call without Step 4, Step 4b, and the safety checks.
- Never treat high yield as value before validating coverage quality.
- Use the regular forward yield for Step 1, never a special- or TTM-inclusive figure; near-floor-and-unconfirmed means STEP1-RECHECK, not FAIL.
- A failed or skipped event scan on a triggered name means HOLD-REVIEW plus a blocked first tranche — never silently skip Step 4b.
- Keep assumptions explicit; missing adjusted-EPS or other required data fails safe to HOLD-REVIEW, never a silent PASS.