Office Lease Analysis: Gross vs NNN vs Modified Gross — and the Expense Stop / Base Year Mechanics That Decide Who Pays What

Key Takeaways

Why Lease Structure Is a Daily Underwriting Decision

In multi-tenant office, the lease structure determines who absorbs operating expense inflation, who carries vacancy-related expense distortion, and how Year 1 NOI is calculated. The 2026 environment makes this consequential in a way the 2010s SERP literature does not capture. Per CBRE's office operating-cost research, property insurance was up 17.4% in 2024 and has continued double-digit growth through 2025–2026, cumulative since 2023 in the 30–50% range across major metros; property taxes were up 4.3% in 2024 and have sustained mid-single-digit growth as municipalities replenish post-COVID budget gaps; utilities are elevated as AI and data-center load strain regional grids. The JLL Q1 2026 U.S. Office Market Dynamics report documents the trend at the institutional sell-side and the Cushman & Wakefield Q1 2026 U.S. Office MarketBeat confirms it at the metro level.

A tenant on a triple-net structure absorbs all of that inflation by contract. A tenant on a modified gross structure with a stale base year — signed in 2018 against a $18/SF Year 1 OpEx that has grown to $24/SF in 2026 — absorbs the cumulative escalation against that 2018 floor. A landlord on a full-service gross structure with no escalation mechanism absorbs the inflation themselves and watches NOI compress. Per the Investment Grade Q1 2026 net-lease cap rate report, office NNN cap rates sit at 7.90%, down 10 bps from Q4 2025 but still 242 bps over the 10-year Treasury at 4.38% — the widest spread in the recent cycle, reflecting the market's pricing of tenant credit risk under sustained OpEx pass-through pressure.

The institutional discipline is to trust nothing in the broker's lease summary until you have read the economic terms back to the rent roll, the base year, the expense stop, the gross-up provision, and the OpEx pass-through clause. The same physical building under three different lease-structure assumptions produces three different Year 1 NOI numbers — and at a 7% cap rate, the spread is north of $20M in valuation. Lease structure is a valuation input. This article walks the math.

The Three Structures: FSG, MG, NNN

The three structures define a spectrum of risk transfer between landlord and tenant. Full-service gross leaves the landlord fully exposed; triple-net pushes the entire expense base to the tenant; modified gross is the negotiated middle where the structure depends on the specific clauses. Holland & Knight's March 2026 "Who Pays for What" analysis frames the legal-economic distinction precisely: under NNN and MG, "the party at risk if 'extras' increase is the Tenant"; under FSG, "the party at risk if 'extras' increase is the Landlord."

Full-service gross (FSG). Tenant pays one number — the face rent — and the landlord absorbs every operating expense behind it: property tax, insurance, utilities, janitorial, security, management fee, repairs, maintenance, landscaping, snow removal, HVAC and elevator service, administrative. The historical office standard in major-metro Class A pre-2015. Tenant has perfect cost certainty; landlord prices the expected OpEx into the face rent and absorbs the variance. In a stable OpEx environment, FSG is competitive because the landlord earns the spread between underwritten OpEx and actual. In a 2026-style inflation environment, FSG compresses landlord NOI as actual OpEx outruns the underwriting assumptions that were priced into the face rent.

Modified gross (MG). The contested middle. Tenant pays base rent plus escalations above a defined baseline. The baseline is either a base year (the actual Year 1 OpEx, used as the floor) or an expense stop (a negotiated $/SF figure that may or may not equal Year 1 actual). Above the floor, the tenant absorbs escalations on a pro-rata share basis. Most common in mid-tier office and in major-metro Class A signed pre-2020 when the structure was the negotiated compromise between FSG and NNN. MG admits sub-types: base-year stop, fixed-dollar expense stop, hybrid with separate stops for property tax, insurance, and CAM. Two MG leases in the same building can have wildly different economics depending on the base year vintage, the stop type, the gross-up clause, and the exclusion list. Always read every MG lease clause by clause.

