Manhattan Bridge Capital: A 10% Dividend Our Model Won’t Buy

Forget the name — this is not a bank and it owns no buildings. It is a bridge lender: a house-flipper finds a property, needs cash in days rather than the weeks a mortgage takes, and pays a high rate for the speed. Manhattan Bridge writes the cheque, takes the house as collateral, collects interest for about a year, and gets its money back when the flipper sells or refinances. Three things make that work.

The profitability is real: about 59 cents of every revenue dollar reaches net income, on a 11.6% return on capital and a 12% return on equity. Those are honest numbers for a lender running at just 0.45× leverage. But look at the fourth card.

Every stock on Finrys runs through a Graham-inspired 8-point quality checklist before any valuation. Manhattan Bridge passes seven — cheap on earnings, cheap on cash flow, high returns, no dilution, almost no debt. It fails exactly one, and it fails it by a whisker: profit is up 15.6% over five years against a 20% bar.

Here is the single most important thing to understand before you trust any of the ratios above. A lender’s risk does not live in its margins — it lives in its loan book. And Manhattan Bridge carries an allowance for credit losses of exactly zero. Not a small one. Zero. Management’s justification is a genuinely remarkable record: 1,340+ loans since 2007, not one completed foreclosure. On that history, zero is arguably honest.

But look at how the book stays clean. Of the $60.67M outstanding at year-end, about $39.81M consists of loans that are already past their original due date — including one that was due in 2016 — and are still counted as performing because, in the 10-K’s own words, the borrowers “have either signed an extension agreement or are in the process of signing” one.

This is what you are actually buying. The quarterly dividend is $0.11 a share — $0.44 a year, a 10.2% yield at $4.30. That is a genuinely high income stream, and it is paid in cash, quarterly, by a company with no meaningful debt.

But it was $0.115 until February 2026, when the board cut it. That is a small cut, and small cuts are easy to wave away — except that this one is the arithmetic catching up. For FY2025 the company declared roughly 103% of its reported profit as dividends. When the loan book shrinks, interest income shrinks with it, and a REIT paying out everything it earns has only one lever left.

Our model puts fair value at $3.77 — below the $4.30 price — and calls the stock Overvalued. On a 9.8× P/E, that sounds absurd. It isn’t, and the reason is worth understanding, because it applies to every lender you will ever screen.

Manhattan Bridge is the mirror image of the wide-moat names we usually write about. Adobe is a great business at a questionable price. This is a well-run business at a fair price — and the question is not whether it is cheap, but whether the income is durable.

On business quality, it is far better than its size suggests: 11.6% returns on capital, 0.45× leverage, no dilution, a founder with 22.8% of the shares, and a credit record — zero foreclosures in 1,340+ loans — that most lenders would kill for. On price, our model says you are paying about what it is worth: fair value $3.77 against a $4.30 price, with the 10-Cap landing almost exactly on today’s quote. What you get for that price is a 10.2% yield from a shrinking loan book, backed by a zero credit-loss allowance whose credibility rests on extensions the auditor has formally flagged. If the extensions are what management says they are, this is a solid income holding that will never make you rich. If they aren’t, the first loss lands straight in the earnings — and there is nothing set aside to catch it. Track it on Finrys, decide which of those you believe, and size the position accordingly.