The decision, the way the market tells it
1290 Avenue of the Americas is a 2-million-square-foot Class-A office tower in Midtown Manhattan, owned by Vornado Realty Trust, a publicly traded real estate investment trust (a REIT is a company that owns income-producing property and answers to public shareholders). It is exactly the kind of asset that shows up in a decarbonization pitch deck.
The number driving that pitch is public and specific. A widely circulated projection puts this building's Local Law 97 penalty at roughly $835,000 per year for the 2030 to 2039 compliance period, rising to as much as ~$1.3 million per year after that. (Local Law 97 is New York City's rule that caps how much greenhouse gas a large building may emit each year and fines the rest.)
If you underwrite New York office, you have seen this line item. It arrives as a row in a model, “annual LL97 penalty, 2030+,” and it does real work: it sizes a capital budget, it justifies a retrofit, it moves a hold-or-sell decision.
We didn't set out to catch anyone. We ran the decision the way an investment committee actually uses it, “is this projected penalty a sound basis for a capital decarbonization decision?” through a governed framework that does one thing: it refuses to let a decision advance faster than its evidence allows, and it tells you precisely what evidence would let it. Here is what came back, using nothing but public records, including the building's own 14 years of mandatory energy disclosure, pulled and verified by the framework itself.
Who moves the emissions matters as much as the emissions.
The one question that governs the decision
The projection answers a simple implicit question: “this building is over its cap, so how big is the fine?” The framework's governing question for a commercial building of this type is different, and it is a standing prior for this whole asset class, not a discovery about this one address. For every large multi-tenant building it screens, the framework asks the same thing first:
Does the owner actually control the dominant load being penalized, and is the number a property of the building, or an artifact of the year it was measured and a grid factor the owner doesn't control?
In plain terms: in a full tower of tenants who run their own floors, their own hours, and their own equipment, who moves the emissions matters as much as the emissions. And a penalty is only worth what its input year is worth, the single slice of consumption the whole projection is built on.
The reason this class-level question lands hard here is concrete, and it is what the rest of this case is about: the projection everyone cites is a calculator output applied to one year of consumption (2022), and the building's own disclosure shows that number is anything but stable. Hold the question. Everything below is what the public record can, and cannot, say about it.
What the public record actually shows
New York City has required this building to report its measured energy use every year since 2010, under Local Law 84. That is not a model. It is the meter.
| Year | Site EUI¹ | Weather-normalized² | ENERGY STAR³ | Declared occupancy⁴ |
|---|---|---|---|---|
| 2010 | — | 117.6 | — | — |
| 2011 | — | 117.5 | — | — |
| 2012 | 115.5 | 112.2 | 75 | — |
| 2013 | 120.0 | 118.5 | — | 100% |
| 2014 | 101.9 | 101.4 | 70 | 100% |
| 2015 | 107.1 | 104.0 | 70 | 100% |
| 2016 | 108.3 | 106.1 | 67 | 100% |
| 2017 | 107.8 | 109.1 | 68 | 100% |
| 2018 | 120.1 | 117.7 | — | 100% |
| 2019 | 105.7 | 105.7 | 55 | 100% |
| 2020 · pandemic | 93.8 | 92.0 | 53 | 100% |
| 2021 · pandemic | 93.1 | 91.9 | 57 | 95% |
| 2022 · the penalty's input year | 104.7 | 103.4 | 58 | 95% |
| 2023 | 96.5 | 100.1 | 60 | 95% |
¹ Site EUI (kBtu/ft²): energy used per square foot per year, the building's “miles per gallon.” Lower is better. ² Weather-normalized: the city's own correction for hot and cold years, so you compare the building to itself, not to the weather. ³ ENERGY STAR (1–100): how this building ranks against similar buildings nationwide, a percentile, not a measurement. ⁴ Declared occupancy: the leased share the owner reports, not how many people were actually in the building. Note it barely moves (95–100%) while the energy number swings wildly.
Weather-normalized EUI removes the weather, so what you see is the building compared to itself. The series swings from 91.9 to 118.5, about a quarter of the value, with no capital project in between. The shaded band marks the pandemic years; the marker is 2022, the single year the penalty projection is built on.
Reported for calendar year 2020: 14,470.9 metric tons of CO₂-equivalent emissions, across 1,928,127 ft² of office (built 1963).