Triple net (NNN). Tenant pays base rent plus its pro-rata share of all three operating expense categories — property tax, insurance, and common area maintenance (CAM, the catch-all for everything else). The "three nets" framing is the structural origin of the name. Common in single-tenant retail and industrial; increasingly common in multi-tenant office Class A and medical office post-2020 as landlords push risk to tenants in the OpEx-inflation environment. Sub-variants in office: absolute net (tenant carries even structural and capital expenses, rare in office), double net (tenant carries tax and insurance but landlord retains CAM, occasional in single-tenant office), and modified NNN (tenant carries all three but with capped escalations or specific category exclusions).

The comparison at a glance:

Table 1. The three office lease structures on the spectrum from full landlord risk to full tenant risk. Modified gross occupies the contested middle where two leases in the same building can have materially different economics.

Base Year — and the Institutional Games

The base year is the dollar amount of operating expenses in the lease's defined first year, used as the floor against which subsequent-year escalations are calculated. Tenant pays base rent flat each year; landlord recovers the dollar difference between the current-year OpEx and the base-year OpEx as additional rent. The math is simple at the surface: $18/SF base year, $24/SF current year, $6/SF escalation that the tenant pays in addition to base rent. The institutional reading is harder, because the games landlords play with the base year compound the further into a lease term the tenant goes.

The institutional games begin at lease signing and compound over the term.

The stale base year. Landlord signs a new tenant in Year 5 of the building's life against the Year 5 base year — which captures four years of accumulated OpEx inflation as the new tenant's immediate escalation baseline. Tenant is structurally disadvantaged from Day 1. The market convention is that the base year should be the lease's Year 1, not the building's Year 5; sophisticated tenant reps negotiate "year of occupancy" base year (the actual first year of the tenant's occupancy) and push back against landlords who try to lock in the property's recent peak as the floor.

Tax base year manipulation. The property tax bill in a building's first year of operation is often abnormally low — either the building was under construction for part of the year, the assessment was based on land-only value, or the assessor hasn't yet reassessed at the stabilized value. Landlords sometimes sign tenants against that pre-stabilization tax base year, knowing the assessed value will normalize upward in the following years and the tenant will absorb the full normalization as escalation. The institutional discipline is to pull assessor records, identify when the property was first assessed at its stabilized value, and require that the base year tax bill reflect the stabilized assessment — not the construction-period assessment.

Insurance carrier swap. Landlord places insurance in the base year with a high-cost carrier (or at the peak of the insurance cycle), then switches to a lower-cost carrier in subsequent years without adjusting the base-year insurance figure. The "actual" Year 2 insurance is now lower than the base-year figure on paper, but the tenant has no visibility into the carrier change, and any genuine market-wide insurance inflation that comes in subsequent years escalates against the artificially high base-year floor. The tenant absorbs less than the actual inflation; this game cuts in favor of the tenant when discovered, but landlords routinely set up the symmetric version — placing insurance in the base year with the cheapest carrier they can find, then switching to a higher-cost carrier in Year 2 and passing the differential through as inflation. The carrier-change disclosure is the discovery point.

Asymmetric gross-up. Landlord grosses up the operating expenses in escalation years (to full occupancy, capturing what variable OpEx would have been if the building were full) but does not gross up the base year. The escalation arithmetic now compares a grossed-up current year against a raw base year. The tenant pays an inflated escalation that captures the building's vacancy as if it were occupancy. The institutional discipline is to require symmetric gross-up application: either both years grossed up or neither.

Tenant protections. Sophisticated tenant counsel negotiates three protections. The rolling base year: the base year resets each year to the prior actual, so the tenant only absorbs year-over-year inflation, never cumulative inflation against a stale floor. The constant-dollar base year: the base year is inflation-adjusted using CPI or a negotiated index, neutralizing the stale-base-year game. The fair-and-equitable gross-up symmetry clause: the gross-up applies symmetrically across the base year and all escalation years. These three protections, when stacked, neutralize roughly 80% of the seller-side base-year games. Few mid-tier landlords offer them in the standard form; they are won at the LOI stage.

