Boston Properties: Occupancy Is Climbing. The Market Still Thinks the Office Is Dead.

Boston Properties occupies 88.4% of its portfolio today, up from 87.4% three months ago and 86.7% six months ago. Over 1.8 million square feet of new leases were signed in the second quarter at a weighted-average term of 9.9 years — 29% above the company's own 10-year historical average. The gap between space that's already leased and space that's actually occupied stands at 290 basis points, with roughly 85% of those incoming tenants expected to move in before year-end.

The company's stock has lost ground for years, dragged by the story that remote work killed the city-center office. The operating numbers are telling a different story now.

The question is whether the market is still pricing the old narrative while the business quietly moves toward the new one.

The old story

The conventional wisdom over the past two years has been straightforward: people stopped coming back to the office, CBD occupancy collapsed, rents would follow, and office REITs were stranded assets. BXP, the largest publicly traded office REIT in the United States, concentrated in six gateway markets — Boston, New York, Los Angeles, San Francisco, Seattle, and Washington, D.C. — has been one of the most visible casualties. The stock fell sharply from its pre-pandemic highs and remained under pressure as vacancy rates climbed and refinancing risk grew.

That story hasn't been entirely wrong. Occupancy did fall. Rents did soften in parts of the portfolio. The capital structure — $17.4 billion in total debt against $7.7 billion in equity — has been uncomfortable. But "the office is dead" is a headline, not an operating metric. And the operating metrics have been moving in the other direction.

The occupancy climb

Three consecutive quarters of rising portfolio occupancy is the first piece of evidence that the worst is behind BXP's existing portfolio.

End of 2025: 86.7%. End of Q1 2026: 87.4%. End of Q2 2026: 88.4%. Management achieved 170 of its originally projected 200 basis points of full-year occupancy improvement in the first half of the year alone.

This isn't statistical noise. In an office portfolio measured in tens of millions of square feet, a 100-basis-point quarterly move requires substantial tenant commitment. About 86% of that latest gain came from the existing portfolio itself. The central business district portfolio — which generates roughly 91% of annualized rental obligations — sits at 90.7% occupied and 93.6% leased.

The spread between leased and occupied space is the forward-looking signal. That 290-basis-point gap represents about 1.3 million square feet of committed leases that haven't yet started. If 85% of those commence before year-end 2026, as management expects, it translates to roughly another 80-90 basis points of occupancy gain still flowing through the pipeline.

Management has raised its full-year 2026 occupancy outlook to an average of 88.9%, with year-end expected to be closer to 90% than 89%. Longer-term targets are 91% by end of 2027 and a stabilized 94-95% by 2028. Those are ambitious numbers. The question is whether the leasing momentum supports them.

Leasing volume and rent direction

Volume supports the occupancy trajectory. The 106 leases totaling 1.8 million square feet in Q2 covered 380,000 square feet of vacant space and 600,000 square feet of 2026-2027 expirations. The remaining volume came from tenant expansions and relocations, which are the strongest kind of demand signal because they reflect companies choosing to grow rather than merely stay put.

Rent direction matters as much as volume. Management reported meaningful rent increases across several markets:

• San Francisco: Embedded rent growth of 30-40% at 680 Folsom, 535 Mission, and Salesforce Tower.
• New York: Rents consistently 10-15% above prior leases at high-quality Midtown buildings, with 20% year-over-year increases at lower-stack affordable properties.
• Boston: Renewal rent growth embedded between 20-25% at properties like 888 Boylston Street.

These are not across-the-board improvements. Washington, D.C. leasing spreads remain under pressure. Parts of West Los Angeles and Seattle show weakness. BXP's portfolio is concentrated enough that market-by-market variation matters. But the three deepest markets — New York, Boston, San Francisco — are the ones driving recovery, and they carry the most weight.

The demand driver management keeps citing is AI. Companies like Boston Dynamics signed a 322,000-square-foot lease at Reservoir Place in Waltham, Massachusetts, for a robotics and AI innovation center. Similar AI-driven leasing is happening at 680 Folsom and 50 Hawthorne in San Francisco, where management described a tenant demand pipeline pushing 9 million square feet across its portfolio. The argument is simple: AI companies need premier, high-quality workplaces to attract talent, and those workplaces are exactly what BXP owns.

The cash-flow reality

Here's where the picture gets more complicated. Free cash flow — operating cash flow minus capital expenditures — has declined over the past year. Trailing twelve-month FCF was $275 million as of the Q2 period, down from roughly $330 million at the end of 2025 and $591 million two years ago. Capital expenditures have been rising, from roughly $764 million two years ago to nearly $975 million today.

