COPT DEFENSE PROPERTIES (CDP) stock intrinsic value fair value estimate & owner earnings analysis

COPT Defense Properties owns about 200 buildings that sit beside, or inside, United States defence and intelligence installations - Fort Meade and the Baltimore-Washington corridor in Maryland, Redstone Arsenal in Huntsville, Lackland Air Force Base in San Antonio and Northern Virginia - plus 33 data-centre shells in Northern Virginia leased to a single Fortune 100 cloud company. At the end of 2025, 90% of rent came from this defence and IT portfolio and 10% from six ordinary Washington-area offices it intends to dispose of. The tenants need space next to the mission and frequently pay for secured-space build-outs themselves, which is why the defence portfolio has ended every one of the last nine years at least 93% occupied and has retained at least 75% of expiring space for ten straight years. The risks are concentration: the US Government is 35% of rent on mostly one-year renewable leases that can be dropped if Congress fails to fund them, the top three tenants are 51%, and the cloud customer is 11%. Earnings measure. We take cash from operations, strip out working-capital swings, deduct stock pay, and deduct the spending needed to keep buildings leased - tenant improvements, capital improvements, leasing commissions and lease incentives - while excluding spending on new buildings, which adds capacity. On that basis owner earnings were roughly $1.25-$1.59 a share from 2016 to 2023 and stepped up to $1.92 in 2024 and $1.89 in 2025 as newly built, fully leased buildings came on line; the 2025 dip reflects US Government rent paid late (receivables rose $16 million), and cash from operations in the first half of 2026 was $187 million against $160 million a year earlier, confirming the higher level. These figures are computed on a stricter basis than our previous write-up, so the previous numbers are not comparable. Because the company carries $2.6 billion of debt, we value the business before interest - about $290 million a year in both 2024 and 2025 - and then subtract what it owes. We anchor on that 2024-2025 level because the portfolio was materially enlarged by deliveries from 2022 onward and earlier years describe a smaller business. Growth credited. Last year's actual change was a decline of about 1%; we assume about 2% in year one because the late government payments that caused the dip have already reversed in the 2026 half-year cash flow, and three fully pre-leased buildings under construction are scheduled to open in 2027. Beyond that we credit only 1.5% a year: the company has developed and leased a handful of buildings a year for decades, funded from its own cash flow and modest borrowing, but the per-share result over 2016-2025 grew only about 2% a year, and we do not credit new data-centre awards, the two speculative buildings under development, or any rise in defence budgets. Base case: before-interest cash flow rises from about $296 million to $314 million by 2028 as the 2027 deliveries lease up, then grows 1.5% a year, held flat after year ten because the moat - secured space next to permanent installations - keeps operating but no further growth can be evidenced. The business is worth about $5.6 billion. It owes about $2.55 billion: $2.59 billion of borrowings at 30 June 2026, about $130 million still to spend on committed developments, and $44 million of outside partners' interests, less $24 million of cash and about $190 million of credit for $1.4 billion of 2.0%-2.9% notes running to 2029-2033 that cost far less than market rates. That leaves about $3.06 billion, or $26.64 a share. Bear case: a repeat of the 2011-2013 federal budget sequestration and shutdown period, when defence-office landlords in this region saw leasing stall and occupancy and rents fall. We take before-interest cash flow down about 15% over three years, to roughly $248 million, as government one-year leases are not renewed, the Other offices lose tenants, and the cloud customer stops awarding new shells; it recovers only partly, to about $265 million, and stays flat. Even at the trough the business covers interest on its whole debt refinanced at 6.5% (about $170 million) by roughly 1.5 times, so no forced equity issuance is modelled. Worth about $4.4 billion, owing about $2.55 billion, leaving $16.30 a share. Bull case: deliveries lease on schedule, retention and cash rent increases on renewals (2.7% in 2025) continue, and the company keeps adding fully pre-leased buildings at its demonstrated pace, giving about 2% annual growth after 2028 and about 1% a year for a further ten years. Worth about $6.3 billion, owing about $2.55 billion, leaving $32.49 a share. Margin-of-safety price. We are fairly confident in the earnings level: two consecutive years near $290 million before interest, corroborated by the first half of 2026, from a highly occupied portfolio whose tenants are hard to move. The worst plausible event is a prolonged federal funding fight that lets the government walk from one-year leases at the same time as the 2028-2029 maturities ($345 million of exchangeable notes in 2028 and $550 million in 2029) need refinancing; with debt at roughly nine times before-interest cash flow, that is the scenario that would hurt, though the 2011-2013 episode shows that the buildings stayed largely leased. If we are wrong about growth (none at all) and about the cheap debt (refinanced at market rates), the shares are worth around $20. During the holding period, the dividend - about $1.22 a share, based on the $35 million quarterly distribution payable at year end - is covered by owner earnings of about $1.89. Measured against $17 the dividend alone represents about 7% a year, the bear-case value is roughly that price, and a move toward the base value over three to four years supplies the remainder of a 20% annual return in the base case. That is why the margin-of-safety price is $17 and not higher.