What the series says, stated precisely, and no more: over the record the weather-normalized number swings from 91.9 (2021) to 118.5 (2013), a ~27-point range, roughly a quarter of the value, with no capital project in between. Declared occupancy explains almost none of that swing. It sits at 95 to 100% the entire time while the energy number moves 27 points, so the disclosed occupancy field is not the variable that moves the meter, and the variable that does, how densely the building is actually used, is not in the public record at all. The pandemic years dropped hard (to ~92), and then 2022 rebounded to 103.4, essentially back to the pre-pandemic 2019 level (105.7). So the penalty's input year is not an artificially low pandemic reading. It is one point on a line that bounces, and 2023 fell again to 100.1.
The projected penalty is pinned to a single year of a volatile series that no party has reconciled to a defined normal year of operation. Pick 2021 and the building looks efficient. Pick 2018 and it looks far worse. Pick 2022, as the projection did, and you get the number in the press. The public record can prove the year is a choice. It cannot tell you which year is the building's true baseline, because that needs evidence the public record doesn't hold.
Same building, same meter. The only thing that changed is which year you point at. The projection pointed at 2022.
Projected and assessed are not the same word.
The formula, and the two inputs the owner does not control
The penalty itself is arithmetic:
Two of those inputs are not the owner's, at least not on the timescale of the decision. The grid emissions factor. A building's emissions aren't measured directly; they're calculated from its electricity use times a coefficient for how dirty the grid is. As New York's grid gets cleaner, that coefficient falls, and the same building's reported emissions fall with it, with no retrofit, no capital, no change on site. The owner does not control the grid. And the future caps: the cap this building is measured against tightens on a schedule set by the city, not the owner.
The run's own assumption register names exactly these: that “the utility grid emissions factor applied to the building's electricity will remain at present levels through the compliance period,” and that “tenant mix and occupancy density will remain comparable.” Both are assumptions inside the projection, not facts, and both sit outside the owner's control. A capital plan justified by the penalty is, in part, a bet on variables the owner cannot move.
Projected vs. assessed: the framework searched for an actual penalty on this property and found none. The city's Department of Buildings only begins assessing LL97 penalties for calendar year 2024 emissions from May 1, 2026. Every figure in circulation for this building, the $835k and the $1.3M, is a projection from the Building Energy Exchange calculator applied to 2022 data, as reported by City Limits. It is a forecast, not a bill. Projected and assessed are not the same word.
Why the honest answer is validate first
Here is where most analysis stops, with a shrug: “we can't really know until 2026.” A governed read does the opposite. It states exactly how much it knows, on a fixed ladder, and exactly what would move it up.
| Level | What it means | This building |
|---|---|---|
| L0 | we haven't observed it | site boundary, floor area, operating schedule, tenant control, utility bills, primary fuel |
| L1 | someone told us / a benchmark | a handful of context fields |
| L2 | typical for this kind of building | one structural prior |
| L3 | documented for this building | measured EUI, weather-normalized EUI, ENERGY STAR, emissions (from LL84) |
| L4 | independently verified / metered | — |
A chain is only as strong as its weakest link. The framework calls it the meet: a cluster can't be graded higher than its weakest member.
The building's energy numbers reach L3, they are the city's mandated measurement, and the framework ingested them. But a capital decision doesn't run on the energy number alone. It runs on a cluster of variables. The “fuel & energy” cluster needs four things: the measured EUI ✅ (L3), the emissions ✅ (L3), the utility bills ❌ (L0), and the primary-fuel / control picture ❌ (L0). Two of four are documented. Two are not seen. So the cluster, and with it the decision, is held at the floor. In this run, every one of the ten evidence clusters sits at L0, and the count is blunt: 6 variables at L3, 1 at L2, 3 at L1, and 80 at L0. Of the three buildings we ran this way, this one reached the fewest documented variables, and it still had the city's full 14-year series.
The verdict: Decision-Blocked Asset Brief, posture validation-first. Read that precisely. It is not “no.” It is not “the penalty is fake.” It is: “not on this evidence, validate the projection before you size capital against it, and here is exactly what closes the gap.” The framework holds the number in validate until a real baseline exists. Across the run, the only action it marks act-now is a single one: request the operator's evidence. Everything else, eight separate investigation lanes, stays open until that evidence arrives.
