Rithm Property Trust Inc. (RPT) stock intrinsic value fair value estimate & owner earnings analysis

Rithm Property Trust (formerly Great Ajax) is a small, externally managed mortgage REIT. Since June 2024 it has been run by an affiliate of Rithm Capital, which is moving it away from buying seasoned residential mortgages toward commercial real estate credit and short-term residential transition (fix-and-flip) loans. At June 30, 2026 it held $929M of assets: $348M of seasoned residential loans (about 29% California, 12% Florida, 9% New York), $118M of residential transition loans bought in the first half of 2026, $84M of commercial mortgage bonds (down from $274M at year-end after sales), $190M of residential mortgage bonds and beneficial interests, $79M of commercial real estate equity stakes (including a 3.9% interest in the Paramount office portfolio in New York and San Francisco) and $67M of cash. It owed $644M: $311M of short-term repurchase lines maturing between July 2026 and April 2028, $212M of securitization bonds that can only be repaid from the loans inside those trusts, and $110M of unsecured notes due September 1, 2027 whose coupon stepped up from 8.875% to 9.875% after a 2024 downgrade. Ahead of the common also sit $52.1M of 9.875% Series C preferred shares. Common book equity is about $233M, or $30.00 on 7.77M shares. What the business earns. This is a lender, so borrowing is its raw material and interest is an operating cost; we value what is left for common holders after interest, operating costs and preferred dividends, and do not subtract the debt a second time. The assets earn roughly 5.5% to 6% on their carrying value (2026 first-half interest income of $24.9M), the borrowings cost about 5% on average, and operating costs run about $18M a year (manager fee, servicing, professional fees), so after $5.2M of preferred dividends the common shares earned a loss: -$2.7M in 2025 (about -1% of book) and -$2.6M in the first half of 2026, with the second quarter slightly positive at $0.09 per share only because of investment gains. Our cash-flow measure has been negative in almost every year for ten years; it understates a lender that books loan discount accretion as investing cash, so we used reported earnings to common as the earnings stream and checked that both point the same way. The common dividend ($0.36 a quarter) is being paid out of capital. How we valued it. We discount everything common holders can expect to receive at 6%. Because the loans and bonds are carried at amortized cost yielding close to 6%, their carrying value is a fair starting point at our rate. The company's own fair-value estimates are lower ($304M for the seasoned loans against $348M carried), but those reflect the higher yields market buyers demand; the loans sit in trusts whose bonds have no claim on the rest of the company, and are being held to collection, so we take only the weakest assets to fair value, and only in the bear case; the question is how much the vehicle's costs consume. Every year the common equity earns below 6%, it loses value against book, and we charge that shortfall year by year for ten years, then credit the remaining book at year ten, when the seasoned mortgages will have returned most of their principal. We credit no growth: the transition-loan purchases and $110.9M of signed funding commitments are real, but they have not yet produced a profitable quarter before gains, so they appear only in the bull case. Bear case ($10.64 per share). Losses continue at about 3% of book a year (between the 2025 result and the roughly 4.6% annual pace of the first half of 2026 before investment gains), the $110M notes are refinanced in 2027 at about two points more than today, the beneficial interests are realized at the $76.6M the company itself estimates they are worth rather than the $102.2M carried (a $25.6M loss), and the Paramount office stake loses 15% ($9.3M). Total equity would drop below the $240M net-asset-value minimum in the lending covenants within about a year, forcing the dividend to stop and the balance sheet to shrink. The assets would be worth about $779M against $696M owed to lenders and preferred holders, leaving about $83M for common holders. Interest income still covers interest cost about 1.4 times, so the case carries its own debt. Base case ($16.32 per share). Earnings stay near the 2025 level, about -1% of book, through the ten years; the notes are refinanced near today's 9.875% (they trade close to par, with no sign of distress); the dividend continues for two years and then stops as the covenant cushion shrinks. The assets would be worth about $823M against $696M owed, leaving about $127M. The year-one assumption (-1% return on book) matches the latest full year, so no improvement is assumed. Bull case ($23.77 per share). The repositioning works: the transition-loan book and the signed funding commitments lift the return on common equity to about 1% in 2027 and 4% from 2028, still below our 6% rate. The assets would be worth about $881M against $696M owed, leaving about $185M. Change since our last estimate. We previously valued the shares at $19, $26 and $33 by haircutting the assets from book. We now value the company as the ongoing vehicle it is: its lenders are covered by assets that earn about what we require, but a sub-scale, externally managed structure with 9.875% notes and 9.875% preferred earns well below 6% for common holders, and that shortfall, accumulated over years, is worth about $106M in the base case. Book value per share also slipped from $31.62 to $30.00 during 2026. What could make this estimate wrong. If the transition-loan and commercial real estate book earns well above its financing cost, the return on equity could approach or exceed 6% and the shares would be worth close to book. If the office stake, the residential beneficial interests or the transition loans suffer credit losses, or refinancing of the 2027 notes proves harder, value falls toward or below the bear case. Rithm holds warrants on 0.6M shares at $32.16, above all three cases, so we assume no dilution.