Commercial Real Estate Analysis Investments

What Actually Matters When You Analyze a Deal

I've spent more time than I want to admit building pro formas for industrial buildings, medical offices, and small multifamily properties. Most people coming into this think the analysis is about the spreadsheet. It's not. The spreadsheet is just where you end up after you've already made all your real judgments about the asset. The actual work happens before you open Excel. You need to understand lease structures first. Triple net, gross, modified gross, percentage leases — each one shifts where the risk sits. If you don't know who's paying for the roof replacement when it fails in year four, your cash flow projection is wrong regardless of how clean your numbers are. I've seen people underwrite a full NNNA grocery-anchored strip center as if they were responsible for structural repairs. They weren't. The landlord had capped operating expense pass-throughs. That single term changed the entire risk profile.

Commercial Real Estate Analysis Investments

At its core, commercial real estate analysis investments is the process of estimating what a property will actually produce in cash, adjusting for every real-world leak — vacancies, capital expenditures, tenant turn, lease expirations — and then comparing that result against your required return. It sounds simple. The complications come from how many assumptions stack on top of each other. Here's how I approach it, not as a rigid framework but as a set of habits that keep me from making expensive mistakes.

The Workflow I Actually Use

I start with the lease rent roll. Not a summary. The actual rent roll with term dates, escalation clauses, renewal options, use clauses, exclusivity provisions, and TI allowance schedules. Most deals never show you this upfront. You have to ask for it during due diligence and then read it line by line. Next I build a vacancy buffer that's honest about the market, not optimistic about the seller's marketing claims. A Class B office space in a secondary market with a 12% vacancy rate is not going to lease up to 5% just because the pro forma says so. I use the trailing twelve months of actual market absorption data for that submarket, adjusted for any new supply coming online. If there's a 200,000 square foot warehouse completing three blocks away next year, your asking rent of $18 per square foot needs a reality check. Then I layer in capital expenditure reserves. This is where most underwriters get sloppy. They throw a flat number at it — $2 per square foot annually, maybe — without thinking about what's actually aging out. I track roof life, HVAC replacement cycles, parking lot resurfacing, and tenant improvement allowances separately. A building with twenty-five percent of its square footage leased to tenants on month-to-month arrangements needs a higher TI reserve than one with a credit tenant signing a ten-year deal. The math is different even if the rent is the same.

After expenses, I calculate net operating income and run the key metrics: cap rate, cash-on-cash return, internal rate of return across a hold period, and debt service coverage ratio. But I don't stop at one scenario. I run a base case, a downside case, and an upside case. The downside case isn't some dramatic "apocalypse" scenario. It's just what happens if your optimistic vacancy assumption is wrong by five percentage points and your expense growth is two percent higher than expected. That alone will kill a deal that looked fine on paper.

The Specific Problem I Ran Into

Two years ago I was analyzing a small medical office building in the Midwest. The seller's pro forma looked decent. Cap rate around 7.5 percent, stable occupancy. I went through the lease roll and noticed something nobody else seemed to catch. One of the anchor tenants had a lease that said they could terminate early at year three with ninety days notice and a penalty equal to three months' rent. The original underwriter had treated that lease as if it ran through year five, the full stated term. If that tenant walked at year three, the building went from ninety-two percent occupied to sixty-eight percent overnight, and medical tenants are notoriously slow to re-lease because of build-out requirements and zoning restrictions. The workaround was straightforward once I saw it. I recalculated the underwriting with the worst-case termination timing — year three — and factored in an eighteen-month lease-up period at seventy percent of asking rent. The deal flipped from profitable to underwater pretty quickly. I walked away from it. The seller's agent was confused when I brought it up. They had never thought about the early termination clause because they'd been looking at the rent roll as a static document rather than a living schedule of obligations that could change.

What Beginners Miss

One thing that trips people up repeatedly is confusing stabilized NOI with projected NOI. A property might be currently leased to a mix of short-term tenants and month-to-month occupants. The current cash flow is real but it's not what the deal will produce once everything matures. Underwriting to current NOI on a turnover-heavy asset is like judging a restaurant's profitability by its lunch rush instead of its annual revenue. You need to underwrite to stabilized conditions, which means projecting what every unit will rent for when it's properly leased at market rate, not what it's currently generating during a transition period. Another blind spot is ignoring the cost of capital in your own analysis. People will obsess over a half-point difference in the cap rate and not think about whether they can actually get financed at the rate they're assuming. Right now, commercial loan rates are meaningfully higher than they were five years ago, and debt yield requirements have tightened. A deal that makes sense at a 6.5 percent interest rate with a thirty-year amortization might not work at 8 percent with a twenty-five-year term. The debt service jumps enough to materially change your equity return. I always run my financing assumptions independently from the property-level underwriting so I'm not accidentally double-counting some benefit or missing a cost entirely.

Tools I Use and Why

I build most of my models in Excel because it's what everyone else on the deal team needs to see. I also keep a simplified version in Google Sheets so I can share it without sending a proprietary file. The choice doesn't matter much. What matters is the discipline of building it from scratch rather than adapting someone else's template. Every template I've ever inherited had at least one hardcoded assumption buried somewhere that didn't apply to the asset I was analyzing. Finding it usually happens too late. For market data, I pull from CoStar when I have access through a broker, CBRE research reports for macro trends, and local municipal planning department websites for pipeline supply. The municipal site seems obvious until you're three weeks into due diligence and realize the city approved a rezoning that will change the surrounding land use but nobody told you about it.

When This Methodology Falls Apart

Commercial real estate analysis investments does not work well for unique or highly specialized properties. A single-tenant industrial facility built for one specific manufacturer, a specialized self-storage operation with unusual climate control requirements, or a restaurant-rented retail space where the tenant's business viability drives the lease payment — these all have too many idiosyncratic variables to model reliably. In those cases, the best approach is a conservative underwrite based on alternative use value and a shorter hold period, ideally with an exit strategy that doesn't depend on the current tenant staying. If you can't realistically imagine five different tenants for a space, you probably shouldn't be analyzing it as a core investment. It's a development risk, not a value-add or stabilized play, and the analysis framework needs to reflect that. Another limitation is that the methodology assumes you have access to decent lease data. In off-market deals or auctions where documentation is thin, your entire model is built on incomplete information. The workaround is to explicitly state your assumptions as conditional variables and model the range of outcomes rather than a single point estimate. A deal where you know fifty percent of what's in the rent roll is not a deal where you should be confident in a precise IRR number. The precision is fake. Anyone who gives you a five-year return to the decimal in that situation is either guessing or lying. The honest takeaway is that the analysis is only as good as the input assumptions, and most people are far too confident in their assumptions. The market doesn't punish bad numbers. It punishes overconfident ones.