Expense Stops vs Base Year

The expense stop is the base year's negotiated cousin. Rather than defining the floor as "the actual Year 1 OpEx," the expense stop defines it as "$X per square foot, regardless of actual." The distinction matters at the clause level because the expense stop decouples the landlord's recovery from the landlord's actual cost — in either direction. The convention in office literature (Adventures in CRE's glossary, PropertyMetrics, Commercial Real Estate Loans glossary) is that an expense stop is "the ceiling on landlord exposure and the floor on tenant exposure" — an accurate framing, but the institutional reading goes one level deeper into the sub-types.

Base-year stop. The stop is set equal to the actual Year 1 OpEx. The most common sub-type and the form that gets conflated with "base year" in the broker-tier literature. Operationally identical to a base year for the first year; the distinction surfaces in subsequent-year reconciliation mechanics and audit rights, where stop language and base-year language can have different procedural implications. Tenant-friendly when Year 1 is a high-OpEx year (the floor is set at the peak); landlord- friendly when Year 1 is a low-OpEx year (the floor is set at the trough).

Fixed-dollar stop. The stop is a negotiated number independent of actual — for example, $8.50/SF. Used in build-to-suit and master leases where the tenant wants absolute certainty about the floor. If actual Year 1 OpEx comes in at $9.50/SF, the landlord absorbs the $1/SF gap. If actual comes in at $7.50/SF, the landlord pockets the $1/SF spread. The tenant has perfect certainty about the floor; the landlord has perfect risk against the gap. Sophisticated landlords negotiate fixed-dollar stops with escalators (the stop itself grows 2–3% annually) to neutralize the inflation risk.

Component stops. Separate stops for each category — property tax, insurance, and CAM. The sophisticated tenant negotiation; protects against one expense category running materially hot while the others stay tame. For example, the tenant can negotiate an insurance stop at the trailing three-year average insurance figure rather than the most recent year, neutralizing the 2024 insurance spike. Common in 100,000+ SF Class A office leases negotiated with experienced tenant counsel.

The clause-level distinction between expense stop and base year matters for audit rights, gross-up application, and the renewal economics question. A base year resets every time the tenant signs a new lease (or renewal); an expense stop is negotiated separately and can be carried forward across renewals at the original number if the lease provides for it (rare; usually the renewal resets the stop). The institutional discipline on diligence: confirm the lease's defined floor type, the floor's vintage, the sub-type (fixed-dollar vs base-year vs component), and the renewal carry-forward language. Three questions, three minutes, materially different economic implications.

The Gross-Up Provision

The gross-up provision is the lease's mechanism for neutralizing vacancy-related distortion in operating expense reconciliation. The problem the gross-up solves: in a 70%-occupied building, the variable operating expenses — electricity in occupied suites, janitorial service, management fees that scale with revenue, supplies — are artificially low because the vacant 30% of the building is not consuming them. The base year of a tenant who signed when the building was 70% occupied would lock in an artificially low floor, and the landlord would be exposed when occupancy improved and variable OpEx normalized upward against the stale floor. The gross-up adjusts the variable OpEx to what it would have been at full occupancy — or, more commonly, at a 95% occupancy benchmark — then charges each tenant their pro-rata share of the grossed-up number.

The canonical legal reference is Holland & Hart's gross-up provisions white paper, which establishes the variable-vs-fixed distinction at the legal level. The institutional practitioner reference is AQUILA Commercial's tenant-rep-side analysis, which walks the mechanics at the clause level from the tenant's perspective.

The institutional discipline checks four things in every gross-up clause:

One: variable-only application. Gross-up applies only to variable expenses (utilities, janitorial, management fees, supplies, repairs that scale with occupancy). Fixed expenses — property tax, insurance premiums, depreciation, debt service — do not vary with occupancy and should not be grossed up. Landlords sometimes apply gross-up to fixed expenses (intentionally or through poor reconciliation software). This is a pure overcharge; property tax does not increase because the building leased up. Severity: material concern, surfaces on every CAM audit where the gross-up clause is loosely drafted.

