Adamas Trust's September Stress Test: When Agency Convexity Ate the Book Value
A 19.6% mark-to-market hit, a $2.9 billion TBA hedge, and a dividend that still covers — the mortgage REIT's Q3 was a lesson in duration risk.
ADAM · Earnings Call · 2026-10-29
Adamas Trust's September Stress Test
Adamas Trust went into the third quarter as a leveraged owner of agency mortgage-backed securities and came out with one of the most violent mark-to-market hits a mortgage REIT can report without a credit event. Adjusted book value fell 19.6% to $8.88 per share; GAAP book value was $7.93. The GAAP net loss was “$171 million, or $1.89 per share” — Kristine R. Nario-Eng, Chief Financial Officer · 2026-10-29. The stock has already voted: ADAM is down 23% over the last 90 trading days and sits nearly 30% below its August peak. Yet the same call raised the dividend and reported earnings available for distribution of $0.30 per share, covering the payout for a seventh straight quarter. That is the tension.
The mechanics: convexity, not credit
Jason Serrano did not mince words. “The third quarter, particularly September, was marked by a sharp increase in interest rate volatility and widening mortgage spreads, creating one of the most challenging environments for agency mortgages in years.” — Jason T. Serrano, Chief Executive Officer · 2026-10-29 The 10-year Treasury rose 82 basis points in the quarter to roughly 5.29%, the first close above 5% since 2007; agency current-coupon spreads widened from 107 to 125 basis points, with nearly three-quarters of that in September. For a portfolio built around higher-coupon agency pools and a duration gap of about 1.4 years, that was close to a worst-case combination. Negative convexity turned the rate move into a book-value accelerant, not just a yield event.
Nick Mah gave the cleanest decomposition: “The agency portfolio accounted for 78% of the decline in adjusted book value, nearly all of it mark-to-market.” — Nicholas Mah, President · 2026-10-29 The rest was structural:
We estimate that convexity was the largest contributor to the decline, accounting for 36% followed by the duration gap at 32%, with the remainder attributable to spread and portfolio changes.
The $2.9 billion TBA short, sized at 23% of the gross portfolio and 37% of the agency long book, contributed a positive $0.29 per share. But it was a hedge against tail risk, not a cure for the marks. Management “did not liquidate any of these core assets” — Jason T. Serrano, Chief Executive Officer · 2026-10-29; the spec-pool balance actually grew from $6.4 billion to $6.9 billion. That choice preserves the recapture option, but it also means the book value recovery is now a market call, not a management call.
From deployment to defense
The contrast with the prior calls is the story. In February, Mah was talking about growing the agency allocation: “we do expect the Agency portfolio to grow. So right now, it's at 56% of equity capital. We do expect it to grow into the 60s” — Nicholas Mah, President · 2026-02-19. In July, Serrano was still framing capital allocation around a “15% plus type of equity returns” — Jason Serrano, Head of Investor Relations or similar senior IR role · 2026-07-30, and the quarter-to-date book value update was a modest “down approximately 2.3%” — Nicholas Mah, President · 2026-07-30. Three months later, the language is rate volatility, agency spreads, and defense. Agency recourse leverage was cut to 5.9x from 8.3x; company economic recourse leverage ended at 4.8x. Available cash rose to $259 million, up 42% from the second quarter, and total excess liquidity capacity reached $447 million. The keyword map has flipped accordingly: available cash is now a top theme, while the prior quarter's language around incremental capital and business purpose loan growth has receded.
This was not simply a macro story. The Fed rate hike wave was global — the market's own keyword heat around high rates and inflation was real, and management tied the inflation impulse to the prolonged Iran conflict. But most mortgage REITs did not lose a fifth of adjusted book value in a single quarter. Adamas did because its agency book was the earnings engine and the risk concentration at the same time.
What survived — and what could still break
The recurring earnings base held. Adjusted net interest income rose to $51.5 million from $50.3 million; agency-adjusted NII grew 7% to $39.3 million, and single-family credit-adjusted NII rose 4% to $21.3 million. In the latest filed quarter, net interest income was $50 million, up 38% year over year, which is why the board could raise the quarterly dividend to $0.30 and still point to a $1.20 annual payout equal to about 13.5% of September 30 adjusted book value. The spread did compress — 140 basis points versus 148 — but the asset base grew enough to keep the dollar income moving higher.
The credit book is not the problem. BPL rental delinquencies fell 10 basis points to 1.4%; the BPL bridge portfolio is running off; multifamily mezzanine payoffs continue at a 55% annualized rate. The problem is that the agency book's marks are now the swing factor for book value, and the fourth-quarter recovery is early. Management says “company-adjusted book value is up approximately 4% as of Monday, 10/26” — Jason T. Serrano, Chief Executive Officer · 2026-10-29, after cutting the TBA short by 76% to $680 million and widening the duration gap back to about 0.6 years. Mah argues that “tail risk of outcomes in spread and convexity have significantly declined” — Nicholas Mah, President · 2026-10-29. Maybe. But the 10-year is still near 5%, the Iran conflict is unresolved, and the TBA short remains a live tool. A single renewed spread-widening episode could turn the October retracement into another mark.
Bottom line
Adamas Trust's Q3 was a company-specific stress test inside a market-wide rate move. The market has already priced a lot of it: a 23% 90-day decline and a nearly 30% drawdown from the August peak. The call's real signal is not the dividend headline — it is the pivot from BPL Rental growth and incremental capital deployment to duration gap management and liquidity. If spreads stabilize, the same high-quality collateral bought at lower prices could produce the mid-to-high-teens returns management is now advertising. If they do not, the seventh consecutive quarter of dividend coverage will be doing a lot of work holding the story together.