Manufacturing & Flex Real Estate: Reshoring, R&D, and the 2024–2026 Capex Wave

Key Takeaways

Two Asset Classes, One Label

Manufacturing and flex sit at the two ends of the industrial spectrum, and they get lumped together as "industrial" on every cap-rate sheet that hasn't been updated since 2019. Manufacturing real estate is closer to single-tenant net lease: credit underwriting against a Tier-1 corporate tenant, 10–25 year initial term, heavy specialized infrastructure, tenant capex investment dwarfing the landlord basis. Flex real estate is closer to multi-tenant office: 25–50% office buildout, shorter terms, hybrid R&D / light manufacturing / lab / assembly use, mid-market and small-tenant credit, operational asset management. Neither underwrites like the modern distribution warehouse on the same broker pitch deck. The 2024–2026 reshoring wave has changed the manufacturing side meaningfully; the post-2024 biotech R&D reset has reshaped the flex side just as much.

The reshoring narrative is the loudest single story in industrial CRE since e-commerce. Per theReshoring Initiative, over 2 million US manufacturing jobs have been announced via reshoring and FDI since 2010, with 244,000 in 2024 alone and 88% of the 2024 cohort in high or medium-high tech industries. TheSemiconductor Industry Associationcounts 160+ announced projects across 30 states totaling more than $828B in private semiconductor-ecosystem investment since 2020. The SelectUSAprogram reports more than $535B in facilitated investment since inception. CHIPS Act grants, the section 48D advanced-manufacturing investment tax credit, and state-EDC incentive stacks have moved a generation of corporate capex onto US soil.

The institutional discipline is to read the announcement-side honestly. PerIoT Analyticsanalyzing Census Bureau Construction Spendingdata, US manufacturing construction spending is down about 21% from its June 2024 peak, driven by a 44% pullback in computer, electronic and electrical construction since July 2024. Excluding that category, manufacturing construction rose 5.6% in nominal terms (about 2.3% in real terms) between February 2025 and March 2026. Intel's Ohio One fab, slated to begin operations in 2026, is, per Intel's early-2025 update, now expected in 2030–2031. Samsung's Taylor, Texas fab sat more than 90% complete in early 2025 before an equipment installation ceremony was set for April 24, 2026. Stellantis took €22.2B in H2 2025 reset charges and agreed to sell its 49% stake in the NextStar Energy battery JV to LG Energy Solution. Reshoring Initiative founder Harry Moser put 2025 announcements at 244,000 jobs, essentially flat with 2024. The announcement side is holding; the gap is in announcement-to-completion conversion and megaproject execution. The narrative is real, and the delivery is uneven, and the institutional underwriter has to hold both.

What follows is the practitioner reference for the manufacturing & flex slice of industrial CRE. We walk the industrial spectrum end to end, distinguish manufacturing CTL underwriting from flex/R&D operational underwriting, ground the reshoring story in the dataset rather than the narrative, walk the biotech R&D Cambridge vacancy reset reshaping the flex side, and close with a dual worked example - a Phoenix advanced-manufacturing CTL on one end and a Cambridge Kendall Square biotech R&D flex on the other. Two deals labeled "industrial." Two completely different underwriting paths. Cap rates 75 basis points apart, buyer pools that barely overlap, and the 2024–2026 macro hitting each from opposite sides.

The Industrial Spectrum

The institutional industrial taxonomy is a continuum, not a binary. The collapsing of "industrial" into one column on the cap-rate sheet conceals five distinct subtypes with materially different office shares, lease terms, tenant profiles, TI requirements, and cap-rate ranges. The table below is the spine; every subsequent section of this article returns to it.

Table 1: The five-subtype industrial spectrum. Cap-rate and TI / fit-out PSF ranges are Apers illustrative assumptions, not survey data. Clean-room cost tiers followACH Engineering. The boldface rows are this article's central pair: flex/R&D and advanced manufacturing.

The 2024–2026 Reshoring Narrative - Factually

Treat the reshoring story as a dataset, not a politics. The 2022 CHIPS and Science Act gave the Commerce Department $50B for semiconductor programs, including $39B of manufacturing incentives and $11B for R&D (NIST), alongside an advanced-manufacturing investment tax credit underIRC section 48Dthat was 25% and that a 2025 law (Pub. L. 119-21) raised to 35% for property placed in service after December 31, 2025; the credit does not apply to property whose construction begins after December 31, 2026 (IRC 48D(e)). The 2022 Inflation Reduction Act added section45Xadvanced-manufacturing production credits, which the 2025 law then amended, including prohibited-foreign-entity restrictions. State EDCs in Arizona, Texas, Ohio, Georgia, the Carolinas, Indiana, Michigan, and Tennessee layer their own incentive packages on top of the federal stack. The capital intent, regardless of one's view of the policy, is concrete.

