In late 2024 a major Manhattan trophy office sale produced a trade-press story I’ve reread several times since. The asset traded at a meaningful premium to comparable inventory, and an analyst quoted in the coverage explained the math by reference to the asset’s embodied-carbon footprint. The building, completed in the mid-2010s, had been delivered with a higher-than-typical structural and cladding-carbon load. Under the buyer’s screen, that load now read as a financeable risk premium. The seller absorbed it; the trade still closed; the trade press, for the first time I’d seen, treated embodied carbon as a price-discovery variable rather than an ESG-report footnote.
The math has flipped. The default instinct, in commercial real estate, to tear down the underperforming asset and rebuild it newer-and-greener is no longer the trade. The trade is to instrument what’s already standing.
This piece is about why that math has flipped, what the math actually shows on a normal Class B asset, and what an owner sitting on a 1980s building should be doing about it this quarter.
The construction-cost picture in 2026.
Start with delivery cost. Per-square-foot all-in delivery for institutional commercial product in major US markets is now running 1.6 to 2.2 times pre-pandemic levels, depending on type and market. The Federal Reserve’s FRED Producer Price Index series for commercial-building construction tells the macro story. Major-market contractor surveys from Turner and Cumming Group fill in the detail. The labor side, the materials side, and the financing side have all moved in the same direction simultaneously. There is no relief curve in sight before 2028.
The implication for an owner with an existing asset is structural. A new-build replacement at $700 to $1,100 per square foot in a US gateway market does not pencil against a same-bone retrofit at $80 to $220 per square foot in the same market. The retrofit-versus-replace math, on dollars alone, was already tilted toward retrofit in 2019. It is decisively tilted in 2026.
That’s before we account for carbon.
The embodied-carbon math, in dollars.
Embodied carbon is the carbon emitted during the production, transportation, and installation of construction materials. It is fixed at delivery; it is not recoverable through operational efficiency. For most commercial buildings, embodied carbon sits between 350 and 750 kilograms of CO2-equivalent per square meter, depending on structure and cladding type. Concrete and steel dominate. Glass and aluminum contribute meaningfully.
The most rigorous framework for translating this into investment math is Simon Sturgis and Gareth Roberts’ Targeting Zero: Embodied and Whole Life Carbon Explained (RIBA Publishing / Routledge, multiple editions), originally published by the RIBA and updated through several editions. The Carbon Leadership Forum Embodied Carbon Benchmark Study provides the empirical reference numbers. Architecture 2030‘s Carbon Smart Materials Palette provides the alternative-material reference.
Translate the kilograms into dollars at three reasonable carbon-price points: $50 per ton CO2e, $100 per ton, $200 per ton. For a 300,000-square-foot replacement build at typical structural and cladding carbon intensity, embodied-carbon liability runs $1.5 million at $50/ton, $3 million at $100/ton, and $6 million at $200/ton, just for the new-build piece. This is on top of the cost of construction itself.
At today’s voluntary carbon prices, that liability is contingent. Under reasonable forecasts of mandatory carbon pricing through the rest of the decade — which Sturgis’ work, the IFC EDGE methodology, and several major European jurisdictions are converging on — that liability becomes balance-sheet. The asset that hasn’t yet emitted the carbon has an option to avoid emitting it. The asset that has emitted it carries it.
The regulatory horizon.
The regulatory landscape on building-level carbon disclosure and limit-setting has moved meaningfully in the last 24 months. Four developments matter most.
New York City’s Local Law 97, in its current form and proposed expansion, will price building-level operational carbon for the city’s commercial inventory. Penalty structures escalate through 2030. The current debate about expanding LL97 to embodied carbon for major renovations is real and will likely conclude this regulatory cycle.
California’s SB 253, the Climate Corporate Data Accountability Act (signed October 2023) requires Scope 1, 2, and eventually Scope 3 emissions reporting from large public and private companies operating in California. Real-estate sponsors with California exposure are caught. Scope 3 emissions reporting includes embodied carbon from major renovation and new construction.
The European Union’s Corporate Sustainability Reporting Directive (CSRD) and its ESRS technical standards apply extraterritorially to US-domiciled sponsors with EU operations or EU-listed debt. CSRD reporting includes Scope 3 for buildings.
The federal SEC climate-disclosure rule, currently in litigation, is expected to require Scope 1 and 2 disclosure for large registrants. Scope 3 is excluded in the final rule but may return under a future rulemaking.
The net effect: by 2028, the cost of producing a building’s embodied-carbon-and-operational-emissions disclosure package is a material line item for most institutional sponsors. The cost of failing to produce it is rising bid-side discount, narrower lender pool, and screened-out access to a growing share of institutional capital.
The retrofit depth question.