The increase in capex is intentional. Management raised its full-year 2026 leasing capex guidance to roughly $500 million, up from a prior $400 million estimate, because leasing premium space requires investment — tenant improvements, lobby upgrades, amenity spending. The company is also developing seven projects totaling 3.5 million square feet, including the 343 Madison Avenue tower in New York, which just closed a $1.2 billion construction loan and is 50% pre-leased.

For a REIT, free cash flow as calculated by general market data providers doesn't capture the full story. The industry-standard metric is Funds From Operations, which adds back depreciation — a large non-cash charge for real estate companies. BXP's Q2 FFO was $1.78 per share, beating consensus by $0.08. Full-year 2026 FFO guidance was raised to $6.99-$7.05 per share, a midpoint of roughly $7.02. On a $390 million annualized basis (based on approximately 55.5 million shares), the dividend of $2.83 per share consumes about 40% of FFO. That payout ratio is sustainable, and management has now paid dividends for 30 consecutive years.

The same-property net operating income growth guidance was raised to 1.8-2.6% for the full year, up from prior estimates. Same-property NOI measures the operating performance of the stabilized portfolio, excluding development and recent acquisitions — it's the cleanest read on whether the existing assets are actually improving. Management noted that cash same-property NOI may stay flat in 2026 due to free rent periods delaying cash collection, with meaningful improvement expected in 2027.

This is the bridge between the occupancy gains and the financial payoff: occupancy improvement and rent growth today flow into NOI improvement next year, which then shows up in cash flow with a lag due to free rent. The timeline matters. The numbers are moving in the right direction now, but the full cash impact arrives later.

The balance-sheet constraint

The $17.4 billion in debt remains the structural overhang. Debt-to-equity sits at roughly 202%, which is elevated even for office REITs. The company faces a $1 billion unsecured bond maturity in October 2026, carrying a GAAP interest rate of 3.5%. Management may use up to $300 million of its $494 million cash balance to reduce the refinancing size.

Asset sales are providing partial relief. BXP has raised $370 million in net proceeds year-to-date, with total proceeds exceeding $1.2 billion since its last investor day. The company expects $1.7 billion in cumulative net proceeds by year-end 2026. But management estimated that 2026 asset sales will dilute per-share value by roughly $0.11, and some of those sales may be at below-replacement-value prices in a still-cautious capital market.

The debt markets themselves remain functional — unsecured credit spreads are trading in the low 100s of basis points — which suggests refinancing at manageable rates is possible. But the scale of the balance sheet means there's always one maturity event that could create a problem. That's the risk investors need to carry.

What the market is pricing

At roughly $70 per share, BXP trades at about 14x trailing EV/EBITDA and approximately 37x forward earnings. The dividend yield sits near 4% on a trailing basis, with a forward yield closer to 5.6% after the company reduced its quarterly dividend in the past year. The stock is down roughly 1.7% over the trailing twelve months and has traded between a 52-week low of $49.72 and a high of $79.33.

These multiples tell you the market hasn't fully bought the recovery story. A 37x forward earnings multiple on a REIT with 202% debt-to-equity suggests investors are still pricing meaningful risk — refinancing, leasing execution, and the possibility that occupancy gains stall before reaching stabilization. A 14x EV/EBITDA multiple is not cheap for an office REIT that owns a premier portfolio in gateway markets. It's not expensive either, relative to what it would be if occupancy, rent growth, and NOI were all trending up while debt was being paid down.

The aggregate analyst signal from AInvest labels BXP a Buy, with a strong liquidity rating of 7.71 out of 10. That doesn't determine the case, but it does suggest the sell-side hasn't turned away from the name.

The case and the risk

The case for BXP rests on a specific sequence: occupancy continues climbing from the current 88.4% toward the 91% target by 2027, rent growth at renewals remains in the 10-25% range at premium assets, same-property NOI accelerates in 2027 as free rent periods end, and the $1 billion bond refinancing in October proceeds without disruption. That sequence produces a business where FFO grows, the dividend is comfortably covered, and the stock re-rates from a beaten-down office name to a premium-workplace compounder.

The bear case is equally concrete: AI-driven leasing demand doesn't materialize at the pace management describes, occupancy gains stall or reverse, the heavy capex burden continues to compress free cash flow, and refinancing costs rise if credit conditions deteriorate. The $17.4 billion debt load is the leverage that turns moderate operating disappointment into a genuine problem. If occupancy fails to reach 90% by year-end 2026 — management's own near-term target — the recovery narrative loses its foundation.

The condition to watch is the occupancy trajectory and same-property NOI growth in the coming quarters. If occupancy continues its current pace and NOI accelerates in 2027 as management projects, the stock has room to re-rate. If occupancy stalls below 88% and capex keeps eating cash flow, the old story hasn't been disproven — it's just been delayed.

The market is still pricing the old risk profile while the operating setup is already getting cleaner. That's the gap. Whether it closes in BXP's favor depends on whether the next two quarters of leasing data confirm the trend or break it.