Announced volumes - what the data says

Per theReshoring Initiative 2024 Annual Report(Harry Moser's dataset), 244,000 manufacturing jobs were announced in 2024 via reshoring and FDI, and over 2 million jobs have been announced since 2010, of which an estimated 1.7 million have been filled. 88% of 2024 announcements landed in high or medium-high tech industries (90% in early 2025 data).

Per the Semiconductor Industry Association, the CHIPS-era pipeline carries 160+ announced projects across 30 states totaling more than $828.0B in private investment since 2020. The headline projects are familiar to any industrial acquirer: TSMC's three-fab Phoenix complex (Fab 21 phase 1 in volume since late 2024; phase 2 equipment install Q3 2026, production 2027); Samsung's Taylor, Texas fab; Intel's Ohio One in New Albany; Micron's Boise and Clay, NY expansions; GlobalFoundries' Malta, NY expansion; and the Asian-supply-chain follow-on of equipment, materials, and packaging investments around each anchor. The EV battery cohort layers in Ford BlueOval City (Stanton, TN, with SK On), GM Ultium (Lordstown, Spring Hill, Lansing), the Stellantis-LG NextStar Energy plant (Windsor / Indiana), and the Hyundai-LG Metaplant in Bryan County, GA. Biotech reshoring - mid-cycle Eli Lilly Lebanon, IN ($9B), Novo Nordisk Clayton, NC, AstraZeneca Maryland, and the federal-BARDA-funded contract-manufacturing buildout - layers in another vertical.

SelectUSA, a separate all-sector program at the US Department of Commerce, reports it has facilitated more than $535B of investment supporting over 290,000 jobs since inception. Per theCBRE 2026 US Industrial Outlook, Louisville, Nashville, Cincinnati, Chicago, Detroit, and Kansas City are the named manufacturing-expansion markets in 2026, the inland reshoring corridor. In Apers' read, Phoenix, Austin, Atlanta, and the Carolinas anchor the semiconductor and EV battery clusters.

The translation to CRE demand

Per theJLL Shifting Trade Policiesguide, manufacturing accounted for nearly 20% of JLL's industrial leasing requirements in 2024 and is expected to grow to as much as 30% by 2028. JLL's manufacturing-related engagements are up more than 350% since 2018. CBRE's 2026 Outlook expects build-to-suit development to increase to meet more specialized occupier requirements, with speculative development minimal in 2026. ThePrologis-GIC $1.6B US build-to-suit logistics JV(initial portfolio of about 4.1M SF) shows institutional appetite for build-to-suit industrial; it is a logistics program, not a manufacturing-specific one.

The Honest Tempering of the Narrative

The institutional read is that announced volume is a poor leading indicator for breaking-ground volume. Discount accordingly. The 2024–2026 dataset on the delivery side is unambiguous - and citing it is what separates a serious manufacturing CRE underwriting note from a build-to-suit pitch deck.

Manufacturing construction spending has decelerated

Per IoT Analytics analyzing Census Bureau Construction Spending data, US manufacturing construction spending peaked at $239B in June 2024 and has since declined 21%. The pullback is concentrated: computer, electronic and electrical construction, which includes semiconductor fabs, is down 44% from its July 2024 peak. Excluding that category, manufacturing construction rose 5.6% in nominal terms (about 2.3% in real terms) between February 2025 and March 2026, well below what the Reshoring Initiative announcement-volume figure suggests. The Census-reported construction-spending trajectory is the leading indicator the institutional underwriter should be tracking, not the announced-jobs figure.

The specific delays are material

Three project-level data points define the announcement-vs-delivery gap. PerCNBCand Manufacturing Dive, Intel's $28B Ohio One project, slated to begin operations in 2026, is, per Intel's early-2025 update, now expected in 2030–2031; Manufacturing Dive attributes the repeated delays in part to changing corporate strategy, market dynamics and limited capital.

Per SiliconANGLE, Samsung's Taylor, Texas fab was more than 90% complete as of March 2025, but Samsung was "in no hurry" to move chipmaking equipment in. Per EVXL, Samsung scheduled its Taylor equipment installation ceremony for April 24, 2026, andTrendForce, citing Munhwa Ilbo, says trial operations could begin as early as late September or early October 2026, with mass production in 2027. PerStellantis' February 2026 6-K, the company took €22.2B in H2 2025 reset charges and cancelled the Ram 1500 BEV; in the same filing, LG Energy Solutionagreed to acquire full ownership of NextStar Energy, with Stellantis selling its 49% equity stake. The EV-cycle reset has cascaded through the broader EV battery plant CRE narrative. Ford and GM have publicly trimmed their EV battery footprints in parallel.

