Empire State Realty Trust, Inc. (ESRT) stock intrinsic value fair value estimate & owner earnings analysis

Empire State Realty Trust owns about 7.9 million square feet of office space, 0.8 million square feet of retail and 743 apartments, all in New York City, plus the Empire State Building Observatory. The portfolio was 90.3% occupied and 93.6% leased at the end of 2025, with about $552 million of annualized rent. Three buildings produced 55.6% of rental revenue in 2025, and the Empire State Building alone produced 32.3%. Observatory revenue was $128.3 million in 2025, down 5.9% from 2024 on fewer international visitors. How we measure earnings. We start from operating cash flow, remove the swings in receivables, payables and deposits, and subtract stock compensation and the cost of keeping the buildings competitive. That last cost is large here. Additions to buildings and improvements (mostly tenant build-outs and building upgrades) were $198.5 million in 2025 against operating cash flow of $249.1 million; leasing commissions are already inside operating cash flow. Of the 2025 additions, $63.9 million was building capital work outside tenant improvements and leasing commissions, and another $94.2 million of tenant work and commissions is already committed under signed leases. On this basis, earnings per share (counting operating partnership units, about 270 million in all) were roughly $0.29 in 2021, $0.36 in 2022, $0.33 in 2023, $0.25 in 2024 and $0.01 in 2025. Years before 2021 reflect the 2017-2019 building-upgrade program and the 2020 pandemic, so we treat them as a different business. What changed in 2025. Earnings fell about 95% from 2024 because capital spending rose about $26 million while Observatory revenue fell and operating cash flow slipped $11.8 million. The five-year middle value is about $0.29, but two consecutive weaker years have not yet been shown to be temporary, so the base case starts from $0.20 (the 2023-2025 average) rather than $0.29. We assume no growth in the first year: the actual trend is down, and we found no dated, documented reason for it to reverse. In December 2025 the company bought 130 Mercer Street in SoHo for $386.0 million and sold Metro Center in Stamford for $64.0 million; we give no credit for the acquisition because its net contribution after financing cannot yet be measured from the filings. Base case: $3.33 per share. Earnings hold at about $0.20 a share indefinitely. The buildings, including the Empire State Building, are long-lived assets in a market where the company has leased and re-leased space for decades, and all building spending is already deducted, so we assume neither growth nor decline. We credit no rent increases, no Observatory recovery and no acquisition income, because none is contracted or proven at a defensible pace. Bear case: $0.05 per share. The 2025 pattern persists: earnings stay near $0.01 a share as tenant spending stays high and the Observatory remains at 2025 levels. The company has about $2.4 billion of debt at a 4.48% average rate, with $149.6 million due in 2028, $398.9 million in 2029 and $513.1 million in 2030. If that debt is refinanced near today's long-term government yields, interest costs rise by roughly the size of the remaining earnings, so we credit value only through about 2029 and nothing after. We did not model a negative outcome because the company can reduce discretionary spending, but we have no filing evidence of how far. Bull case: $5.83 per share. Capital spending returns to the 2021-2024 pattern and earnings return to about $0.35 a share, close to the $0.29-$0.36 the business earned in 2021-2023, held flat. This needs the 2025 spending step-up to prove temporary; the signed leases (93.6% leased versus 90.3% occupied) support, but do not guarantee, that. What would make this wrong. Higher-than-expected tenant spending or a sustained Observatory decline would push earnings toward the bear case; heavy dependence on three buildings means one large lease loss matters. Weaker points in the analysis: our capital-spending treatment counts all building additions as ongoing costs, which may understate earnings if part of the 2025 spend was one-time redevelopment. The filings do not split that out, so we did not adjust for it. The most recent annual report is for 2025 (filed March 2026); a second-quarter 2026 report was filed in August 2026 but we did not use it to change the earnings level.