Most deals look good in the pitchbook. The ones we close look good after the stress test. The gap between those two is where the edge lives — and it's the part of the work nobody markets.
Every deal that comes across our desk gets the same treatment. We don't have a separate process for "interesting" deals and "obvious passes." The same diligence machine runs on every LOI candidate, because the only way to know which is which is to do the work.
The four scenarios we run on every deal
Our base case isn't the one we underwrite to. It's the one we use to start the conversation. We run every deal through four scenarios before we'll sign an LOI:
1. Lender stress case
What does the deal look like if we have to refinance at the current senior rate plus 200 basis points? If the answer is "we're a forced seller at the wrong number," the leverage is too aggressive. We don't size debt for the rate environment we hope for — we size for the one that's actually in front of us, with margin.
2. Lease-up stress case
What if the value-add lease-up takes nine months longer than we modeled? What if asking rents come in 8% below our underwriting at execution? Our IRR has to survive both of those stresses simultaneously, not just one at a time. The dirty secret of most rosy underwrites is that they stress one variable at a time.
3. Tenant concentration case
If our largest tenant doesn't renew, what's the cost — in vacancy, in lease-up timing, in capex to re-tenant — and how does that change the equity return? If we can't get comfortable with that outcome at the asking price, the price needs to come down or we walk.
4. Exit cap shock case
What if we exit at our entry cap rate instead of the compressed cap we're modeling? The realized returns should still pencil. We don't underwrite cap compression as part of our base case — we treat it as upside. If the deal only works on the back of cap compression, it's a beta trade, not an alpha trade.
If a deal can't survive all four stresses simultaneously, we walk. We've walked from more deals than we've closed. Every walk has been the right call.
The line items that have killed our last six passes
We track every deal we pass on. Not in a dramatic way — just a running list of which line item, on which deal, was the one that broke the trade. Here's the distribution from the last six passes:
- Insurance trajectory (2 deals). Florida insurance is the single biggest non-controllable expense line we underwrite. Every deal we close has a 15% YoY insurance step-up baked into the model. Two recent passes had insurance quotes come in 40%+ over our underwriting on the existing carrier — and the bid we got from the broker for next year's renewal was wider still.
- Roof / capex disclosure mismatch (1 deal). The Phase I said the roof had "remaining useful life." The follow-up roofing consultant gave it 18–30 months. The seller wouldn't credit the buyer at LOI. That's roughly a half-million dollars of capex hidden in plain sight.
- Tenant concentration credit (1 deal). Two tenants on the rent roll were technically separate entities but had the same ownership group and shared back-office staff. The "diversified rent roll" was actually concentrated. Pricing didn't reflect it.
- Tax abatement expiration (1 deal). A pending property tax abatement set to expire mid-hold would have erased the going-in coupon. The seller's pro-forma quietly assumed renewal at current terms. We checked with the county; renewal was not assured.
- Exit cap stretch (1 deal). The seller's deal book underwrote a 25-bps cap compression to make the IRR pencil. Strip out the compression and the deal returned a 9% levered IRR, not the 15% claimed. We let someone else write the check.
None of these were bad assets. They were assets being sold at prices that didn't reflect the specific line item that mattered. That's the work — finding the line item that matters on every deal.
What the network sees before we sign an LOI
The diligence we run in-house gets us 80% of the way to a decision. The last 20% — the part that separates real underwriting from financial-model underwriting — comes from the network we've built.
- Our brokers tell us what the actual recent comp set is, not the cherry-picked comps the seller's broker handed us. They tell us what's about to come to market and whether we'd rather wait.
- Our lenders pre-clear the cap structure before we LOI. We don't get to LOI on a structure that doesn't work for our debt — that's a waste of everyone's time.
- Our contractors walk the building with us during diligence. The actual capex number — not the "deferred maintenance" number from the rent roll — comes from contractors who'll execute the work, not from a desktop assumption.
- Our operators tell us what the actual day-to-day cost-to-operate looks like on a comparable asset they're already running. The operating expense line is the most lied-about line in commercial real estate, and it's the one our operators can call from memory.
This network is the reason our underwriting holds up after we close. It's the reason we can move fast on the deals that pass the stress test, and walk fast on the ones that don't.
The deals that close vs. the deals that don't
For every deal we close, we look at six to ten that don't survive the diligence stage. That ratio is the discipline. The investors who write checks with us aren't paying for our willingness to deploy capital — they're paying for our willingness to not deploy it on the wrong deal.
"The deal that gets away is rarely the deal you should have done. The deal that breaks the fund is almost always the one you talked yourself into."
This is the work. It's not glamorous, it's not in the pitchbook, and it doesn't make for great LinkedIn content. But it's the difference between a fund that compounds and a fund that gives back its returns at the next refinance cycle. Underwriting is the edge.