The bright spots

The data is not uniformly bearish. TSMC Arizona is delivering. Phase 1 has been in volume production since late 2024; phase 2 (3nm) equipment install is scheduled Q3 2026 with production 2027, ahead of the original schedule. The Phoenix cluster is the clearest proof-point that semiconductor reshoring can deliver. The Hyundai-LG Bryan County, GA Metaplant began EV production in 2024. Eli Lilly's advanced-manufacturing buildout in Indiana, North Carolina, and elsewhere is proceeding on schedule.

The honest forward call

Per anApril 6, 2026 Reshoring Initiative release, founder Harry Moser put 2025 announcements at 244,000 jobs, matching the 2024 total. The early-2025 projection of a steep drop to 174,000 did not materialize. The announcement side is not where the honest caveat lives. The reshoring tailwind is real, the announcement volume is holding, and the delivery is concentrated in fewer, slower, larger projects than the announced-volume headline suggests.

Why Manufacturing Is Not Distribution

A manufacturing facility is on the cap-rate sheet under "industrial," and the broker pitch deck treats it as interchangeable with a Class A logistics box on the same Sun Belt corridor. It is not. The institutional distinction matters because the underwriting framework, the buyer pool, the debt sizing, the residual logic, and the relationship to the macro cycle are all different. Six points carry the distinction.

1. Credit underwriting, not real-estate underwriting. A manufacturing credit-tenant lease (CTL) underwrites primarily to the tenant - corporate credit, balance-sheet capex commitment to the asset, strategic role of the facility in the network. Real-estate residual matters less because re-tenanting an advanced manufacturing facility is hard. CTL underwriting reads more like bond credit work than like rent-roll work. The buyer pool overlaps net-lease (Realty Income, which absorbed Spirit Realty in January 2024, and W. P. Carey) more than core industrial (Prologis, Brookfield, EastGroup), even though the asset is labeled industrial.

2. Lease term is long. 10–25 years on advanced manufacturing CTL versus 5–10 years on distribution. Tenant capex is large; the tenant won't sign short, and the landlord doesn't want short either. The lease structure pulls toward absolute net - tenant covers everything including structure and roof on the most credit-driven deals.

3. Tenant capex co-investment is the moat. A semiconductor or EV battery tenant invests heavily in equipment inside the box; Deal A's tenant carries about $1.5B. At the leading edge the figure is far larger (an illustrative $15–25B for a single leading-edge logic fab). That capex is, effectively, a giant security deposit that makes mid-term renewal probability much higher than a distribution tenant's. The moat is institutional - the tenant cannot move the equipment without writing off the investment.

4. Infrastructure spec is heavier and more specific. 2–5 MW utility service versus ~1 MW for a Class A distribution warehouse. 1,000+ PSF floor load versus 500 PSF. Redundant utilities, specialized HVAC (clean-room ISO classes), waste-water and chemical handling for semiconductor and pharma, heavy crane capacity for OEM auto and battery, vibration isolation for semiconductor lithography. Per ACH Engineering, clean-room build-out runs $300–600 PSF for ISO 5-6 rooms and $600–1,000+ PSF for ISO 1-4 rooms (the high-precision tier used in pharma and semiconductor manufacturing), with ISO 7-9 rooms at $100–300 PSF, against distribution's $15–25 PSF TI envelope. In Apers' read, power availability can be the binding gating constraint on site selection for large industrial users.

5. Build-to-suit-dominant. Advanced manufacturing is developed overwhelmingly to a tenant's specification, and CBRE's 2026 Outlook expects speculative development to be minimal. The Prologis-GIC $1.6B joint venture is a build-to-suit logistics program rather than a manufacturing one, but it shows institutional capital funding BTS industrial. Trammell Crow, Hines Industrial, Stonemont, and Industrial Realty Group are the active developer cohort.

6. Different geography. The manufacturing reshoring map - the Midwest belt (Louisville, Nashville, Cincinnati, Chicago, Detroit, Kansas City), Phoenix and Austin (semiconductors), Atlanta and the Carolinas (Hyundai/LG, BMW, VinFast, Toyota), Indiana (Stellantis-LG, Lilly), Tennessee (Ford BlueOval) - overlaps but does not equal the modern distribution geography. The distribution map is the inland-port + coastal-port matrix (Dallas, the Inland Empire, Atlanta, Memphis, Columbus, Indianapolis, Northeast NJ/Lehigh). Sun Belt overlap is meaningful, but the manufacturing geography is more concentrated around utility-rich and water-rich sites that the distribution map doesn't prioritize.