Two: the gross-up threshold. Institutional convention is 95–100% of full occupancy. A lower threshold (85%, 90%) inflates the tenant's share because the gross-up captures more of the building's vacant variable OpEx as if it were occupied. The market standard is 95%; 100% is landlord- aggressive; below 95% is tenant-aggressive. Verify the threshold in the lease language.

Three: symmetric application across the base year. The gross-up must apply both to the base year (or the expense stop reconciliation) and to escalation years. Asymmetric application — grossing up escalation years but not the base year — inflates the tenant's escalation by capturing the building's vacancy gap as if it were inflation. This is one of the highest-frequency seller-side games on MG and NNN reconciliations in 2026, particularly on buildings that leased up materially between the base year and the audit year. Symmetric application is the institutional standard.

Four: occupancy measurement methodology. The gross-up calculation requires the landlord to define "occupied" — is it leased, occupied, or revenue-generating? Each gives a different answer. A suite that is leased but in tenant improvement build-out is not consuming variable OpEx; counting it as occupied inflates the gross-up base. The institutional discipline is to require "revenue-generating occupancy" as the gross-up base — the suite has a paying tenant in occupancy.

CAM, Pro-Rata Share, and the Pass-Through Math

Common area maintenance (CAM) is the catch-all for the operating expenses that aren't property tax or insurance. In a multi-tenant office building, CAM includes utilities for common areas (lobby, corridors, elevators, garage), janitorial for common areas, security, repairs and maintenance of common-area systems (HVAC, elevator, fire/life safety), landscaping, snow removal, exterior cleaning, management fee, administrative cost, supplies, and lobby attendant or concierge salaries. CAM is the engine of MG and NNN pass-throughs; the institutional reading walks the math.

Pro-rata share. Tenant SF divided by rentable building SF. The BOMA standard is rentable SF (which includes load factor for common areas) rather than usable SF (the tenant's direct space) or leasable SF (whatever the landlord defines). The BOMA International standard for office floor measurement (ANSI/BOMA Z65.1) defines rentable area precisely; landlords sometimes deviate from it to inflate the denominator selectively. Verify the lease's pro-rata share definition references BOMA-rentable area explicitly, not "the building's total square footage as determined by Landlord."

Estimate / true-up cycle. Landlords typically collect monthly CAM estimates against the budget, then reconcile to actuals at year-end. The reconciliation produces either an additional payment due from the tenant (if actuals exceeded budget) or a credit to the tenant (if budget exceeded actuals). The reconciliation is delivered to the tenant typically within 90–180 days of year-end; tenant has a defined window (typically 30–90 days, sometimes one year) to dispute. Per CapVeri's institutional research on CAM reconciliation, over 70% of tenants contest at least part of their CAM reconciliation each year. The contestable surface is large because the line between operating and capital is fuzzy, the gross-up methodology is rarely documented, and the inclusion/exclusion list is often interpreted ambiguously by landlord accounting.

Common inclusions. Property tax, insurance premiums, utilities (electric, water, sewer, gas), janitorial, security, management fee (2–4% of EGI is institutional norm per BOMA Experience Exchange Report), repairs and maintenance, landscaping, snow removal, HVAC service contracts, elevator service contracts, fire/life safety, supplies, lobby attendant, administrative cost.

Common exclusions. Capital expenditures (HVAC replacement, roof, elevator modernization, building system replacement), leasing commissions, tenant improvement allowances, free rent (concessions), executive compensation, marketing and promotional costs, depreciation, debt service, penalties and violations, capital reserves, advertising, ground lease payments (separately recoverable in most leases), and the landlord's corporate-level overhead. The exclusion list is where the games happen — landlords routinely push capital items into operating, particularly HVAC replacement (the capital line item with the highest dollar magnitude on a 30–40 year-old building).

The institutional CAM reconciliation discipline: every reconciliation should be reconciled to (a) the prior year's actual reconciliation (line-by-line variance), (b) the BOMA EER benchmark for the building's class and metro, (c) the IREM Income/Expense IQ data, (d) the lease's explicit inclusion/exclusion list, and (e) the budget vs actual variance with line-item explanation. The diligence packet is four documents, the institutional reading is line-by-line, and the contestable surface is roughly 70% of reconciliations.

