A Tighter Market Than It Looks: What CoStar’s Data Says About Office Vacancy, Rents, and Cap Rates

The office headlines have barely changed in five years: vacancy is high, values are down, stay away. Underneath them, the market has moved a long way. Buildings have been converted or torn down, new construction has nearly stopped, and tenants searching for quality space in many markets are struggling to find it.

To sort out what is really happening, I welcomed Phil Mobley, National Director of U.S. Office Analytics at CoStar, back to America’s Commercial Real Estate Show. Phil leads the development of CoStar’s house view of the office market, and his numbers paint a picture that owners and investors should study closely.

The short version is that the office market is recovering, unevenly, and the recovery rewards the right buildings with the right owners.

The Recovery Is Real, and Vacancy Is Still Structurally Higher

By CoStar’s measure, the national office vacancy rate sits at about 13.6%. That is down 50 basis points over the past 15 months, which Phil describes as an unequivocal recovery. It is also 100 basis points above the previous peak, set right after the Great Recession.

Phil’s point is that the sector is operating under a different equilibrium than it once did. Vacancy will likely stay higher than we have measured before, and yet rents are rising. Over the last 12 months, office rents grew 2.5%, the strongest year-over-year rate since the pandemic. A market where landlords raise rents while headline vacancy stays elevated tells you the headline number is hiding something.

The K-Shaped Office Market

What it hides is concentration. Phil’s research divides the office inventory by vacancy level, and the shift over the past decade is striking.

• High-Vacancy Buildings: Roughly 20% of buildings have vacancy of at least 25%. Historically they held about two-thirds of all vacant space. Today they hold about 80%.
• High-Occupancy Buildings: Buildings less than 10% vacant used to hold about 10% of all vacant space. Today it is closer to 4%.
• Quality Is in the Eye of the Beholder: The winners are defined less by Class A, B, or C labels and more by location, amenities, submarket, and where other tenants are moving.
• Market by Market: Midtown Manhattan is seeing trickle-down demand into Class B space on Park and Madison Avenues, while Washington, DC, Chicago, and Los Angeles are still working to backfill lost occupancy.

Phil calls this the office sector’s version of the K-shaped recovery. I would add one more factor that tenant representatives weigh heavily: the ownership and the capital stack. Phil agreed that a landlord’s ability to fund lobby and common area renovations, generous tenant improvement (TI) packages, or pre-built spec suites is getting real traction, particularly with smaller occupiers who need speed to market, such as an AI startup that wants to move in quickly.

Smaller Deals in a Supply-Constrained Market

With very little new supply coming online, large Fortune 100 type companies have limited options to relocate from space they have occupied for 10, 20, or 30 years. There are exceptions, such as American Express essentially pre-leasing a 2 million square foot tower at Two World Trade Center, but most large occupiers are staying put.

That leaves the market open to smaller tenants. The number of lease deals has fully recovered, while average deal sizes run 12% to 15% below the pre-pandemic average, so total leasing volume has not fully caught up. Phil expects that to continue for a simple reason: you cannot lease availabilities that do not exist.

On attendance, Phil believes we are mostly where we are: roughly 70% to 80% of pre-pandemic weekly attendance, a figure that has barely moved in a year. He compares office to a car, a low-utilization asset that still carries high value when you need it, and notes that 10% to 15% of knowledge workers never worked effectively from home.

AI and the Office Workforce

So far, Phil says, AI has been unambiguously positive for office demand, because AI firms and adjacent tech companies have hired people and leased space ahead of expected growth. There is some evidence of slower entry-level hiring, though it is hard to say how much of that is AI itself.

At Bull Realty, we use AI heavily. It has made us more productive, and we have actually hired more people, putting them on different work. Some of my clients have reduced headcount instead. Phil’s view is that a larger disruption may only show up after a broader recession, if companies lay off knowledge workers and choose not to rehire. CoStar’s base case, built on the Oxford Economics forecast, does not include a recession, but the error bars on AI’s long-term effect are wide.

Investment Sales and the Cap Rate Surprise

Office transaction volume was up about 10% year over year through Q3, though it has moved sideways in recent months as the 10-year Treasury swings. Prices corrected about 20% from the peak, compared with 30% during the Great Recession, and that correction took almost four years. Values rebounded roughly 5% off the trough over the following 12 months, then flattened.

• Cap Rate Forecast: Office cap rates peaked around mid-2025, alongside the peak in vacancy, and CoStar forecasts them to keep compressing. That stands apart from CoStar’s industrial and multifamily outlooks, which call for flat or slightly expanding cap rates.
• Risk Premium: The office risk premium over the 10-year has been very high because demand was so hard to read. With a finite universe of investable office and more clarity on what makes a building competitive, that premium has room to shrink.
• User Buyers: Owner-users bought 12% to 15% of office assets before 2020 and about 30% a year or so ago. That share is now 20% to 25% as institutional investors return.

Institutional capital is coming back because, as Phil put it, buyers have stopped waiting for a unicorn rate scenario and now understand the game in office. As I like to say, date the rate, marry the deal.

Final Thoughts: Landlord Capital Is King

CoStar forecasts a long, slow decline in vacancy that flattens out above 13% through the end of the decade. Positive absorption has been about 20 million square feet over the past year, compared with roughly 60 million in a typical year in the 2010s, and job growth is expected to run 0.3% to 0.5% a year. The improvement will come mostly from shrinking supply. Phil’s key point is that removing obsolete space does little for lease economics, because that space was never competitive. Tenants shopping today already face a much tighter market than the headline vacancy rate suggests.

I still believe office may be the buy of the decade. Phil agrees that opportunity exists for investors who know the asset, its market, and its tenant base, and who have a plan. Hope is not a strategy. His closing advice is that the landlord’s capital is key and maybe king, whether it goes into spec suites, TI reserves, whole-building renovations, or a low-basis acquisition repositioned to compete with the best of the market. I am selling a building right now that sat effectively off the market for years because tenant reps would not show it. With a new owner and a stable capital stack that can pay commissions and fund tenant improvements, buildings like that can turn around quickly.

Have a Commercial Real Estate Strategy to Discuss?

With more than $8 billion in closed commercial real estate transactions, Michael Bull, CCIM, brings extensive transaction and market expertise to owners, investors, and companies. As host of America’s Commercial Real Estate Show and an active broker with a strong presence in Atlanta and throughout the Southeast, Michael is available to discuss property strategies, market conditions, acquisitions, dispositions, and other commercial real estate endeavors.

Michael Bull, CCIM
Michael@BullRealty.com
404-876-1640 x 101