Retrofit is not a single decision. The Rocky Mountain Institute (RMI, 2022, Reinventing Existing Buildings) work on retrofit depth is the cleanest published framework. Three tiers.
Light retrofit. Lighting upgrade, BMS tune-up, predictive-maintenance program, instrumentation overlay. Capex on the order of $5 to $25 per square foot. Operational-carbon reduction of 10 to 25 percent. Quick payback. Minimal tenant disruption. The math nearly always works.
Medium retrofit. Envelope upgrades, electrification of building systems, recommissioning of major MEP, structured-cabling and connectivity uplift. Capex on the order of $40 to $90 per square foot. Operational-carbon reduction of 25 to 55 percent. Multi-year payback at current energy prices, faster under carbon-pricing assumptions.
Deep retrofit. Major envelope replacement, structural upgrades for adaptive reuse, full MEP replacement, façade re-design. Capex on the order of $120 to $280 per square foot. Operational-carbon reduction of 55 to 80 percent. Payback period of seven to fifteen years, sensitive to a long list of assumptions.
Each tier has a different IRR profile. The mistake most owners make is treating retrofit as a single decision. The right disposition is sequential: instrument first, then retrofit toward the depth the data justifies, then iterate.
Where instrumented retrofit beats deep retrofit.
This is the most important claim in the essay. Light retrofit plus instrumentation, on most Class B assets in most US markets, produces better risk-adjusted returns than deep retrofit alone.
The reason: deep retrofit pays for performance the building’s current tenants may not value at current lease rates. The carbon savings accrue to the owner’s ESG profile, but the rent uplift is often modest at the next lease renewal. The hold period required to recover the deep-retrofit capex is long. The capital is illiquid until the next sale.
Light retrofit plus instrumentation produces measurable performance and a documented digital revenue line in months, not years. The asset becomes underwritable at a tighter cap rate within the current hold. The capex is recovered through cap-rate compression at exit rather than through rent uplift alone. The data accrued during the hold informs whether deep retrofit, when the right lease-rollover window opens, is justified — at which point the numbers are sharper than they would have been at the start.
The instrumented-first path is also lower regret. If the asset turns over to a new owner mid-hold, the new owner inherits the instrumentation and the data, and can make their own retrofit-depth decision against richer information. The deep-retrofit-first path encumbers the asset with a long-tail capex line that may or may not match the next owner’s thesis.
The financing instruments.
Retrofit financing is more developed than most owners realize. Four product lines matter.
C-PACE (Commercial Property Assessed Clean Energy) is now available in roughly 38 US states and the District of Columbia. C-PACE is structured as a long-tenor, fixed-rate, assessment-secured loan that runs with the property. It typically funds 100 percent of qualifying energy and resilience capex, including instrumentation and predictive-controls programs. The capital is non-recourse to the owner and survives transfer.
Sustainability-linked loans are now standard at the major commercial banks for institutional sponsors with documented sustainability KPIs. The pricing structure typically embeds 10 to 25 basis points of pricing tightness against documented performance over the term.
Green bonds and transition-finance instruments at the corporate level let sponsors capitalize on portfolio-wide retrofit programs. The pricing differential against vanilla debt is real but variable; the access to the capital pool is the more meaningful advantage.
Federal and state-level tax credits, particularly the residual structure of the Inflation Reduction Act’s commercial-buildings credits, continue to subsidize specific equipment classes through the rest of the decade. These are bolt-ons to the financing stack, not the primary instrument.
A real comp set.
Apply the framework to a single concrete asset and the numbers become persuasive.
A 300,000-square-foot 1980s Class B office in a US gateway market, currently 65 percent occupied, generating $4.8 million NOI, at an in-place cap rate of 8.0 percent. The owner has three options.
Option A: Replace. Demolish and rebuild at 320,000 square feet of efficient new product. All-in cost at $850 per square foot: $272 million. Embodied-carbon footprint: roughly 2,100 tons CO2e per 100,000 square feet, so roughly 6,700 tons total; at $100/ton, $670,000 of embodied-carbon liability. Five-year delivery timeline before stabilized cash flow. IRR, fully loaded against current rents and reasonable lease-up assumptions: marginal-to-negative in most US markets.
Option B: Deep retrofit. Façade replacement, electrification, full MEP rebuild, 18-month renovation requiring temporary tenant relocation. All-in cost at $180 per square foot: $54 million. Operational-carbon reduction: 60 percent. Rent uplift after re-tenanting: $4 to $7 per square foot. IRR: positive but extended, sensitive to lease-up.