The 2026 OpEx Inflation Overlay

The 2026 OpEx environment is the freshness wedge that distinguishes this article from the 2018-era SERP literature. Property tax, insurance, and utilities are all moving up materially — not in coordination, but in compounding succession — and the flow-through implications differ sharply by lease structure.

Table 2. The 2026 OpEx inflation overlay. Property insurance is the standout — up 17.4% in 2024 with cumulative 2023–2026 movement in the 30–50% range across major metros. Property tax and utilities compound on top. CPI explains approximately 95% of OpEx movement on a 12-month lag per CBRE.

The CPI lag is the underwriting wedge: CBRE's research establishes that roughly 95% of OpEx movement is explained by lagged CPI, with a 1% CPI move flowing through to approximately 1.3% OpEx movement about 12 months later. The 2024–2025 CPI runs of 2.7–3.5% therefore predict 3.5–4.5% OpEx growth into 2026, and the property tax / insurance asymmetry adds an additional structural increment on top.

The flow-through by structure:

• FSG. Landlord absorbs 100% of the inflation. In a 33% cumulative OpEx inflation environment, FSG-heavy buildings show 8–12% NOI compression vs original underwriting absent renegotiation. The leverage point: as FSG tenants roll, the landlord can convert to MG (or to NNN with appropriate face-rent compensation) and arrest the compression. Buildings unable to convert because tenant credit allows the tenant to demand FSG renewal at face are the structural losers in the 2026 environment.

• MG with stale base year (pre-2020). Tenant pays escalations against an old floor. On a 2018 base year MG with 2026 reconciliation, the tenant absorbs the cumulative 2018–2026 inflation (~33%) as escalation. Renewal economics increasingly contested — tenants negotiate hard against rolling the lease forward at face because they know renewal on a new MG with a 2026 base year structurally resets their floor and drops their effective rent by $6–$8/SF. The landlord's renewal economics are now hostage to the base year reset.

• MG with current base year (signed 2024+). Tenant pays escalations against a recent floor; manageable, and consistent with how the structure was originally designed to work. These leases are not the source of CAM audit volume in 2026.

• NNN. Tenant pays 100% pass-through. Tenant credit risk rises as OpEx absorbs a larger share of the tenant's cash flow. Per Investment Grade Q1 2026, office NNN cap rates at 7.90% with 242 bps spread over the 10-year Treasury — the market's explicit pricing of tenant credit risk under sustained pass-through pressure. The landlord retains base rent durability but the tenant's downside is now structural OpEx exposure.

The institutional forward note: JLL's office outlook highlights AI-enabled predictive controls cutting energy and maintenance OpEx 10–30% on stabilized assets. Landlords on FSG — who absorb 100% of the savings — have the largest incentive to deploy. Landlords on NNN pass any savings through to the tenant in subsequent-year reconciliation; the economic incentive structure differs by lease type and affects capital allocation across a stabilized portfolio.

The CAM Reconciliation Red Flag Taxonomy

The institutional reading uses a structured taxonomy of CAM reconciliation red flags. Each red flag maps to the lease clause where it surfaces, the cross-check document that confirms or refutes it, the severity grade (Audit Trigger / Diligence Item / Material Concern / Deal-Killer), and the specific reconciliation or underwriting adjustment. The twelve flags below cover roughly 85% of the institutional patterns that surface on multi-tenant office CAM audits in 2026.

Table 3. The CAM reconciliation red flag taxonomy. Severity grades: Diligence (price into a buffer); Material Concern (specific reconciliation adjustment); Deal-Killer (renegotiate or walk at lease signing). Roughly 70% of institutional CAM reconciliations include at least one of these patterns; sophisticated tenant reps audit every reconciliation against the full taxonomy.

