The deal we walked away from (and why)

Let me tell you about a deal we walked away from. Not just any deal: this one had all the bells and whistles. Flashy, smart, compelling on paper, the kind that makes your eyebrows lift when the teaser deck opens. You know that feeling when you want something to be true? Not consciously biased, just quietly hoping this might be the one. That's the feeling I want to walk you through, because saying no to it reveals more about how real investors decide than any winning deal I could describe. We don't get rich from saying yes. We get wealthy from walking away from almost good enough.

The deal came from a respected sponsor, someone we'd watched operate, with credibility, solid exits, a decent platform. A large ground-up multifamily project in a fast-growing market. I am going to describe this one loosely, and more loosely than I did the first time I told the story, because the details I used then, the region, the unit count, the incentive package, the sponsor's project number, add up to a name for anyone who works in that market. The lesson does not need them. What matters is the shape: class A, a submarket with real buzz, strong demographic tailwinds, and an attractive projected return net of fees. On paper, gorgeous.

Three cracks, none of them scandalous

So we did what we always do and put it through our screening process. It passed the first gate, which never means yes, only that it's worth our time. What the screen caught was subtle: a softening absorption trend in that submarket. Not a crash. Not even a red alert. Just softness. On its own that kills nothing. But softness is how trends introduce themselves, so we pulled the thread.

The sponsor was underwriting lease-up at eleven months, stabilized occupancy inside a year. When our team pulled real comparable data, nearby properties were giving away free months and one was still well short of stabilized more than a year in. That might sound like splitting hairs, and it isn't. When you shave three to six months off an absorption assumption, your cash flow shifts, your break-even shifts, your exit timeline shifts. In development, time isn't just money, it's velocity. Lose velocity and the whole capital stack starts to wobble.

But the real crack was in the sponsor's pipeline. This was one of several concurrent builds: same market, same window, same trades, overlapping teams. And here's a lesson we've learned the hard way: concentration hides behind momentum. Everybody wants to back the busy sponsor, cranes up, capital raised, headlines made. But if one of those dominoes falls, the others are standing very close. So we made calls: general contractors, brokers, our own person on the ground. What came back was the ordinary texture of a market under strain, labor tight and sub costs drifting upward, plus enough hesitation in how people answered that we stopped asking. I am not going to repeat the specifics of what we were told. It was secondhand, we never verified it, and passing on a deal is a decision that requires far less proof than repeating an allegation in public does.

And I want to be honest about what we did not find. No fraud. No cooked numbers. No scandal. This was a perfectly average moment in capital allocation: a decent deal, a decent team, a decent market. That is exactly the trap. It's never the obvious disasters that burn you, it's the almost-good ones, where you talk yourself into the upside, the deck is handsome, your team is split, and you catch yourself thinking maybe I'll just do one instead of three. Mediocrity wrapped in momentum is where portfolios die.

The whisper circuit

This is what the runtime didn't get to teach, and I add it here as the extended class: how to make those ground-truth calls yourself, because that's the part of this story most investors can't replicate and assume requires a team. It doesn't. It requires four calls, none of them to the sponsor.

Call one: a general contractor or subcontractor working in that submarket. You are not asking their opinion of the project. You are asking about their world: how is labor availability right now, what have material and sub prices done in the last ninety days, and is anyone in this market paying slowly? Slow payment is the earliest tremor there is, and trades always know before lenders do.

Call two: a leasing agent at a competing property. Again, not what they think of the new development, but what they are actually doing today: what concessions are you offering right now, and how long is a unit sitting? Asking rents are marketing. Concessions are the truth serum, because free months are what a market gives up when it's soft and nobody has updated the sign yet.

Call three: an investor from one of the sponsor's earlier deals, and specifically one who is not on the reference list you were given. Ask what surprised them, how the sponsor communicated when something slipped, and whether they'd do it again on the same terms.

Call four: a lender or debt broker active in that market and asset type. Ask what terms they would quote today for this profile. The debt market prices reality faster than any equity deck does, and if a lender's answer is meaningfully worse than the assumptions in the model, you have your answer without needing anyone's permission.

Then three rules for reading what you hear. Concessions over asking rents, always. The trend over the level, because a market softening from an excellent base is more dangerous than a mediocre market that's stable. And tone counts: a pause before someone declines to comment is data, and so is enthusiasm that arrives too fast. Finally, do the counting exercise nobody does: list every active project your sponsor is running and mark which ones share a market, a trade base, a lender, or a timeline. Every shared dependency collapses the count. Four projects with four shared dependencies is not a diversified operator. It's one bet, reported as four.

One more rule, and it is the one I most had to learn: everything on this circuit is gathered privately, under an implicit understanding, from people who were candid because it wasn't going anywhere. It is for your decision. It is not material for your marketing.

The mirror

I am not going to tell you how that project turned out. I know some of it, and I have told that part of the story before, and I have come to think that was a mistake. Reporting the troubles of a deal I declined, using information I gathered privately and secondhand, is not diligence being shared. It is closer to settling a score in a game I chose not to play. It also flatters my judgment in a way the facts may not support, because deals go wrong for reasons that have nothing to do with what I noticed, and deals I passed on have gone beautifully while I quietly explained to myself why I was still right.

What I can tell you is that we slept fine the entire time, and not because we were clever. Because we honored the process, trusted the framework, and listened to the whispers underneath the noise. That is the only part of this I can actually take credit for.

So here is what I'll leave you with, and the mirror question that goes with it. The deal you say no to will define you more than the one you say yes to. Anyone can wire money when it's exciting. So the next time you're in that moment, with everyone else nodding, ask yourself plainly: does this deal honor my framework, or am I hoping it will? If the answer is hope, walk. Walking is wealth.

If you want to keep this conversation going each week, there is The Sunday Memo. Because the whispers arrive long before the headlines.

Founder of Infinity⁹. Here I write in my own voice.