Last Friday, April 17 2026, I asked Claude one question: "Is Coca-Cola's dividend safer than Realty Income's right now?" Ninety seconds later I had a side-by-side table of yields, payout ratios, credit ratings, and a one-line "biggest risk" for each. I did not open a single 10-K. This article is the exact workflow I used, run on three real US dividend stocks (Johnson & Johnson, Coca-Cola, and Realty Income), with the actual Claude outputs shown in full. Whether you are building a retirement income portfolio or just tired of yield-chasing into dividend traps, this is the fastest way I have found to pressure-test a dividend before you buy it.
Most dividend screeners give you yield. That is the trap. A 7% yield on a stock about to cut looks identical to a 7% yield on a stock about to raise, until the cut hits and the share price drops 30%. What you actually need is payout ratio, free cash flow coverage, credit rating, dividend growth streak, and a real answer to "what scenario breaks this?" That is exactly what Claude can give you in a single prompt. Use this workflow on any US dividend stock you are evaluating.
Time from blank chat to three-stock comparison table: under 2 minutes.
Why dividend research breaks most investors
Dividend investing looks simple from the outside: find a high yield, collect the checks. The reality is that the three numbers that actually matter (payout ratio, free cash flow coverage, and credit rating) are buried across three different documents (10-K, 10-Q, S&P report). Comparing them across two or three companies side by side takes an afternoon of spreadsheet work. Nine out of ten retail investors skip it and buy on yield alone. That is why "dividend traps" exist: stocks where the yield is high precisely because the market already knows the dividend is unsafe.
The workflow below fixes that. Claude reads, compares, and flags in one shot. You then spot-check two or three numbers against the actual filings before you place an order. The AI does the grunt work. You keep the judgment.
The 6-step workflow for dividend research with Claude
This is the exact framework I use for every dividend stock I consider, and it works for any US-listed dividend payer, including Dividend Kings, Aristocrats, REITs, BDCs, or a single stock you are curious about. Each step builds on the last.
Step 1: Define what safe means to you
Before you prompt Claude, answer one question for yourself: are you buying for income today or dividend growth over decades? A retiree wants a high, stable yield with a low payout ratio. A 30-year-old building a compounding machine wants a lower yield but 8%+ dividend growth. Tell Claude which camp you are in so the analysis is weighted correctly.
Step 2: Feed Claude the basics in one paragraph
Name the tickers, the sector, your time horizon, and what you care about most. Thirty seconds of setup saves you from generic output. The more specific the brief, the sharper the answer.
Step 3: Ask for the safety table
This is the single most valuable prompt in the entire workflow. Do not ask "which is better?", ask for a structured comparison table with the seven rows that actually determine dividend safety: yield, payout ratio, dividend streak, 5-year dividend CAGR, free cash flow coverage, credit rating, and biggest risk.
Step 4: Run the stress test
A comparison table tells you what the dividend looks like today. A stress test tells you what kills it tomorrow. Ask Claude for the one scenario most likely to force a dividend cut in the next five years, plus a specific quarterly-report metric to watch and a "yellow line" threshold. This is the step most investors skip, and exactly where dividend traps hide.
Step 5: Spot-check two numbers
Claude can hallucinate a decimal point. Pick two figures from the table (usually yield and payout ratio) and verify them on the company's investor relations page or a primary data source before you act. If both match within a percentage point, the rest of the table is almost certainly directionally correct. If either is materially off, re-run the prompt with the correct baseline.
Step 6: Make the buy-or-skip decision
With a comparison table, a stress test, and two verified anchor numbers, you now know more about these three dividends than 95% of retail investors who own them. Rank them by the metric that matters most to you (safety, yield, or growth), note the watch-items for future quarters, and size your position accordingly.
Real example: comparing JNJ, KO, and Realty Income
Here is what came back when I fed Claude the exact prompt from Step 3, asking for a side-by-side dividend comparison of Johnson & Johnson, Coca-Cola, and Realty Income with a 10+ year horizon and safety weighted over yield. Figures are approximate as of April 2026 and labeled as such. Treat them as a research starting point, not a trading signal.
The comparison table Claude produced
- Dividend yield: JNJ ~2.2%, KO ~2.7%, O ~5.0%
- Payout ratio (TTM, earnings-based): JNJ ~50%, KO ~75%, O ~275% (but for REITs use AFFO, and O's AFFO payout is ~74%)
- Consecutive years of dividend increases: JNJ 64 years, KO 64 years, O 31 years (monthly REIT)
- 5-year dividend growth (CAGR): JNJ ~5%, KO ~4.5%, O ~3%
- Free cash flow coverage of dividend: JNJ ~1.7x to 1.9x (very strong), KO ~1.3x to 1.4x (adequate), O ~1.35x AFFO coverage
- S&P credit rating: JNJ AAA (one of only two in the US), KO A+, O A-
- Biggest dividend risk (one line each): JNJ is talc litigation plus pharma patent cliffs; KO is currency exposure plus slow volume growth in developed markets; O is rate sensitivity plus tenant concentration
The REIT caveat matters. Realty Income's earnings-based payout ratio looks alarming at 275%, but REITs are required to distribute most of their taxable income and are judged on AFFO (adjusted funds from operations), not net income. On AFFO the payout is ~74%, healthy for the sector. Claude flagged this automatically. A yield screener would not.
