A logistics real estate portfolio is a pool of income-producing industrial properties: distribution, fulfillment, last-mile, cold storage, and light industrial, held and managed as one investment. Underwriting is the discipline of turning buildings and leases into forward cash flows, then testing whether those cash flows can support the purchase price and the debt when conditions get less friendly.
This guide explains how to underwrite a logistics real estate portfolio in a way that avoids “pretty IRRs” and instead surfaces the real drivers of cash, covenants, and exit value. The payoff is simple: you will know what you own, what can break, and which assumptions actually deserve debate.
Understand what you are (and are not) buying
This isn’t underwriting an operating company, even when tenants run logistics networks. You are not buying the tenant’s margin. You are buying contracted rent plus an option on future leasing spreads, minus capex, downtime, and incentives, all inside a legal and tax wrapper that decides who controls cash and what happens in a workout.
The central question is simple: does NOI durability and the economics of re-leasing justify the entry basis and leverage, given uncertain market rent and exit liquidity? If you can’t answer that in plain English, the spreadsheet is doing the thinking for you.
Define “logistics” early so your comps stay honest
Separate property types because risk is not uniform
“Logistics” usually means bulk warehouses, cross-docks, and last-mile facilities with real throughput and access to population centers. Many portfolios also include cold storage, truck terminals, container depots, outdoor storage, and light manufacturing. You should separate those buckets on day one because rent drivers, capex, and tenant behavior differ.
The boundary matters because cap rates and lender terms often assume a plain warehouse box. Cold storage can demand heavy systems capex, larger TIs (tenant improvements), longer downtime, and a narrower user set. Last-mile can post high rents, but it can also age fast if truck courts, parking, power, or delivery layouts can’t keep up.
If someone calls a portfolio “logistics” while it carries meaningful specialized or quasi-manufacturing exposure, you lose clean comparables. That mispricing shows up later as “surprises,” which are usually just mislabeled risk.
Map the incentives of sponsors, lenders, and tenants
Incentives also diverge. Sponsors often earn fees tied to gross asset value and upside outcomes, which can tilt assumptions toward higher leverage and faster rent marks. Lenders care about debt service and collateral liquidity, so they punish short WALT (weighted average lease term), tenant concentration, and near-term capex cliffs. Tenants optimize network cost and service levels, which can make occupancy decisions lumpy and correlated across markets.
Build a minimum underwriting architecture that actually works
A portfolio model has to bridge bottoms-up property detail with portfolio-level liquidity and debt mechanics. If you skip that bridge, you’ll get a clean IRR and an ugly cash event.
Start with a property-level lease schedule that is actually usable. For each unit or building, model base rent, recoveries, abatements, free rent, step-ups, expense stops, options, and who pays taxes, insurance, and CAM (common area maintenance). Use net effective rent, not headline rent. Teams lose money by modeling the contract and ignoring the concessions.
Then build a re-leasing engine by submarket and product type. Feed it market rent, vacancy, downtime, TI, leasing commissions, renewal probability, and any local friction like permitting or utility lead times. This engine should convert expirations into stabilized cash flows with explicit timing.
Separate cash burn during downtime from capitalized leasing costs. Debt covenants and distribution capacity often hinge on timing, not annual NOI. A deal can look fine on paper and still trap cash when leasing costs hit before rent starts.
Finally, model the capital stack and the cash cascade in monthly or quarterly detail. Run collections, expenses, reserves, debt service, sweeps, capex, and distributions through the loan documents and the JV waterfall. Portfolio underwriting fails most often at this level, not at the market rent input.
Translate the return equation into underwriteable numbers
Equity returns in logistics real estate come down to three variables: entry basis, the NOI path through rollover and capex, and exit pricing. Rent growth matters only if it survives incentives, downtime, and capex. Cap rate compression is not a plan; it’s a bet, and you should price it like one.
Treat going-in yield and stabilized yield separately. Going-in cap rate uses in-place NOI and purchase price. Stabilized yield uses forward NOI after rent marks, rollover, and planned capex, divided by purchase price plus capex. When the stabilized yield depends on heroic releasing spreads, the thesis is leasing execution, not core logistics.
On exit, apply a terminal cap rate to year-N NOI that reflects normalized vacancy and a reserve for capital. A disciplined underwrite uses a terminal cap rate at least as conservative as entry, unless you can show why the assets will be more liquid and lower risk at exit. Portfolios can sometimes get more liquid if you sell in smaller lots, but that introduces time, transaction costs, and execution risk. Put numbers on those frictions.
