The Print

Blackstone Mortgage Trust (NYSE:BXMT) earned its dividend last quarter. Starwood Property Trust (NYSE:STWD) did not. Two commercial mortgage REITs, two yields near 10%, and a coverage ratio that points the wrong way.

In Q1 2026, BXMT generated distributable earnings prior to realized gains and losses of $0.49 per share against a $0.47 dividend — its third consecutive quarter of coverage, roughly 104%. The yield sits near 9.4%.

Starwood reported distributable earnings of $0.39 against a $0.48 dividend it has held for more than a decade — coverage of about 81%. Even on management’s adjusted figure of $0.47, it still lands a penny short. The yield runs above 10%.

Take only those two numbers and BXMT wins. But coverage at a mortgage REIT is measured before the part that actually moves book value — and that is where the two companies separate.

What The Coverage Number Hides

BXMT’s coverage is calculated prior to realized gains and losses. Include them, and full distributable earnings for the quarter were $0.21 per share, against a GAAP net loss of $0.04. The difference includes a $46 million realized loss on the foreclosure of a San Francisco hotel loan, plus a $55 million CECL provision — roughly 80% of it specific to individual loans, including an impaired studio and a Dallas multifamily asset. Book value fell 2.7% to $20.20, and total CECL reserves now stand at $1.80 per share.

The dividend was covered — but the buffer behind it, book value, shrank in the same quarter.

The important number is not the reserve itself. It is what sits behind the reserve once it is used. A reserve is a shock absorber. The question is what happens after the shock arrives.

For BXMT, the answer is structural. It has reduced office exposure aggressively and built a liability structure designed to withstand stress. But none of that changes the central fact: it remains a single-engine lender. When a loan resolves at a loss, the damage ultimately lands in one place — book value. And the dividend was already reset once, from $0.62 to $0.47 in 2024. Today’s coverage is being measured against a payout that has already been recalibrated.

To BXMT’s credit, the liability side remains unusually resilient. Roughly 86% of its financing sits in non-mark-to-market structures, and the company faces no corporate debt maturities until 2027. The pressure is not funding risk. It is asset-resolution risk.

Same Cushion, Different Places For The Loss

Here is the number that reframes the comparison. Starwood carries almost exactly the same reserve. Its $676 million of reserves — $455 million of CECL plus $221 million in REO — work out to $1.82 per share against undepreciated book value of $18.97, within two cents of BXMT’s $1.80.

The reserve cushions are almost identical. The structures standing behind them are not.

Start with why Starwood missed. Management frames the shortfall as self-inflicted rather than credit-driven: cash parked at low yields, the resolution of roughly $900 million in non-accrual and REO assets it expects to clear in 2026, and the dilutive ramp of a newly acquired net lease platform. Whether that framing holds depends on whether those assets resolve near the marks Starwood is carrying. But the segment detail makes the structural point on its own: the commercial and residential lending arm produced $0.45 per share of distributable earnings, with another $0.15 from investing and servicing, $0.08 from property, and $0.06 from infrastructure, before $0.35 of corporate-level costs brought the total to $0.39.

The miss sits below the operating segments, not inside them.

That distinction matters. A credit loss weakens an engine. A corporate-level investment decision reduces what the engines produce. One reflects asset performance. The other reflects capital allocation.

And Starwood does not have one engine. It has five — including an investing and servicing arm anchored by LNR, the only special servicer in the country carrying Fitch’s highest CSS1 rating, a countercyclical business that tends to earn more precisely when the lending book is under stress. A pure-play lender has no equivalent. The leverage reflects the same gap: Starwood ended the quarter at 2.59x debt-to-undepreciated equity against BXMT’s 3.7x — lower leverage against a far less concentrated book.

When losses emerge, the two structures absorb them differently. BXMT absorbs the loss in book value. Starwood has four other places for it to land before book value is the one that takes the hit.

BXMT’s dividend is covered. Starwood’s is not. But coverage measures whether a dividend is paid. It does not tell you where the next loss goes. On that question, the structures are fundamentally different.

The yields are nearly identical. The loss-absorption mechanisms underneath them are not.

Source: Blackstone Mortgage Trust and Starwood Property Trust Q1 2026 earnings releases and earnings calls (April 29 and May 8, 2026); company investor relations.

The author holds no position in any security mentioned. Generalized research, not personalized investment advice.

Read the weekly structural income letter at jungmoku.substack.com.

Benzinga Disclaimer: This article is from an unpaid external contributor. It does not represent Benzinga’s reporting and has not been edited for content or accuracy.