The seller-side patterns are not theoretical. The recurring patterns on multi-tenant Class A office in 2026: HVAC replacement billed as repairs in OpEx (capital-in-operating, flag #1); 5–6% management fees rather than the 2–4% institutional norm (flag #2); gross-up applied to property tax (flag #4); pro-rata denominator using leasable rather than BOMA-rentable (flag #6); base-year property tax captured from the under-construction year before stabilization (flag #9, the highest-severity); insurance carrier swapped without adjustment to the base-year insurance figure (flag #10). The institutional CAM auditor walks the twelve-flag table once per reconciliation; per CapVeri's research, the average institutional audit recovers 4–9% of the year's CAM pass-through — a material refund line.

Worked Example: 200,000 SF Class A Boston CBD, Three Tenants, Three Structures

The integrated worked example. Setup before reading the lease pool:

Tenant A: Law Firm. 50,000 SF (25% of building). FSG @ $65/SF, 10-year lease commencing 2026. The law firm signs a fresh FSG lease in 2026 at $65/SF face rent. No escalation; one number, all-in. Landlord absorbs the tenant's 25% pro-rata share of the building's 2026 OpEx: 25% × $4,800,000 = $1,200,000 absorbed annually. Tenant pays $65/SF × 50,000 SF = $3,250,000 in face rent. Landlord nets $3,250,000 − $1,200,000 = $2,050,000 from Tenant A. Tenant's effective rent: $65/SF (no escalation absorbed). The landlord has effectively quoted a face rent that prices in expected 2026 OpEx plus a margin — but in a 33% cumulative inflation environment, the law firm has perfect cost certainty and the landlord absorbs any further 2027–2035 OpEx growth on the absorbed share.

Tenant B: Financial Services. 75,000 SF (37.5% of building). MG with 2018 base year @ $58/SF base rent, Year 8 of 10-year lease. The financial services tenant signed in 2018 against the 2018 building OpEx of $18/SF as the base year. Tenant pays $58/SF base rent flat each year. Escalation in 2026: ($24 − $18) × 75,000 SF = $450,000 annually, or $6/SF on top of base rent. Tenant's effective gross rent in 2026: $58 + $6 = $64/SF. Landlord absorbs the 2018 base year level of OpEx as locked-in cost: 37.5% × $3,600,000 (2018 OpEx) = $1,350,000 absorbed annually. Landlord receives $58/SF × 75,000 SF = $4,350,000 base rent + $450,000 escalation = $4,800,000 from Tenant B. Landlord nets $4,800,000 − $1,350,000 = $3,450,000 from Tenant B. The 2018 base year is doing the work — Tenant B absorbs eight years of accumulated inflation against the 2018 floor.

Tenant C: Tech Company. 75,000 SF (37.5% of building). NNN @ $52/SF base + pro-rata OpEx, Year 3 of 12-year lease commencing 2024. The tech tenant signed in 2024 on a triple-net structure at $52/SF base rent plus pro-rata share of all operating expenses. 2026 pro-rata pass-through: 37.5% × $4,800,000 = $1,800,000, or $24/SF on top of base rent. Tenant's effective gross rent: $52 + $24 = $76/SF. Landlord absorbs zero of the tenant's pro-rata share (the pass-through is a wash). Landlord receives $52/SF × 75,000 SF = $3,900,000 base rent + $1,800,000 pass-through = $5,700,000 from Tenant C. Landlord nets $5,700,000 − $1,800,000 = $3,900,000 from Tenant C (the pass-through revenue cancels the OpEx the landlord paid and then recovered). Tenant C carries the full $24/SF OpEx exposure; if 2027 OpEx grows to $26/SF, Tenant C absorbs the full $150,000 incremental.

Table 4. The three-tenant, three-structure building. Total landlord NOI (before vacancy adjustment and shared common-area OpEx) is $9.4M on the mixed lease pool. The structural composition matters: the same physical OpEx flows through differently to each tenant.