Flex / R&D - the Hybrid Category

Flex sits between distribution and pure office. In this article, flex means a building with at least 25% office buildout, with the remainder in light industrial, R&D, lab, or assembly use. The practical office share runs 25–50%, with the upper end approaching small-floor-plate office. Flex is multi-tenant, smaller-floorplate, higher-parking, and considerably more operationally intensive than distribution. At Table 1 midpoints the cap rate runs about 100 basis points wider than core logistics (7.0% vs 6.0%), reflecting the operational complexity and the shorter average lease term.

Flex sub-uses - the four populations

Biotech / life-sciences R&D is the highest-value flex sub-use. The institutional submarkets, including Kendall Square (Cambridge), South San Francisco, San Diego UTC, the Route 128 corridor, Princeton-NJ, and Research Triangle Park, are assumed here to have carried 4.5–5.25% cap rates at the 2021 peak and to have since reset wider with the 2024–2026 biotech demand pullback. Specialist sponsors dominate: Alexandria Real Estate Equities, IQHQ, Longfellow, BioMed Realty (Blackstone), Healthpeak's life-sciences book, Ventas's life-sciences book, Tishman Speyer's life-sciences platform. Lab re-lease TI is underwritten at $150–250 PSF in this article (an illustrative assumption).

Tech R&D is the Silicon Valley archetype. Cushman & Wakefield's Q2 2026 Silicon Valley R&D MarketBeat prints 13.3% vacancy (+30 bps QoQ from 13.0%, +60 bps YoY from 12.7%) - soft but not collapsed. Tech-R&D demand correlates to venture funding and corporate R&D capex; the AI infrastructure cycle is a tailwind on certain submarkets (the Westside of Sunnyvale-Mountain View, parts of Seattle, Austin) and a headwind on others (the deeper Valley R&D corridor that traditionally housed semiconductor R&D operations).

Light industrial / contract manufacturing flex is the workhorse - small-bay buildings serving local manufacturers, contract assemblers, and distributors with showroom requirements. The EastGroup Properties (Sun Belt) and First Industrial (national Tier-1 industrial REIT) benchmark portfolios anchor the institutional comp set. Cap rates run 6.0–7.0% (Table 1). The reshoring tailwind benefits this cohort indirectly - contract manufacturers serving the Tier-1 OEMs cluster around the advanced-manufacturing anchors, and the demand for 50K–150K SF shallow-bay multi-tenant product has held up better than the demand for the largest distribution boxes.

Specialty flex - data center adjacent, life-sciences manufacturing, advanced energyis the emergent fourth population. Edge data-center and AI-inference colocation in 50K–200K SF flex bays. Life-sciences manufacturing - the CDMO and contract-pharma sites that sit between R&D and full GMP commercial production - in 80K–300K SF flex with selective clean-room build-out. Advanced-energy R&D and pilot production for battery, hydrogen, and grid-scale energy storage tenants. Each is a discrete underwriting case; the common thread is high TI, longer-than-typical-flex lease terms, and a buyer pool that pulls toward specialist sponsors.

Flex operational reality

This article's flex underwriting assumes buildings of 50K–300K SF total with 5K–45K SF suites; these operating figures are illustrative assumptions. The leasing rhythm is different from distribution: assume 6–12 months of vacancy between tenants versus 3–6 months for distribution. Parking ratios of 2.5–4.5 spaces per 1,000 SF run closer to office than to distribution's 1.0–2.0. TI burn on rollover is the dominant operating expense, and for biotech R&D flex, TI on re-lease is assumed at $150–250 PSF on lab fit-out before any tenant-specific specification. The institutional flex underwriter spends most diligence time on the rent roll, the rollover schedule, the submarket comp set, and the TI burn forecast - the same operational discipline a multi-tenant office underwriter applies, adjusted for the industrial-spec reality.

The Biotech R&D Flex Reset

The fresh wedge on the flex side is the 2024–2026 biotech reset. The institutional biotech submarkets expanded aggressively from 2018 through 2021 against a venture and IPO-funded biotech wave that pulled forward years of demand. The 2022–2024 funding contraction and the post-2024 demand pullback have left the lab market structurally oversupplied in the secondary submarkets and softer than expected even in the prime. The Q4 2025 / Q1 2026 vacancy prints are the cleanest measurable proof.

