ARMOUR's Dividend Is Now Bigger Than Its Business

A 25bp Fed hike and the widest mortgage spreads of the year cost ARR 17% of book value — and management is refusing to blink on the payout.

ARR · Earnings Call · 2026-10-22

The Script Flipped, And The Tape Already Knew

For the better part of a year, ARMOUR's call was a bet on the easing cycle. In July, Desmond Macauley described a book hedged to roughly zero duration with “a slight positive bias there. And that's because we think that the Fed could stay on hold for longer” — Desmond Macauley, Co-Chief Investment Officer · 2026-07-23. Sergey Losyev had flagged the risk a year earlier in almost exactly the right words, calling out that “a pause in the easing cycle or something that would cause them to walk back their projections would be a potential source of volatility in the market” — Sergey Losyev, Co-Chief Investment Officer · 2025-10-23. That is precisely what happened. September delivered a 25 basis point hike — the first since 2023 — and Treasuries repriced 70 to 80 basis points over the quarter. Scott Ulm did not dress it up: “Persistent inflation prompted the Federal Reserve to raise rates by 25 basis points in September, its first hike since 2023 and signal further tightening” — Scott Ulm, Chief Executive Officer · 2026-10-22. The damage was concentrated and late. Gordon Harper's framing is the cleanest evidence: “Our results for the quarter reflect the significant widening of mortgage spreads in the last 3 weeks of the quarter as interest rates rose and the markets reacted to a more hawkish Federal Reserve” — Gordon Harper, Chief Financial Officer · 2026-10-22. Losyev quantified it — option-adjusted spread widened 20 basis points, with roughly 80% of that in September alone. Book value fell to $14.55 from $16.99-ish at June 30, a 17% decline, and the stock's own tape tells the same story: down 25.5% over 90 days and 27.3% from its June 30 peak of $17.45. This is not an ARMOUR-specific event. The global keyword set for the quarter puts Fed rate hike near the very top of the market-wide agenda, and the reporter tape is littered with regional banks working the same theme — SSB and ORRF on the Fed hike cycle, UCBI on the September move, SMBK on a December one. ARMOUR sits inside a broad rate shock, but its expression of it is the purest: no credit book, no deposit beta, just the Agency MBS market and mortgage spreads taking the hit directly.

A Dividend That Now Out-Earns The Business

Distributable earnings were $0.71 per share. The monthly dividend totals $0.72 for the quarter. That gap — one cent — is the entire story of the equity right now.

Recent volatility in interest rates and Agency MBS spreads has pressured both our share price and book value, resulting in an unusually high dividend at current market prices. We remain constructive on the Agency MBS market and believe many of the factors driving recent volatility are likely to resolve over the coming months and into next year.
Scott Ulm, Chief Executive Officer · 2026-10-22

Note what this phrasing does: it reframes an uncomfortable yield as a market anomaly rather than a payout problem. Losyev leaned on the same logic, saying earnings “are going to be in the range of the dividend for the coming quarter” — Sergey Losyev, Co-Chief Investment Officer · 2026-10-22, and the company had already declared November at $0.24. Meanwhile the GAAP line was a $327 million net loss, or $2.26 per share, and total economic return for the quarter was negative 12.9%. The accounting loss is marks, not cash — but it is real book value. The capital actions were equally ambiguous. ARMOUR raised $205.7 million of common and $2.5 million of preferred in the third quarter through its ATM, then turned around and repurchased 363,700 common shares at the end of September. “We look at issuing under the ATM when we have an attractive price... we should always be on both sides of the market” — Sergey Losyev, Co-Chief Investment Officer · 2026-10-22, Losyev explained. The buyback is a rounding error against the issuance, but the instinct is telling: they were selling near book early in the quarter and buying the dip late.

The Threads That Went Quiet

The more interesting signal is in what dropped out of the conversation. In the second quarter, the top of ARMOUR's keyword stack was the yield curve, followed by swap spreads, a new CMBS position, and that distinctly directional phrase, a bull steepener. None of those appear in this quarter's vocabulary. The new entrants are tactical and defensive: deep discount seasoned MBS, dry powder, and favorable repo conditions. The CMBS thread deserves a flag. Six months ago management was extolling the new commercial MBS position as a convexity substitute for expensive specified pools. This quarter the portfolio is described as invested exclusively in Agency MBS and DUS securities — "consistent with its strategy since the third quarter of 2020." That is a retroactive reframing of a trade that has simply stopped being mentioned. What replaced it is a very specific recycling trade: they sold roughly $900 million of 5- and 10-year DUS pools and deep-discount seasoned paper that outperformed in September, and intend to redeploy into moderate-discount coupons where levered returns now run 15% to 18%. In July, Losyev was telling investors specified pools looked “fully valued here versus TBAs” — Sergey Losyev, Co-Chief Investment Officer · 2026-07-23. A week of September chaos made the same collateral cheap again.

The Bull Case Rests On Two Assumptions

The first is that the hike cycle is nearly done. Ulm's base case is a “shallow Fed hiking cycle rather than a repeat of 2022, '23” — Scott Ulm, Chief Executive Officer · 2026-10-22. Macauley put a number on the divergence: “Markets are pricing about 3 to 4 cuts by the end of 2027. And our view here is the base case is 2 cuts by the end of 2027” — Desmond Macauley, Executive (likely Co-Chief Investment Officer or similar senior investment role) · 2026-10-22 — which, he noted, is hawkish against the dot plot but dovish against the curve. Hence a small long-duration bias in the front end. The second is that the GSEs re-enter as buyers. Ulm sees more than $100 billion of capacity left under their caps and treats them as a backstop at wide spreads. Macauley is watching the specified-pool auction sizes shrink: “we do expect some buying to show up” — Desmond Macauley, Executive (likely Co-Chief Investment Officer or similar senior investment role) · 2026-10-22. The balance sheet is set up to wait. Stockholders' equity of $2.4 billion still sits just 9% below its Q2 peak, and $19.4 billion of repo borrowings against that equity implies about 8.3x, with 7.9x as of October 19 and a swap hedge-to-repo ratio of 89%, the highest since 2022. Liquidity was $1.02 billion. The offsetting strength is real: net interest income of $84 million is up 117% year over year and sits at the top of a nine-year range. The business works. The problem is that the marks move faster than the carry — and in a quarter where the Fed rate path turned against them, the market repriced the equity before it repriced the earnings.