The renewal economics for Tenant B. Year 10 (2028) of Tenant B's lease is the renegotiation. If Tenant B renews on a new MG with a 2026 base year, the floor resets up to $24/SF. Her face rent stays roughly at market ($58–$62/SF; assume $60/SF) but the escalation now applies only to 2027+ inflation, not the 2018–2026 accumulated $6/SF. Tenant B's effective rent drops from $64 back to roughly $60. The landlord absorbs the $4/SF reset (the difference between the prior $64 effective and the new $60 effective) — directly hits NOI. This is the renewal trap that landlords on MG buildings with multiple stale base years face simultaneously: the base year reset on each renewal structurally compresses landlord NOI by the cumulative inflation between the original base year and the renewal date. The institutional underwriter must model the base-year reset explicitly on every MG lease rolling within the hold period.

The reading for the institutional analyst: lease structure mix is part of the asset, not metadata. A building with 60% FSG, 30% MG-stale-base-year, and 10% NNN is a different asset for valuation purposes than the same building with 20% FSG, 30% MG-current-base-year, and 50% NNN. The cap rate the buyer applies will be different; the NOI durability under OpEx stress will be different; the renewal economics will be different. For the underlying valuation mechanics that this NOI flows into, see the cap rate calculator and formula article and the IRR calculator and formula article. For the TI/LC and effective rent overlay that runs alongside this lease-structure analysis, see the office underwriting: TI/LC, free rent, effective rent article. For repositioning class and trophy strategies that intersect with lease pool composition, see the trophy Class A vs Class B repositioning article.

The Institutional Underwriting Overlay

How an institutional underwriter reads the lease pool of a multi-tenant office building. The discipline is a structured cascade.

Step one: pull the lease abstract for every tenant. Verify the lease type, the base year (if MG) and its vintage, the expense stop (if MG) and its type, the gross-up clause language and the threshold, the audit rights clause, the inclusion/exclusion list, the pro-rata share definition, the permitted use clause, and the renewal options. Map each lease to the A.CRE office lease taxonomy and flag deviations.

Step two: reconcile the rent roll's lease type column against actual lease documents. Sample 100% of the top 10 tenants by SF and 25% of the long tail. Rent roll abstracts — particularly those produced by lease administration vendors — routinely miscode lease structure (an MG-with-tax- stop coded as full NNN; a modified-NNN coded as plain NNN; a fixed-dollar expense stop coded as a base-year stop). The miscodes are not small — they affect Year 1 NOI calculations by 5–15% on individual leases.

Step three: build a pro-forma OpEx flowing through each lease structure. For each tenant, calculate the absorbed-share of building OpEx (what landlord pays) and the recovered share (what landlord collects via base rent + escalation + pass-through). Net to the tenant-level NOI contribution. The sum is the building's underwritten Year 1 NOI before vacancy and shared common-area OpEx adjustments.

Step four: stress test 2026 OpEx +5% / +10% / +15%. Identify which tenants absorb the increase (NNN tenants and MG-above-base-year tenants), which the landlord absorbs (FSG tenants and the base-year-locked portion of MG tenants). Build the NOI sensitivity table. A building with 60% FSG / 40% NNN has dramatically different sensitivity to OpEx stress than a building with 20% FSG / 80% NNN.

Step five: look at the lease maturity ladder. Tenants on stale MG base years rolling in the next 24 months will renew at lower effective rents (the base-year reset structurally drops their effective rent because the new base year captures the cumulative 2018–2026 inflation). Model this as a downward effective-rent adjustment at renewal. On a building with substantial stale-base-year MG leases rolling in Year 1–2 of the hold, this is a Year-2–3 NOI step-down that the broker's proforma will not show.

Step six: look at lease structure mix vs market. A predominantly-FSG building in a market that has shifted to NNN signals tenant-rep negotiation leverage at renewal — tenants will push for NNN with lower face rent at renewal to match the prevailing market structure, and the landlord may absorb a structural NOI step-down on the conversion. Conversely, a predominantly-NNN building in a market where Class A is moving back to MG (rare but possible in tenant-leverage markets) signals structural NOI upside.