The Q4 2025 / Q1 2026 vacancy reality

Per Lincoln Property Company's Q1 2026 Lab Market Report (LPC), lab vacancy in Q1 2026 was 26.7% in Boston, 16.0% in Cambridge, 20.5% in the suburbs, and 19.6% metro-wide. Boston's rate declined 50 basis points, the first quarter-over-quarter decrease in over a year, but remains elevated. LPC puts East Cambridge at 10.4%, whileBisnowreports Cushman & Wakefield putting East Cambridge, with more than 11M SF of lab inventory, above 20% vacancy; the two readings come from different broker datasets, so compare each figure only with its own source. Per theBoston Globe / Boston.comcoverage from March 2025, the Massachusetts biotech slump has produced unleased lab space, plunging valuations, and layoffs across the sector.

The Takeda sublease and the anchor-recommit counter-story

Per LPC's Q1 2026 report, Takeda placed approximately 630,000 SF of space in the mid-Cambridge market, which pushed Cambridge sublease availability to 11%, the highest level recorded to date in the Cambridge lab market. That single placement is about 3.1% of Cambridge's 20.4M SF of lab inventory (630,000 / 20,398,146). The bifurcation cuts the other way too: Biogen is building a 585,000 SF global headquarters at 75 Broadway in East Cambridge, due in 2028, and Bisnow reports a trio of AI-company leases (mostly office) in Kendall Square in September and October 2026. The institutional read is that Kendall is bifurcating between prime space with new demand and softer secondary space.

The basis reset and the long-term thesis

This article's Deal B assumes a basis reset: cap rates moving from a 4.5–5.25% prime band at the 2021 peak to a 6.5–7.5% range today (both illustrative assumptions, not survey data). Capital markets are thin. LPC reports no Q1 2026 lab transactions across the region and only a handful of trades in late 2025, among them Phase 3 Real Estate Partners' $25.1 million purchase of 30 Hampshire Street in East Cambridge. The institutional read: the long-term thesis on Kendall is intact (it remains a deep biotech submarket), and the 2024–2026 reset has reset basis, not fundamentally re-priced the long-term sector tailwind. The bet on a Kendall biotech R&D flex acquisition in 2026 is on re-leasing into the recovery at the reset basis.

The secondary submarkets (suburban Boston, Research Triangle Park, suburban San Diego, NJ) have a longer absorption cycle ahead. LPC's Q1 2026 data puts suburban Boston lab vacancy at 20.5% and availability at 32.2%.

Capital Stack Reality

The capital stack on manufacturing CTL is structurally distinct from the capital stack on flex / R&D. Both are distinct from core logistics. The institutional underwriter sizing debt or screening equity requirements on a manufacturing or flex deal has to read the stack against the right comp set, not the industrial-default comp set. The debt terms and leverage ranges below are Apers illustrative assumptions.

Manufacturing CTL acquisition

The manufacturing CTL acquisition competes against the net-lease bid more than the core-industrial bid. The expected senior debt is balance-sheet life-co or insurance paper at 55–65% LTV, 10–25 year fixed rate matched to the lease term - structured like a CTL bond. Equity capital sources: net-lease REITs (Realty Income, which absorbed Spirit Realty in January 2024, and W. P. Carey), private net-lease funds (LCN Capital Partners, Oak Street / Blue Owl, Brookfield's net-lease platform), private REITs in the BREIT / BPP cohort, and the institutional sale-leaseback market. Sale-leaseback is a seller-side route for manufacturers raising balance-sheet capital to fund reshoring capex; the manufacturing CTL bidder pool actively works that pipeline.

Manufacturing build-to-suit development

BTS manufacturing development is overwhelmingly forward-leased - the development risk transfer to the tenant is what makes the project bankable. The Prologis-GIC $1.6B logistics JV is an example of institutional capital funding build-to-suit. Debt is construction-to-perm balance-sheet at 60–65% LTC with a life-co lead, with takeout sized to the lease term. Equity from open-ended core funds and sovereign capital - GIC, ADIA, Norges, the Canadian pension cohort - alongside the institutional industrial REIT balance sheet. The Trammell Crow, Hines Industrial, Stonemont, and Realterm developer cohort is active.

Flex / R&D multi-tenant acquisition

The flex / R&D multi-tenant acquisition sits in the value-add industrial / industrial-flex bid. Debt is from regional banks and debt funds at 60–70% LTV, 5–7 year terms with extension options. Equity from value-add funds, private REITs in the BREIT-style cohort, and family-office capital. Stonemont's IOS specialization, Realterm's transportation-adjacent specialization, and STAG Industrial's smaller-tenant industrial book are reference platforms for the institutional comp set.

Biotech R&D flex specialist stack

Biotech R&D flex has its own specialized capital stack. Senior debt from life companies and the large money-center banks active in life sciences (JPMorgan, Wells Fargo, Bank of America, Goldman). Equity from life-sciences specialist sponsors - Alexandria Real Estate Equities, IQHQ, Longfellow, BioMed Realty (Blackstone), Healthpeak's life-sciences book, Ventas's life-sciences book, the Tishman Speyer / Breakthrough Properties JV. The 2024–2026 reset has hit value, and LPC reports only a handful of lab trades across the region in late 2025 and none in Q1 2026.

