Run your portfolio like an income operator.

The letter is computed, not written.

CaptureGrade™ is a deterministic letter grade, A through F, computed by CaptureDiv's scoring engine from filed and vendor-supplied figures — payout ratio, cash coverage, dividend record, and balance-sheet leverage — and ranked against the company's own peer group of the same security type. It is not produced by artificial intelligence, it is a rank rather than a prediction that a dividend will or will not be cut, and it is not investment advice.

Where the filings carry the evidence, a dividend payer is scored on 22 factors drawn from filed figures — payout ratio, coverage, growth streak, leverage, share count and the rest. Where they do not, we publish no letter rather than a guess. No model writes the grade. Given the same filings, the same inputs produce the same letter.

A bank is not scored like a REIT.

11 security types carry their own lens — which factors apply, which are meaningless, and what a good reading looks like. Three of the biggest groups we cover:

• Bank vs. operating company. A bank is measured on equity to assets; an operating company on net debt to EBITDA. Deposits are funding, not borrowing, and a bank reports no EBITDA — so the industrial measure has nothing to read.
• REIT vs. operating company. A REIT's GAAP payout ratio is inflated by depreciation — a non-cash charge on buildings that generally appreciate — so reported earnings understate what the REIT can actually pay from. Payout ratio is therefore not scored for REITs. Invitation Homes reports 121% of earnings, unscored, and grades B →
• BDC. Distributions come from net investment income, while reported earnings swing with unrealised marks on the loan book. Leverage is not read off debt to equity either: a BDC's binding constraint is the asset coverage requirement of the Investment Company Act of 1940. See the measures we drop for Ares Capital, and why →

Absolute floors sit above the ranking.

Hard caps apply after the ranking, so a name that looks defensible against its cohort is still capped on its own numbers. Paying out more than free cash flow caps at D no matter how the rest of the field scores.

A percentile sits beside the letter.

Two Dividend Kings can both be C and not be the same investment. The percentile says where each one landed against its own security type — REITs ranked among REITs, banks among banks — so a high C and a low C are visibly different.

And it refuses — in writing, per name.

Where the evidence is not there, no letter is published, and the row names the measure that is missing. Whole security types we do not rate — funds, partnerships, royalty trusts, preferreds — say so as a stated decision, by type, not as a blank. Read the reason we publish no grade for Barclays → No sign-in required.

Dividend safety ratings are common. A rating you can reproduce from the inputs, that scores a bank on bank measures, that tells you by name which measure it is missing when it declines to rate, and that says plainly which security types it does not rate at all — that is what the CaptureGrade is.