JNJ is one of only two AAA-rated companies in the US. Microsoft is the other.
Claude's verdict
Claude picked JNJ as the safest for a conservative income investor. The three reasons: the AAA credit rating (lowest balance-sheet risk of the three), the ~50% payout ratio (earnings could drop meaningfully before the dividend is pressured), and the strongest free cash flow coverage. KO came second on brand moat and streak but runs a tighter payout. O offered the best yield but carries interest-rate and leverage sensitivity that JNJ simply does not have. The one caveat Claude added: JNJ's talc litigation is a known overhang, priced in, and S&P kept the AAA through it, but it is worth watching.
The stress test: what breaks each dividend?
The comparison table is step three. The stress test is step four, and it is where you separate real research from a glorified yield screen. I asked Claude for each of JNJ, KO, and O: the one scenario most likely to force a dividend cut in the next five years, the specific quarterly-report metric to watch, and a "yellow line" threshold where worry should kick in. The answers came back in bullet form, tight and actionable.
Johnson & Johnson
- Cut scenario: a catastrophic talc settlement above $30B plus a failed Stelara biosimilar defense, draining cash faster than Innovative Medicine can replace lost revenue
- Watch in quarterly report: free cash flow TTM vs. dividend paid
- Yellow line: FCF coverage drops below 1.3x (currently ~1.7x)
Coca-Cola
- Cut scenario: sustained USD strength (10%+) for 2+ years combined with a GLP-1-driven structural decline in sugary beverage demand in developed markets
- Watch in quarterly report: organic revenue growth ex-currency and ex-acquisitions, specifically unit case volume
- Yellow line: unit case volume goes negative for 2+ consecutive quarters, OR free cash flow drops below $10B annualized (the dividend costs ~$9B per year)
Realty Income
- Cut scenario: higher-for-longer rates (5%+ for 3+ years) force refinancing at punitive spreads while a major tenant segment (drugstores, dollar stores) sees multiple bankruptcies
- Watch in quarterly report: AFFO payout ratio plus occupancy rate
- Yellow line: AFFO payout ratio above 85%, OR occupancy drops below 97% (currently 98.9%), OR Net Debt/EBITDA above 6.0x (currently 5.4x)
JNJ's yellow line is the hardest to trip. Realty Income has the most yellow lines to monitor because REITs run closer to the edge by design.
That last line from Claude is the entire point of the stress test. You now have three specific numbers to scan for in each company's next quarterly filing. If any yellow line gets hit, you have a pre-agreed trigger to reassess, not a panic reaction to a headline.
The 6-step workflow at a glance
The reusable prompt template
Copy this prompt, fill in the brackets with your tickers, and run it on any US dividend stock pair or triple you are considering. It is the same prompt that generated the JNJ/KO/O comparison above.
- I am comparing [TICKER A], [TICKER B], and [TICKER C] for a long-term income portfolio. My horizon is [X years] and I care about dividend safety first, yield second.
- Build a side-by-side dividend comparison table with these rows: current dividend yield, payout ratio (TTM), consecutive years of dividend increases, 5-year dividend growth rate (CAGR), free cash flow coverage of the dividend, credit rating, and one-line biggest dividend risk for each.
- Use approximate figures and label them clearly as approximate.
- Then flag which one looks safest for a conservative income investor, and explain in 2-3 sentences why.
Follow it with the stress-test prompt from step four. Two prompts, one chat session, every dividend stress-tested the same way.
What to verify before you buy
Claude is a research accelerator, not a data oracle. Before you place any trade, verify these three anchor numbers against primary sources on the company's investor relations site:
- Current dividend per share and the most recent ex-dividend date
- Trailing twelve-month earnings per share (for payout ratio) and trailing FCF (for coverage)
- Latest S&P or Moody's credit rating, which can change quietly
If the numbers Claude quoted match these within a percentage point, the whole analysis stands. If any are meaningfully off, re-run the prompt with the corrected anchor and see whether the verdict changes.
What to watch next
Three things are worth tracking on these three names over the next twelve months. For JNJ, the outcome of the multi-billion dollar talc consolidation and whether S&P reaffirms AAA. For KO, whether GLP-1 drug adoption is visibly compressing developed-market volume in the quarterly cadence. For Realty Income, the trajectory of the 10-year Treasury yield and any uptick in tenant credit events among drugstore or dollar-store operators. If any of those move through a yellow line Claude named, the verdict changes and it is time to re-run the stress test.
The bottom line
Dividend investing rewards patience, but only if you pick dividends that survive. The ninety-second prompt at the top of this article, the one that started my Friday afternoon research, compressed what used to be a full day of spreadsheet work into a single structured answer. It will not tell you whether to buy. It will tell you whether a dividend is safer than its yield suggests, which is the question that actually matters. Use the six-step workflow on every dividend you are serious about, verify two anchor numbers before you act, and keep the yellow lines on a note next to your watchlist. That is how you stay out of dividend traps without giving up yield.
References
- Johnson & Johnson Investor Relations
- Coca-Cola Company Investor Relations
- Realty Income Investor Relations
- S&P Global Ratings
- Claude.ai