Deal A - Phoenix Semiconductor CTL

The first worked example. An advanced-manufacturing CTL acquisition in the North Phoenix TSMC cluster. The asset profile follows the institutional spec for a Tier-1 semiconductor equipment supplier facility: 240,000 SF on 40 acres, 2024 vintage, 50 ft clear height, 1,200 PSF floor load, 4 MW utility service with redundant feeds, eight dock doors plus two drive-ins, 15% office buildout. The tenant is a Tier-1 semiconductor equipment manufacturer at BBB+ corporate credit on a 15-year absolute NNN lease that commenced in 2024 (13 years remaining at a 2026 sale, expiring 2039), with two 5-year FMV options and 2.0% annual escalators. Current in-place base rent is $18.00 PSF NNN, which on 240,000 SF gives in-place NOI of $4.32M (absolute net, so NOI equals rent). Tenant equipment investment in the facility is approximately $1.5B; landlord TI is zero because the tenant delivered the clean-room build-out as tenant capex. The deal is on the market at a 6.10% cap rate, implying a $70.8M purchase price ($295 PSF).

Deal A worked example: illustrative, not deal-specific. The Phoenix advanced-manufacturing CTL profile underwrites as net-lease credit. At Table 1 midpoints the 6.10% cap sits about 40 bps tighter than light industrial (6.50%) and about 90 bps tighter than flex/R&D (7.00%) because the tenant credit and the equipment moat carry the deal. Debt cost (6.25%) exceeds the 6.10% going-in cap rate, so the loan is negatively levered at purchase.

Deal A IRR assumptions (illustrative): 10-year hold, NOI of $4.32M growing 2.0% a year, 60% LTV at 6.25% with 25-year amortization, 1% sale cost; exit price is year-11 NOI divided by the exit cap rate. An exit cap rate of 6.10% gives a 10.0% levered IRR (8.0% unlevered); a 6.75% exit cap rate gives 8.5%. The levered IRR exceeds the unlevered IRR despite the negative leverage at purchase because 2.0% annual NOI growth lifts the unlevered return above the 6.25% cost of debt over the hold. With about 3 years of lease term left at exit, a buyer may price at the wider end of that range.

The institutional read on Deal A is that this is a credit underwrite, not a real-estate underwrite. The diligence workload is dominated by the tenant credit memo, the lease document, and the strategic-role-of- facility analysis. Replacement-cost residual is a secondary check; the primary case for the cap rate is the BBB+ tenant credit, the 13-year remaining term, and the strategic role of the facility within the tenant's Phoenix-cluster footprint. The renewal probability at the 2039 expiry is high because the $1.5B of tenant equipment capex is not portable - the tenant cannot relocate without writing off the equipment investment. The deal trades into the net-lease bidder pool: Realty Income, W. P. Carey, the private net-lease funds, and the industrial CTL sleeves at the large sovereign capital pools (GIC, ADIA, Norges). The traditional industrial REIT bid (Prologis, Brookfield, EastGroup) is structurally less competitive on this deal because the asset profile pulls toward the net-lease comp set and away from the multi-tenant industrial bid.

Deal B - Cambridge Biotech R&D Flex

The second worked example. A Cambridge Kendall Square biotech R&D flex acquisition at the post-2024 reset basis. 180,000 SF urban-infill flex / R&D building, 1.2-acre site, 18 ft clear height (typical for lab/R&D flex), 35% office buildout, 65% lab build-out, 2018 vintage with a 2023 partial renovation. Six tenants, all biotech R&D and contract-pharma R&D, in 12K–45K SF suites on a 3-year weighted-average remaining term (5-year average original terms) with 3% annual escalators. In-place rent averages $52 PSF gross; stabilized economic occupancy 88%, reflecting the post-2024 Cambridge softening. NOI is approximately $5.4M: 180,000 SF x $52 = $9.36M of potential gross rent, less 12% vacancy ($1.12M) leaves $8.24M, less about $2.84M of operating expenses ($15.76 PSF) leaves $5.4M. The deal is on the market at a 6.85% cap rate, implying a $78.8M price ($438 PSF). That cap rate sits inside the 6.5–7.5% range this article assumes for flex/R&D today, against an assumed 4.50–5.25% at the 2021 Kendall peak. Re-leasing capital is the swing factor: TI of $150–$250 PSF across all 180,000 SF would be $27.0M to $45.0M, or 34% to 57% of the price.

