The Occupancy Rate That Isn't What It Says
Office occupancy data carries more definitional ambiguity than almost any other commercial real estate metric, and that ambiguity has real consequences for acquisition underwriting. When a broker OM reports "88% leased," it does not mean 88% of the building's space is being actively used. It means 88% of the space has a lease in effect — which is a very different number than physical utilization, which is different again from economic occupancy when concessions are factored in.
In the post-2024 office market, the spread between leased percentage and physical utilization has been particularly wide. Data from keycard access, desk utilization sensors, and parking garage throughput consistently shows that leased office space in most metro markets is being used at 50-65% of its contractual capacity on a typical Tuesday through Thursday, with significantly lower utilization on Mondays and Fridays. This doesn't necessarily affect near-term income — the tenant is paying rent regardless of utilization — but it directly affects renewal probability, which is the underwriting question that matters most for long-hold acquisitions.
Sublease Overhang: The Shadow Vacancy Problem
The metric that most materially changed office market dynamics coming out of 2020-2022 was sublease availability. Corporate tenants who had committed to large office footprints during pre-pandemic expansion — typically on 7-10 year terms — found themselves with more space than their hybrid workforce policies required. Rather than waiting out the lease, many put significant portions of that space on the sublease market at materially discounted rents.
Sublease space creates a specific underwriting problem because it competes with direct space at a discount. A tenant shopping for 20,000 SF in a major CBD market has options: take direct space at market asking rent with a full TI package, or take sublease space at 15-30% below market with shorter remaining lease terms and a TI package funded by the sublessor. For price-sensitive tenants or those who want shorter commitments, the sublease option is compelling. This suppresses effective demand for direct space and puts downward pressure on market rents — a pressure that doesn't show up cleanly in the headline vacancy statistics because sublease space is often tracked separately.
For acquisition analysis, the relevant metric is total available supply — direct vacancy plus sublease space — not just direct vacancy. A building with 10% direct vacancy sitting in a submarket with another 8% sublease availability in the immediate competitive set is effectively in a 18% supply environment, and the market rent and TI assumptions should reflect that.
Flight to Quality: The Class A Divergence
The standard observation about post-2024 office is that Class A is doing better than Class B and C, and that's true at a high level — but the category is too broad to be analytically useful. "Class A" encompasses trophy buildings with premier amenities and LEED certification in core locations, and it also encompasses 1990s vintage towers with adequate but not exceptional finishes that were repositioned and re-leased during the 2015-2020 cycle.
The flight-to-quality dynamic that benefits office acquisitions is specifically concentrated in what the market has come to call "trophy" or "Tier 1 Class A" — buildings with genuine amenity differentiation: efficient floor plates, high ceiling heights, high-end building lobbies, on-site food and beverage, fitness facilities, outdoor terraces, and increasingly, buildings with strong green certification (LEED Gold/Platinum, ENERGY STAR) that satisfies corporate ESG commitments. These buildings can support asking rents at a premium to the market and are seeing positive leasing activity even in markets where overall Class B is struggling.
Tenant concession trends in the current market are significant at both ends of the quality spectrum. Trophy Class A buildings offering to the best tenants are pricing TI packages in the $60-80 PSF range in gateway CBDs, with 6-9 months of free rent on longer-term leases. Class B buildings competing for tenants are running TI at $30-50 PSF but often also offering more free rent — 9-12 months — to compensate for the quality gap. These concession levels have a meaningful impact on the effective economic occupancy during the free rent period and on the underwritten value of lease-up assumptions.
What WALT Means for Office Risk
Weighted average lease term remaining is particularly important in office underwriting because the tenant decision cycle is long and the cost of re-leasing is high. An office acquisition with a WALT of 3 years is a fundamentally different risk profile than one with a WALT of 7 years, even if the in-place rent and occupancy look identical at purchase.
WALT analysis should go beyond the portfolio average. The distribution matters: a WALT of 5 years that is actually composed of a 40% anchor tenant with 9 years remaining and 60% of tenants with 2-3 years remaining is a near-term rollover exposure story, not a stable 5-year income story. The rollover schedule needs to be modeled tenant-by-tenant, with realistic assumptions for renewal probability, downtime between tenants, and TI and leasing commission costs associated with re-leasing.
We're not saying a short WALT is necessarily a problem — it can be a value-add story if current in-place rents are below market and the rollover gives you the opportunity to re-lease at higher rates. What we're saying is that WALT without a rollover schedule and cost model attached to it is an incomplete analysis.
Tenant Concession Modeling in Office Underwriting
The tenant improvement allowance and free rent period are two components of the office leasing cost structure that are frequently understated in acquisition models. Brokers presenting an OM will often show "effective rent" calculations that amortize TI and free rent over the full lease term — which makes the economics look better than the cash flow actually is.
For acquisition modeling, TI and leasing commissions should be treated as capital expenditures, not as income adjustments. A lease producing $45 PSF annual gross rent with $60 PSF of TI and 8 months free rent has a net cash yield in Year 1 that is substantially below the headline rent figure. On a 10-year lease, the amortized cost is manageable; on a 5-year lease, the amortized TI cost represents a meaningful drag on unlevered returns.
The forward underwriting question for an office acquisition is: what does the next round of TI and leasing commissions cost, and when does it hit? Lenders will ask this. The IC should know the answer before any of them do.
Underwriting Office in a Structurally Uncertain Market
The honest position for acquisition underwriting is that long-term space utilization patterns for office are not yet fully settled. Corporate real estate policies on hybrid work continue to evolve. Some large employers have moved to return-to-office mandates; others have maintained hybrid frameworks. The aggregate demand consequence of these policies at the portfolio level is still working through the system, particularly as legacy leases signed before 2020 roll to renewal decisions.
A viable office acquisition underwriting framework for this environment: stress-test renewal probability on every tenant with more than 15% of building revenue and a lease expiring within 5 years. Use a renewal rate assumption that reflects the submarket's historical renewal rates for comparable building quality, not the portfolio's own history. Model TI and leasing costs on a conservative per-SF basis — current market TI for the quality tier, not the prior cycle's experience. And be explicit about the exit cap rate assumption — office cap rate expansion risk in most markets is still not fully priced, and a model that exits at going-in cap or at minimal expansion is not showing the risk honestly.
Office acquisitions can still make sense — particularly trophy assets with long WALT tenants, strong locations, and the amenity profile to compete for the flight-to-quality demand. The analytical rigor required to underwrite them well is simply higher than it was before 2020, and the margin for modeling error is smaller.