Option C: Light retrofit plus instrumentation. Connectivity-as-a-service, predictive operations, tenant engagement layer, deep building-level instrumentation. All-in cost at $22 per square foot: $6.6 million. Operational-carbon reduction: 18 percent. New digital revenue line: $0.85 per square foot per year. Documented data package supporting CRREM, GRESB, and SBTi-aligned reporting. IRR: positive within twelve months, with material cap-rate compression at exit.
Option C does not preclude Option B. Done first, it informs whether B is the right next move. In most cases, the data shows that the marginal IRR on B, executed after C, is meaningfully better than B done blind.
The lease-rollover wrinkle.
Retrofit-timing is more important than retrofit-depth on most institutional assets. The reason is mechanical: the cash flow uplift from retrofit accrues at the next rent rollover, not at the moment of completion. If a major tenant has just signed a ten-year lease at pre-retrofit rents, the retrofit-IRR clock on that tenant’s leased area doesn’t really start for ten years.
The owner who instruments early, accumulates two years of performance data, and then executes the deep retrofit on a major-tenant lease-rollover window captures three things at once: a freshly-renovated asset, a major-tenant lease at post-retrofit rents, and a portfolio of accumulated performance data to support the new pro forma. The owner who tries to do everything at once typically captures only the first.
This is also a practical reason for the instrument-first approach. The instrumentation runs with the asset across rollovers. The deep capex is best aligned to the rollover calendar. The two work on different clocks.
The neighborhood-opposition angle.
Teardown-and-rebuild has gotten harder, politically, since 2020. The reasons are local, complicated, and not going to reverse in the next decade. Most US gateway markets have seen real estate sit at the center of housing-affordability politics; the politics that follow are not friendly to demolition-and-luxury-rebuild narratives. The entitlement timelines have lengthened. The CEQA and SEQRA litigation risk profiles have widened. The community-benefits negotiations have grown more material.
Retrofit projects don’t trigger these dynamics, or trigger them at a much smaller scale. The asset stays in place. The tenants stay in place. The neighborhood gets a better building, not a longer construction-impact period. The political risk profile is structurally lower. The financing risk profile follows.
This is not the dominant consideration on most underwrites, but it’s not nothing. For institutional sponsors operating in San Francisco, Los Angeles, New York, Seattle, Portland, Boston, or DC, the political-risk delta is real.
A decision framework.
For owners of older Class B and Class C commercial inventory, here’s the disposition I’d recommend this quarter. It’s three steps.
One. Get an embodied-carbon-plus-operational-retrofit assessment on your asset. The numbers above are reasonable national averages but every asset is its own case. The assessment costs $30,000 to $80,000 for a typical 300,000-square-foot asset and answers the questions you’d otherwise underwrite blind.
Two. Instrument first. Light-retrofit-plus-instrumentation is the lowest-regret next move on almost every asset I’ve reviewed. The capex is small, the payback is fast, the data informs every subsequent decision, and the cap-rate compression at exit is real.
Three. Time the deeper retrofit to the lease-rollover calendar. The data accumulated over the first two years tells you which deeper-retrofit components actually pencil at your tenant mix and your hold horizon. The deeper retrofit then runs on a tighter underwrite than it would have at year zero.
This is not the answer that makes for impressive corporate sustainability marketing. It is the answer the numbers reward. Thirty years of putting up buildings taught me that markets reward the path with the better risk-adjusted return. For most Class B inventory in 2026, that path is instrument first, retrofit deep on the rollover.
Sources cited
- Carbon Leadership Forum. (2017, updated 2024). Embodied Carbon Benchmark Study. University of Washington.
- Sturgis, S., & Roberts, G. (2010, multiple editions). Targeting Zero: Embodied and Whole Life Carbon Explained. RIBA Publishing / Routledge.
- Rocky Mountain Institute. (2022). Reinventing Existing Buildings; Retrofit Depth series. rmi.org.
- Architecture 2030. Carbon Smart Materials Palette; The 2030 Challenge. architecture2030.org.
- American Institute of Architects. AIA 2030 Commitment, annual reports.
- National Renewable Energy Laboratory. 2050 Buildings Stock Analysis.
- Federal Reserve Bank of St. Louis. Producer Price Index, Commercial Building Construction (FRED Series WPUSI012011).
- City of New York. Local Law 97 of 2019 and subsequent emissions-limit rules. nyc.gov/buildings.
- California. (2023). SB 253 Climate Corporate Data Accountability Act.
- European Commission. Corporate Sustainability Reporting Directive (CSRD).
- U.S. Department of Energy. C-PACE Property Assessed Clean Energy program reference.
- International Finance Corporation. EDGE Buildings methodology. edgebuildings.com.
Jack Illes is CEO of Smart City Labs. The SCL diagnostic engagement, on a single asset, produces the embodied-carbon-plus-retrofit-depth assessment described above in roughly six weeks. Talk to our team.