Deal B worked example: illustrative, not deal-specific. The Cambridge biotech R&D flex profile underwrites as multi-tenant real estate. The cap rate sits 75 bps wider than Deal A's CTL print because the tenant credit is diffuse, the rollover is operational, and the TI burn on re-lease is meaningful. The bet is on re-leasing into the recovery at a reset basis.

Deal B IRR assumptions (illustrative): 5-year hold, 3% annual growth in rent and operating expenses, 60% LTV at 6.75% with 22-year amortization, 1% sale cost, NOI bridge as above. Lease-up case: occupancy rises from 88% to 95% by year 3, with $200 PSF TI on the 12,600 newly leased SF ($2.52M) and exit at a 6.85% cap rate, giving a 17.4% levered IRR. Downside case: occupancy falls from 88% to 78% by year 3, no new TI; exit at a 6.85% cap rate gives 3.4% and exit at a 7.5% cap rate gives -0.7%. The 17.4%, 3.4% and -0.7% figures are before rollover TI on existing tenants, which matters because the average remaining term is 3 years inside a 5-year hold. If half the occupied area (79,200 SF) re-leases in year 3 with $150 to $250 PSF of landlord TI ($11.9M to $19.8M), the lease-up case levered IRR falls to 11.6% to 7.8%.

The institutional read on Deal B is that this is a real-estate underwrite, not a credit underwrite. The diligence workload is dominated by the submarket comp set, the rent roll, the rollover schedule, the re-leasing velocity analysis (months from move-out to lease commencement in the comp set), and the TI burn forecast on rollover. Tenant credit is a secondary concern because the multi-tenant structure spreads the risk across six counterparties - biotech tenants are often pre-revenue, but no single tenant carries more than an assumed 25–30% of the rent. The deal trades into the life-sciences specialist bidder pool: Alexandria, IQHQ, Longfellow, BioMed Realty (Blackstone), Healthpeak's life-sciences book, and the value-add industrial-flex funds. The traditional industrial REIT bid is structurally less competitive on this deal because the asset profile pulls toward the life-sciences-specialist comp set and away from the multi-tenant flex bid that doesn't underwrite lab fit-out.

The Two-Deal Payoff

Two deals labeled "industrial." Neither underwrites like the other. The side-by-side comparison is the article's payoff - it makes the institutional distinction visceral in a way that no taxonomy table can.

The two-deal payoff. Same industrial label, opposite underwriting frameworks, 75 bps of cap rate between them, buyer pools that barely overlap. The 2024–2026 macro supports the manufacturing CTL side (with the IoT Analytics / Census temper); the 2024–2026 biotech reset shapes the flex/R&D basis on the other. IRR assumptions are stated under each deal table.

The takeaway is operational, not theoretical. The 2026 acquisitions team scanning a fresh OM on a manufacturing or flex deal has to make the same first call the medical-office team makes when reading the rent roll: is this a credit instrument or an operational asset? If the lease document is the artifact you spend two hours reading, it is a credit underwrite. If the rent roll is the artifact, it is a real-estate underwrite. The cap rate, the IRR, the buyer pool, the debt structure, the diligence workload, and the institutional comp set all fall out of that first call. The parallel article on the MOB side (single-tenant credit vs multi-tenant medical office) walks the same bifurcation in a different sector.

Six Mistakes Practitioners Make

• Anchoring on the announced-volume figure without the delivery-side temper. The Reshoring Initiative 244K-jobs-in-2024 figure and the SIA $828B+ semiconductor announcement number are real, and they read like a one-way tailwind for manufacturing CRE. The IoT Analytics / Census construction-spending data, the Intel Ohio slippage, the Samsung Taylor equipment delay, and the Stellantis €22.2B reset are the necessary other half. Announcements held roughly flat at 244K in 2025; the discount belongs on announcement-to-completion conversion, not on the headline count. Underwrite the announcement side at a discount: as an illustrative assumption, haircut multi-year manufacturing announcements to a 50–70% delivery probability when sizing future absorption.

• Using core-industrial cap rates on manufacturing CTL. The core-logistics Class A cap rate (5.5–6.5%) is the wrong comp set for a Tier-1-credit advanced-manufacturing CTL; the net-lease bid (5.75–6.75% on BBB+ credit with long WALT, per Table 1) is the right comp set. Using the low end of core-logistics cap rates overvalues the asset: at 5.5%, Deal A's $4.32M NOI is worth $78.5M, $7.7M above its $70.8M price at 6.10%, and at the 6.75% top of the net-lease range it is worth $64.0M.Bridge Investment Group's net lease team, for example, describes a credit-driven approach to mission-critical industrial assets leased to high-quality credit tenants; bring that comp set into the pricing, not the core-industrial one.