Focus on assumptions that actually move outcomes
Market rent and rent growth: build a path, not a slogan
Market rent is the anchor for releasing spreads. Source it at the submarket and product level, not the metro average, and reconcile at least two data sources. If the portfolio spans countries, don’t copy U.S.-style annual escalators into markets where the norm is different.
Express growth as a path, not a single CAGR. Near-term growth should reflect forward supply and absorption; longer-term growth should revert toward something closer to inflation for mature markets. The IMF forecasted global inflation at 5.8% in 2024 and 4.4% in 2025 (Apr-2024 WEO). That doesn’t set rents, but it does bound long-run nominal assumptions. Perpetual high rent growth is a quiet way to force a good answer.
Vacancy and downtime: underwrite time-to-cash
Model downtime in months from expiration to rent commencement, including build-out, permitting, and inspections. Specialized assets, very large boxes, and buildings with weak truck courts, power, or fire life safety take longer. The cost is not just lost rent; it’s operating carry, leasing costs, and covenant pressure.
In downside cases, assume synchronized rollover for concentrated tenant exposure. If the portfolio leans on e-commerce, consumer discretionary, or one large 3PL, assume correlated non-renewals and longer downtime. Correlation is what makes portfolios different from single assets, and it’s what makes leverage dangerous.
Renewal probability and options: treat options as constraints
Make renewal probability a function of rent versus market, building fit, relocation cost, and landlord capex obligations. Tenants renew when the site is embedded in the network and the rent is tolerable relative to alternatives. They also renew when the landlord writes checks for refresh capex.
Treat options as real constraints, not footnotes. Fixed-strike or CPI-linked options can cap upside when in-place rent is below market. Options can also create decision drift, where tenants wait while negotiating, pushing your downtime into the future at the worst time for covenants. Model options as decision trees with timing, not as automatic renewals.
Expense pass-through and leakage: “NNN” still leaks
Leases run from NNN to modified gross, and even NNN leaks. Leakage shows up in non-recoverables, caps on controllables, exclusions, and timing mismatches. Verify what the landlord actually billed and collected, and test historical recovery ratios.
Insurance and taxes can move fast, and pass-through depends on language and local law. Underwrite gross-to-net drift and stress it. A modest change in expense ratio can overwhelm modest rent growth in a mature portfolio. The impact is immediate: less cash to fund leasing and reserves, and weaker DSCR optics.
Capex: keep it granular, then schedule it
Split capex into four buckets, each with different risk and timing.
- Maintenance capex: Roofs, paving, dock equipment, fire systems, and lighting, modeled as recurring reserves per square foot plus periodic lumps.
- Turnover capex: TI, make-ready, racking removal, slab repairs, and power upgrades, which can be light for some boxes but heavy for others.
- Value-add capex: Expansions, trailer parking, additional docks, solar, and EV charging, each tied to a rent premium or absorption improvement with a permit-ready schedule.
- Compliance capex: Code upgrades triggered by tenant work, environmental items, and resiliency projects that can become mandatory at rollover.
Timing is the point. Portfolios can look capex-light on average and still face a wall in years two to four, the same window when refinance risk peaks. Liquidity, not NOI, is what breaks deals.
Credit loss and default mechanics: don’t assume “re-tenantable” means fast
Industrial underwriting often assumes minimal credit loss because leases are shorter and space is re-tenantable. That can be true for generic boxes in strong nodes. It fails for specialized improvements, weak access, or thin leasing markets, where a warehouse starts behaving like a single-tenant asset.
Model explicit credit loss for non-investment-grade tenants, especially with longer leases and uncertain backfill. Also model the jurisdictional timeline to regain possession. Legal remedies can be slow, and bankruptcy can delay recoveries. The impact is direct: longer cash drag and higher leasing costs, often during a covenant-sensitive period.
Exit liquidity and lot-size strategy: underwrite who the buyer is
Large portfolios can shrink the bidder set and invite a complexity discount. A break-up sale can improve pricing but adds disposition costs, time, and execution risk. If your upside depends on a portfolio premium or break-up value, underwrite both paths and show the time-to-close and cost.
Let debt mechanics, not hope, decide the equity story
People talk about cap rates and rent growth, but realized equity IRR is often decided by the loan agreement. Underwrite the debt as a living set of triggers.
Model interest type, hedging, amortization, maturity, extension conditions, LTV or debt yield covenants, DSCR tests, reserves, and cash management. Many loans shift from springing cash management to hard cash dominion when tests fail. That traps cash and stops distributions even if the assets remain valuable.