The public record can frame the question. Only your compliance folder can answer it.
New York already made you produce the evidence that settles this
You are not the public. You hold the documents the public record doesn't. And here is the quietly useful thing about New York in particular: the city already required you to produce the exact evidence that would settle this. It is sitting in a compliance folder, under the same law family that generated the penalty in the first place.
| The law | What it made you produce | What it lifts |
|---|---|---|
| LL84 · Benchmarking | your annual measured energy number | already used → L3 |
| LL87 · Energy audit + retro-commissioning | a systems-level audit of where the energy actually goes | the systems cluster |
| LL88 · Sub-metering + tenant statements | who controls and pays for each load | the control boundary, the governing variable |
| LL97 · Emissions cap + penalty | your filed compliance basis and applicable cap | the regulatory cluster |
LL84: buildings over 25,000 ft² report measured energy every year; this is the series in the table above. LL87: buildings over 50,000 ft² complete an energy audit and retro-commissioning every 10 years, and that audit is the “why” behind your EUI. LL88: sub-meters in tenant spaces over 5,000 ft², with regular tenant statements; for a Class-A office this is, almost exactly, the control-boundary evidence the framework is asking for, showing which loads are yours and which are the tenants'. LL97: penalty = (reported emissions − your cap) × $268 per ton over the limit, assessed annually, first for CY2024 from May 1, 2026.
What your own evidence does
| You add | The variable that moves | The verdict does this |
|---|---|---|
| your LL88 sub-metering + lease matrix | control boundary: L0 → asset-documented | the governing question gets an answer: how much of the penalized load is even yours to fix |
| your LL87 audit | systems: L0 → asset-documented | the “where the energy goes” gap closes |
| 12 to 24 months of utility bills | utility bills / primary fuel: L0 → asset-documented | the fuel-energy cluster clears its weakest link |
| a defined stabilized baseline year | the input-year problem | the penalty stops being “2022's number” and becomes the building's number |
This is not hypothetical. The framework's stated minimum to discriminate between the rival explanations is precise: a tenant metering map, a lease responsibility matrix, utility bills, and the LL97 filing basis. Same building, same framework. The difference is not a better model, it's your own paperwork. The tool gets more precise as you feed it, not more vague, and it never fills a blank for you: if you don't have a document, it stays a blank, named, on the list.
The honest verdict
On public evidence alone, three things hold, and no more. The penalty projection rests on a single, unstabilized year: the building's own 14-year disclosure swings from ~92 to ~119 on the city's weather-normalized measure while declared occupancy barely moves, so 2022 is a choice of year, not an established baseline, and the framework holds it in validate, not reject. Two of the penalty's inputs aren't the owner's to control, the grid emissions factor and the future caps. Projected is not assessed: no penalty has been charged; the number in circulation is a calculator output, and the city's first assessment is still ahead. None of this says the building will owe nothing, or that anyone acted in bad faith. The point is narrower and more useful: a capital decision this size deserves to know whether its penalty line is a property of the building or an artifact of one year and a grid it doesn't control, before the committee votes, not after. Honestly stated: this asset reached the fewest documented variables of the cases we've run this way, which is itself the finding.
Method: run on the źlab Operational Truth Framework (64 motors, full pipeline) using only public sources: NYC LL84 benchmarking (14 calendar years, 2010 to 2023, fetched and verified by the framework); NYC DOB Local Law 97 penalty structure; the Building Energy Exchange projected-penalty calculator as reported by City Limits. Every figure above is quoted from its source; where the record is silent, we say so. The framework's verdict (Decision-Blocked Asset Brief, validation-first), its evidence ceiling (NOT_OBSERVED), and the evidence that would lift it are reproduced from the run (run:255c3682ebe85082), not paraphrased. The governing “does the owner control the dominant covered load” question is the framework's standing prior for commercial buildings, applied here, not a discovery about this specific address.
Sources: NYC Open Data · Energy & Water Data Disclosure (Local Law 84), BIN 1034510 / BBL 1012670001 · NYC Department of Buildings, Local Law 97 penalty structure and compliance schedule · Building Energy Exchange projected-penalty calculator (2022 consumption basis), as reported by City Limits · Vornado Realty Trust (owner of record).
Your asset could be the next read.
A governed read of your own decision, built with the same discipline as this one, and it gets sharper the moment you add the evidence you already hold.