Most deal-analysis writeups focus on the ones that closed. The more interesting story is usually the deal that didn't — because that's where you see what a sponsor actually believes about risk.
Here's a recent example. Names changed, location obscured, numbers slightly rounded, but the underwriting logic is intact.
The setup
A privately-held multi-tenant industrial property in a secondary South Florida submarket. Roughly 65,000 SF, eight tenants, average bay size 8,000 SF. Built 1988, modest condition, fully sprinklered, 24' clear height. The seller is the original developer's family trust; they've owned it since construction.
List price: $14.5M. In-place NOI: $920k. Going-in cap rate at ask: 6.35%. Asking rent across the rent roll: $11.20/SF NNN. Estimated market rent for the submarket: $14.50–$15.25/SF NNN.
On paper, this is exactly the kind of deal we look for. Below-market rents, motivated long-term seller, no institutional auction process. The mark-to-market math is obvious.
The headline underwrite
Our base case rolled the rent roll over a four-year period as leases expired:
- Year 1: 2 leases rolling at +30% bumps
- Year 2: 3 leases rolling at +28% bumps
- Year 3: 2 leases rolling at +24% bumps
- Year 4: 1 lease rolling at +20% bumps (one of the larger tenants)
Stabilized NOI at year 4: $1.32M. Stabilized cap rate on cost: 8.8%. Exit cap rate assumption: 6.5%. Implied unlevered IRR: 14.2%. Levered IRR at 60% LTV: 19.4%. Equity multiple: ~2.1x.
This is well within the band we typically target.
Where it started to crack
The asset-level math worked. The diligence is where it didn't.
Insurance came in higher than the model
We underwrote insurance at $1.65/SF based on recent broker quotes on comparable assets. The actual quote on this property came back at $2.40/SF — a function of the building's 1988 vintage, the roof's remaining useful life, and one historical wind-related claim from 2017. That's $49k more of fixed expense per year that doesn't come back in rent.
Roof had less life than the inspection initially suggested
The Phase I report flagged the roof as "near end of useful life." The follow-up roofing consultant we hired gave it 18–30 months. Replacement estimate: $480k for a torch-down system, $610k for TPO. Either is a capex hit in the first 24 months that we hadn't fully budgeted.
Tenant concentration was worse than the rent roll showed
Two of the eight tenants — both in the same industry — accounted for 41% of NOI. They were technically separate corporate entities but had the same ownership group and shared a back office. From a credit perspective they were one tenant. A renewal failure or a relocation of that group would create vacancy of more than 17,000 SF — almost a year of lease-up at submarket pace.
When we stress-tested the model assuming the concentrated tenant group didn't renew and we had a 9-month lease-up at flat rents, the equity IRR fell from 19.4% to 8.6%. That was a wider band of outcomes than we were comfortable with at the asking price.
What we offered, and why it wasn't enough
We came back with $13.1M and a structure that included an earn-out tied to the renewal of the concentrated tenant group. The seller's broker came back at $14.0M firm, no earn-out, and indicated they had a backup offer at that number.
Could we have closed at $14.0M? Probably. The base-case IRR would have been around 13% unlevered, 17% levered — still inside our band. But the downside case at $14.0M, with the concentrated tenant risk, put the equity multiple at 1.3x in a non-renewal scenario. That's not a downside we wanted to own at that price.
We walked.
The deal closed three weeks later
A regional sponsor we know closed it at the full $14.0M ask, with a more aggressive cap structure than we'd underwrite to. We'll find out over the next 24 months whether the concentrated tenant renews. We hope they do. If they don't, the new owner is going to have a more interesting equity check than they planned.
None of this is criticism of the buyer. They have a different cost of capital, a different LP structure, and possibly a different read on the tenant credit than we did. The point of the story is that two reasonable sponsors looked at the same deal, ran similar underwrites, and reached different conclusions about the asymmetry of the downside.
The takeaway
A clean base-case IRR is the easy part. The deals that work over a hold cycle are the ones where you've stress-tested the specific line item that, if it moves against you, breaks the trade. That line item is rarely the cap rate or the rent growth assumption. It's usually something that didn't make it into the headline summary — insurance, capex, a single tenant, a buried environmental cost.
"The deal that gets away is rarely the deal you should have done."
We've walked from a lot of deals. We've never regretted the walk. We've occasionally regretted the close.