Mortgage REITs: Read the Payout, Not the Mark-to-Market

Mortgage REITs: Read the Payout, Not the Mark-to-Market

Take a deep breath before you scroll past another double-digit yield from a mortgage REIT. You have heard the stories: the neighbor who bought one, watched its book value collapse when rates jumped, and swore off income stocks forever. The sector gets typed into one of two boxes — income machine, or capital destroyer — and the label sounds like it is stamped on the ticker. It is not. It is stamped on the payout. Two mortgage REITs can trade side by side at nearly the same yield, and one will be quietly filling your account with cash it actually earned from lending, while the other returns chunks of your own principal disguised as interest. The difference is not the yield. The difference is whether the money is produced by the business, or paid for by the balance sheet. That, and not the daily price, is what you are here to decide.

The Engine That Funds the Payout

Understand first what we are actually buying, because the mechanics explain everything that happens afterward. A mortgage REIT is a landlord of a different kind. Instead of owning buildings and collecting rent, it owns pools of mortgage loans and mortgage-backed securities and collects interest. The spread between what those mortgages pay and what it costs to borrow the money is thin, so the business runs on leverage — borrow short-term money cheap, buy longer-dated mortgage bonds that yield more, and pocket the difference after funding and hedging costs. That is the whole business model, and on the biggest agency names the borrowings are roughly eight and a half dollars for every dollar of equity — AGNC's balance sheet currently shows on the order of $109 billion of debt against about $12.5 billion of equity. Stack that leverage onto a modest 1.5-to-2-point spread and you have the machinery for a very fat dividend.

That same leverage is why the sector earned its "capital destroyer" reputation in 2022. The bonds sit on the balance sheet marked to market — repriced every day at whatever the market says they are worth — because, unlike an office tower, an agency mortgage-backed security has a continuous trading price. When the Federal Reserve raised rates with historic speed, bond prices sank and the stated book values of the sector followed. Investors who watched only that number concluded the machine was broken.

Here is the part that changes how you judge these companies: the mark and the cash are two different instruments. When rates rise, the bonds mark down — but the yields on the new mortgages the company can reinvest into rise right along with short-term funding. A year of ugly book value can be followed by a year of strong spread income. The honest scorecard for a mortgage REIT is therefore economic return — the dividend collected plus the change in book value — because that is the total of what the company actually delivered to you, cash in hand plus the real change in what each share owns. Investments that mark to market this way are objectively repriced, which is exactly why watching only one quarter fools people.

The Test: Coverage, Not Yield

So the practical question is not "is it a machine or a destroyer?" It is "how is it paying me?" Three checks answer it.

First, coverage. Is the recurring net interest income — the distributable earnings from the spread trade — enough to cover the dividend with room to spare? A payout ratio at or above 100 percent means the company is paying you out of capital, not cash flow, and part of that "yield" is your own money cycling back into your account. That is the first smell of a destroyer.

Second, which number moved. If book value fell but the dividend held, and book value later recovered, you lived through mark-to-market pain — real, and worth an honest pause, but not destruction. If book value fell and the dividend was cut to match, the cash-flow engine itself was impaired. The difference between those two outcomes is the whole ballgame, and it is visible long before anyone forces you to sell.

Third, the price-to-book tell. When a mortgage REIT trades persistently below book value, the market is telling you something, and the something is usually one of two things: payout that was never fully earned, or the risk of future equity issuance below book that would dilute your shares. Discounts of that kind are the single biggest structural warning in this sector right now. The two giant names have closed that discount, which is its own quiet signal that the market believes their payouts.

Two Names on the Income Side of the Ledger

Start with AGNC, the purest agency play in the group. It pays $0.12 a share every month, about $1.44 a year, which at the current price of roughly $10.90 is a run-rate yield north of 13 percent, and it has paid a monthly dividend through thirteen consecutive years.

Tangible book value came into the second quarter of 2026 at about $8.58 a share, and the company booked a 6.7 percent economic return for that single quarter — the dividend plus the book value gain. The full-year 2025 number was a 22.7 percent economic return, one of the strongest years the name has put up in a long time.