Rates matter because they change refi proceeds and covenant headroom. The Federal Reserve held the federal funds target range at 5.25%-5.50% as of Jan-2024 (FOMC statement). Use that reality to anchor floating-rate exposure and the cost of caps.
Run three interest cases: base, forward curve, and a stressed path where rates are higher at refinance. Then size refinance proceeds using a stressed exit cap rate and lender constraints, not a default 65% LTV takeout. The impact is close certainty: a good deal that can’t refinance is not a good deal.
Get governance and ring-fencing right before it matters
Most portfolios sit in SPVs with separateness covenants and limits on additional debt and asset sales. Those clauses matter when things get tight, because commingled cash and unclear priorities invite creditor disputes and slow decisions.
Cross-collateralization can improve pricing because the lender underwrites the pool. It also reduces release flexibility and can force you to carry weak assets longer than you’d like. Property-level debt improves sale optionality but adds operational weight and can reduce proceeds through smaller loan sizes.
If the portfolio crosses jurisdictions, assume different landlord-tenant rules, transfer taxes, and enforcement processes. In parts of Europe, financing often relies on share pledges and account pledges, and lenders care deeply about enforcement venues and perfection. Enforcement friction changes lender behavior, and lender behavior changes your refinance options. That flows straight into returns.
Model the flow of funds because timing creates winners and losers
The portfolio waterfall is the operational truth. Document it and model it.
Rent is collected into property accounts or a centralized account. Cash management can sweep funds into lender-controlled accounts if covenants trip. Operating expenses get paid, reserves get funded, debt service gets paid, and only then does cash reach equity, subject to lockboxes and distribution tests.
If the sponsor earns property management, leasing, or construction management fees, those often come out before equity distributions. Model fee timing and whether fees subordinate when covenants fail. The impact is straightforward: less liquidity when you need it most.
Also model lender consents for major leases, capex, and asset sales. Delays in consent can extend downtime and raise tenant costs. That shows up as real money and missed quarters, not process.
Diligence: where risk hides in plain sight
The documents that drive cash flows and control rights are unglamorous, and they matter.
The PSA sets price adjustments, proration mechanics, and capex credits. Title and surveys expose access, easements, and encroachments. Leases and abstracts carry recoveries, options, and landlord work. Management and brokerage agreements decide fees, termination rights, exclusivity, and tail periods. Construction contracts decide price certainty and change order risk. Loan and cash management agreements decide triggers, reserves, and remedies. Intercreditors decide standstills and cures. JV documents decide consents, removals, and promote math.
Execution order matters. Lenders often demand finalized leases and estoppels before funding, while sellers resist early tenant outreach. Put that gating risk into the timeline and the close plan, especially in multi-tenant portfolios.
Reps and warranties in real estate are limited. You manage risk with diligence, insurance, escrow, and price. If environmental risk is material, confirm whether insurance is available and whether it covers known conditions.
Stress tests that catch weak deals early
A logistics portfolio shouldn’t require heroics to clear hurdles. Run kill tests early, before heavy diligence spend.
Stress exit cap rate expansion, lower releasing spreads, slower growth, longer downtime, higher TI and leasing commissions, and a refi with higher rates and tighter proceeds. Add a case where one or two major tenants don’t renew in the same year. Pull forward the capex wall on roofs and paving. Then watch DSCR triggers, reserve draws, and cash sweeps. Liquidity events force distressed sales; low IRR does not.
A fresh angle: treat data quality as a risk factor you can price
Portfolio underwriting increasingly fails for a modern reason: the data package is messy, and the model pretends it is clean. A simple, high-value upgrade is to add a “data confidence overlay” to the underwriting memo and the IC discussion.
- Lease data integrity: Track which leases are fully abstracted, which are summary-only, and where recoveries or options are ambiguous.
- Billing evidence: Reconcile recoveries to actual billings and cash receipts, not just lease language.
- Capex evidence: Tie major line items to proposals, unit costs, and site walks rather than blended dollars per square foot.
- Timing evidence: Validate downtime assumptions with broker feedback and permit lead times, not “market standard” guesses.
When confidence is low, you can price it by widening stress ranges, adding reserves, tightening leverage, or demanding seller credits. That is not conservatism for its own sake; it is converting uncertainty into an explicit underwriting input.
Conclusion
Good logistics portfolio underwriting is the practice of turning leases, capex, debt terms, and governance into a timed cash story that still works when the cycle turns. If you define the property boundary, build a re-leasing engine, model the waterfall, and stress liquidity and covenants, you stop debating wishful rent growth and start underwriting what will actually happen.