Step seven: cross-check OpEx assumptions against institutional benchmarks. The BOMA Experience Exchange Report publishes office OpEx benchmarks by class, metro, and building size. The IREM Income/Expense IQ publishes complementary data. CBRE, JLL, and Cushman publish quarterly market reports that overlay metro-specific context. The institutional discipline: every OpEx line item in the proforma should be cross-checked against at least two of these references.

Step eight: debt sizing. Lenders apply a vacancy stress (typically 5–10% of EGI) and an OpEx stress (typically 5–10% of OpEx) on top of underwritten. Lease structure determines whether the OpEx stress hits NOI (FSG) or is fully passed through to tenants (NNN). For office collateral going to CMBS conduit or SASB financing, Trepp's office CMBS data shows that buildings with predominantly-FSG lease pools have shown materially higher DSCR variance under OpEx stress in 2023–2026 — lenders price this risk into spread.

Five Mistakes Practitioners Make

• Treating "modified gross" as a single structure. MG is a category. Two MG leases in the same building can have wildly different economics depending on base year vintage, stop type, gross-up clause symmetry, and the exclusion list. Always read every MG lease clause by clause. Lease-administration vendor outputs that aggregate lease type at the building level systematically misrepresent MG variance.

• Ignoring the base year vintage. A 2018 base year in a 2026 lease is structurally penal to the tenant — the prior eight years of OpEx inflation are baked into immediate escalations against an unchanged floor. Always verify the base year vintage on every MG lease and model the base-year reset on every renewal within the hold period. The reset is a NOI step-down for the landlord that the broker's proforma will not show.

• Missing the gross-up provision asymmetry. Landlord grosses up escalation years but not the base year (or vice versa). Asymmetric application inflates the tenant's share by capturing the building's vacancy gap as if it were inflation. Always verify the gross-up clause applies symmetrically across base year and escalation years — and verify it applies only to variable expenses, never to fixed.

• Conflating face rent and effective rent. Two tenants on $58/SF face rent in the same building can be paying $58/SF effective (FSG) and $82/SF effective (NNN with full pass-through). Always net face rent to comparable effective rent using the building's current OpEx and the lease's pass-through mechanics. The institutional comp-set discipline is on effective rent, not face. For the full effective-rent walkthrough with TI/LC and free rent overlay, see the office underwriting: TI/LC, free rent, effective rent article.

• Skipping the CAM reconciliation history. A landlord with three consecutive years of contested CAM reconciliations signals a pattern, not an exception. Pull three years of reconciliations on diligence; if the building has a chronic dispute history, that history travels with the asset. Cross-check to the 12-flag red-flag taxonomy; recurring flags across multiple years are diagnostic of the landlord's reconciliation methodology and the audit volume to expect post-close.

From Lease Analysis to Institutional Pro Forma

The reading exercise above is the institutional discipline; the modeling step is what consumes it. Once you have walked the three-structure taxonomy, verified the base year vintage on every MG lease, audited the gross-up clause for symmetric application and variable-only scope, reconciled the pro-rata share definitions against BOMA-rentable, mapped the inclusion/exclusion list against the twelve-flag taxonomy, and stress-tested the 2026 OpEx environment by structure — the next step is the full institutional office pro forma. Year 1 NOI, debt sizing, value-add capex timing, stabilized Year 3, exit cap, levered IRR. We know you can build this in Excel from the lease abstracts and the OpEx benchmarks. The reader can do it in Apers in minutes.

Related Articles

• Office Underwriting: TI/LC, Free Rent, and Effective Rent — the leasing concessions and net effective rent overlay that runs alongside lease-structure analysis.
• Trophy Class A vs Class B Repositioning — the asset-class and repositioning strategies that intersect with lease pool composition.
• Cap Rate Calculator and Formula — the valuation mechanics that the lease-structure NOI flows into.
• IRR Calculator and Formula for Real Estate — the returns analysis the full pro forma supports.
• CMBS Conduit vs SASB: Defeasance and Yield Maintenance — the office CMBS financing context for lease-pool-dependent DSCR underwriting.
• Multifamily Underwriting Fundamentals: The Rent Roll Reading Exercise — the parallel institutional reading exercise for multifamily.

FAQ