The first quarter of 2026 was the kind of thing that panics people who only watch the stock: AGNC reported a modest net loss for the quarter, and the mark on the books went the wrong way. The dividend never blinked, and the second quarter printed a gain on book value. If you had judged the company on the March number alone you would have feared a cut that never came and is not coming under current conditions.

Then Annaly, the largest mortgage REIT in the business with a market value still near $17.5 billion and twenty-four consecutive years of common dividends. In the second quarter it raised the quarterly payout to $0.75 a share — about $3.00 a year, a run-rate yield near 13 percent at the current $23.16 price — and here is the detail that matters: the raise was earned.

Annaly's distributable earnings for the quarter came in at $0.79 a share, above the new $0.75 dividend, with reported net income of $1.06, and the quarter delivered a 5.5 percent economic return. Management runs economic leverage around 5.6 times and pairs its agency book with residential credit and mortgage servicing rights, which behave differently under rates and give the machine a gear most of its peers lack.

A High Yield That Was Borrowing From You

Now the name that proves the labels run both ways. Orchid Island Capital carries a headline yield near 18 percent — but its trailing payout ratio ran close to 190 percent, meaning the dividend exceeded earnings over the past year and was being paid, in part, out of what the balance sheet was losing. Book value per share was around $7.08 at the end of the first quarter of 2026, and the company was running a net loss. In April management cut the monthly dividend from $0.12 to $0.10 a share, and the stock dropped about 7 percent in a single session as the market repriced exactly what that cut meant. The shares still trade a few percent below that book figure — around $6.70 against the $7.08 book value — while the one-year price return stayed negative despite the fat yield on the screen.

This is the moment I add the nuance, because it matters: Orchid's second quarter finally turned profitable, with net income of about $89 million, or $0.44 a share, which now sits comfortably above the reduced run-rate payout. That is the raw material for a genuine turnaround, and if book value starts rebuilding, this could become a different story. But the lesson stands on its own. The 18 percent yield was doing the work of handing back principal while book value fell. The mark-to-market was telling the truth; the screen was not.

The Honest Risk

I want to be straight about what I cannot know. Nobody can tell you where rates go next, and I can be fantastically wrong about timing. The current setup is genuinely favorable for the spread trade: the 30-year mortgage sits around 6.5 percent while the yield curve has steepened, agency spreads are in a range analysts consider reasonably tight, and the carry — what the bonds pay you minus what it costs to fund them — is positive.

Several observers expect solid double-digit total returns through 2026 if spreads keep behaving. The risks are the ones that always stalk this sector, and they are not the direction of rates so much as the path: a volatility spike, spreads blowing out, liquidity evaporating — the 2022 playbook in miniature. Fitch reminded investors in August that where discounts below book persist, mortgage REITs lose capital flexibility, meaning a future equity raise would dilute existing holders. That is precisely why the two giant names matter: trading at and around book value, they can defend capital and grow without handing you a handout to new shareholders.

What the Income Investor Does With This

Bring it back to the retiree problem, because that is the whole point. You are not investing in book value; you are funding a life. What you can bank is the dividend, and you can only bank it if the payout is covered by cash flow the business actually earns. So the portfolio action is simple to state. Hold the covered cash-flow machines near book value — the AGNC types and the Annaly types — whose payouts survive a quarter of mark-to-market pain, and watch the engine quarterly rather than the tape daily. When the price dips while coverage holds, that is the market offering you more future income per dollar, the one gift this volatile sector reliably gives. And when you meet a double-digit yield from a name trading well below book, you test the coverage before you trust the yield, because that is where the sector eats capital. One broken dividend does not have to break a retirement plan if the plan is built across many holdings; think in portfolio yield, keep the exposure small enough to hold through a rough quarter, and let the spread machines compound quietly. That is how an "income machine" ends up as real, repeatable retirement income — and how a yield trap ends up as nobody's problem but the one who chased it.