• Underestimating the TI / clean-room build-out cost on advanced manufacturing.$300–1,000+ PSF on advanced clean rooms (ACH Engineering: $300–600 for ISO 5-6, $600–1,000+ for ISO 1-4). Underwriters who carry the $15–25 PSF distribution TI figure into a manufacturing model are using the wrong number. On BTS deals the tenant typically carries the equipment, but landlord shell costs scale with the power, floor-load, vibration, and chemical-handling spec - not the four-walls cost. Size the shell against the spec, not the default.

• Treating biotech R&D flex as a uniform sector. The Q1 2026 dataset bifurcates between prime submarkets with new demand (Kendall) and softening secondary submarkets (suburban Boston, Research Triangle Park, suburban San Diego, NJ). A Kendall acquisition at the reset basis has a meaningfully different lease-up curve than a suburban Boston acquisition at a wider cap. Don't average the sector; read the submarket.

• Underweighting power as a site-selection constraint. For larger industrial users, sufficient power access can be the gating item in many markets. For advanced manufacturing requiring 2–5 MW (or far more for a leading-edge semiconductor fab), power availability can rank ahead of land cost and incentive depth as the threshold criterion (Apers' view). Underwriting a 2026 manufacturing acquisition without explicit power-interconnection diligence is a meaningful institutional risk.

• Mis-sizing landlord TI burn on flex / R&D rollover. Multi-tenant flex carries meaningful TI burn on rollover: $80–150 PSF on non-lab flex, $150–250 PSF on lab re-lease. The 2018 flex underwrite assumed lower TI; the 2026 underwriting environment requires the full lab fit-out range on biotech R&D rollover, particularly for buildings inheriting non-fungible tenant-specific lab configurations. The TI burn is the dominant operating expense on the model and the most-often underestimated.

From Manufacturing/Flex to Apers

The 2026 acquisitions team pulling an OM on a manufacturing CTL or a flex/R&D portfolio runs both deal types through the same institutional model.

AQ-401 Industrial Warehouse / Distribution Center Model underwrites industrial acquisitions with a 10-year cash flow projection: acquisition basis, BTS construction cost or shell capex, lease-up curve with concession burn for multi-tenant deals, TI burn on lab or specialty fit-out, OpEx with utility and tax detail, NOI bridge, cap-rate sensitivity, and debt sizing. The Deal A and Deal B worked examples in this article both flow cleanly through AQ-401 - the CTL underwrite by overlaying the net-lease module on the AQ-401 baseline; the flex/R&D underwrite by overlaying the multi-tenant lease-rollover module. The same model framework, two operational specializations.

For manufacturing CTL deals the spreadsheet work concentrates on tenant credit, lease-term cash-flow projection, lease-expiry renewal sensitivity, residual-value sanity-check, and power-interconnection / environmental risk overlays. For flex / R&D deals the work concentrates on rent-roll diligence, WALT (unit-weighted and space-weighted), rollover schedule, lease-up velocity against the submarket comp set, TI burn forecast on rollover, and theIRR sensitivity to lease-up timing. Both benefit from the AQ-401 structural baseline.

Related Articles

• Industrial Underwriting: Clear Heights, Dock Doors, and the Modern Distribution Spec- the distribution-side baseline that this article's spectrum table reads against; the structural comp on the real-estate-underwriting end of the spectrum.

• Logistics & Distribution: E-commerce Sensitivity and the Last-Mile Spec- the cluster pillar covering the distribution and last-mile populations that sit at the real-estate-underwriting end of the industrial continuum.

• Cold Storage: Specialized Modeling, Tenant Credit, and the Modern Refrigerated Spec- the specialty-industrial cousin to flex/R&D, with its own tenant-credit and TI-burn operational discipline.

• Data Centers: Power Pricing, PUE, and the Hyperscaler Build-to-Suit- the most power-intensive industrial-adjacent typology, where the 2–5 MW manufacturing threshold scales to 50–500+ MW and the underwriting overlaps both CTL and operational-flex frameworks.

• Trophy vs Class A vs Class B Office: The 2026 Stratification- the parallel bifurcation discipline on the office side. Same institutional read, different sector: "office" is no longer a single category; neither is "industrial."

• Medical Office: Single-Tenant Credit vs Multi-Tenant WALT- the credit-vs-operational bifurcation walked on a different asset class. Direct analog to the manufacturing CTL vs flex/R&D distinction in this article.

• IRR Sensitivity Analysis and Stress Testing- the returns machinery that converts stabilized NOI into a levered IRR and stress-tests it against the lease-up timing, cap-rate exit, and rollover-velocity assumptions both deals in this article depend on.