ARMOUR (ARR) Q3 2026 Earnings Call Transcript

ARMOUR Residential REIT reported a Q3 GAAP net loss of $327 million, or $2.26 per common share, and a 17% quarter-over-quarter decline in book value to $14.55 per share; total economic return was negative 12.9%. Management attributed the pressure to mortgage-spread widening and rising rates, including a 70–80-basis-point increase in Treasury yields, while distributable earnings were $0.71 per share against $0.72 in quarterly dividends. The company ended the quarter with $1.02 billion of liquidity and an 89% swap-to-repo hedge ratio, and said it expects earnings available for distribution to be around the dividend level in the coming quarter.

Analysis

ARMOUR Residential REIT’s quarter is a reminder that Agency-only collateral does not mean low equity risk: spread widening, negative convexity and roughly 8x leverage can turn a sharp rates/volatility shock into a disproportionate book-value loss. The second-order risk is capital allocation: issuing common near book value and buying back when shares are below book value can be accretive in principle, but buybacks compete directly with the liquidity buffer needed to withstand repo haircuts and further spread widening. Distributable earnings only approximately matching the near-term dividend leaves little stated coverage cushion; one-quarter guidance is not evidence of a durable payout.

The constructive case depends on spread normalization, not simply Fed cuts. Lower mortgage rates could revive refinancing and weaken the slower-prepayment benefit, while a rapid Treasury rally may extend duration and force hedging or asset sales. Management’s expected GSE support is a potential catalyst, but reported activity and remaining purchase capacity should be verified against official holdings data; it is not a firm floor. The 15–18% levered-return estimate is management’s estimate, not an independently established return.

Near term (days to weeks), volatility and year-end repo conditions dominate. Over 1–3 months, official GSE purchase data and mortgage OAS behavior matter more than the rate-cut narrative. Over 6–18 months, funding costs, dividend coverage and any additional equity issuance determine whether discount-to-book value closes or persists. Contrarian point: the wide spread may create attractive prospective carry, but the yield headline can obscure leverage and payout fragility; neither the dividend nor a GSE backstop makes this a defensive income trade.

Key Decisions for Investors

• Avoid treating ARMOUR Residential REIT’s stated dividend as a covered, stable yield trade. Reassess after the next distributable-earnings report; falsify the cautious view only if earnings cover the dividend with a sustained cushion and book value stabilizes.
• Conditional relative-value idea: initiate a small long in the iShares MBS ETF (MBB) against a duration-matched Treasury hedge only after mortgage OAS stops widening and official data corroborate renewed Fannie Mae/Freddie Mac buying. This targets spread normalization while limiting broad rate exposure; exit if spreads resume widening or Treasury volatility accelerates.
• Do not use FNMA or FMCC as proxies for a near-term Agency-MBS backstop trade: their equity outcomes depend on company-specific policy and conservatorship factors, not just MBS spreads. Track official retained-portfolio data before underwriting the support thesis.
• Alert: monitor repo-to-SOFR funding spreads, ARMOUR’s liquidity and leverage, and subsequent book value/dividend coverage. A material funding squeeze, renewed equity issuance below book value, or another sharp book-value decline would invalidate the recovery thesis and favor staying underweight mortgage REIT risk.