What I’m thinking about: How we took on the most moving-parts deal we’ve ever funded (an entitlement play in the western US) and spent months structuring it toward as close to a layup as we can engineer…where, as it stands today, the risk we’re left managing is mostly the calendar, not the capital.
After months of back and forth, the docs are signed and our capital is moving, with ~$71K out the door so far and more queued behind specific milestones.
This is the deal type we’ve been circling for a while…funding the soft costs to turn a raw parcel into a platted project that a large home builder buys on the back end.
Earlier this year I broke down entitlement deals and how we structured one. For the deal we pulled the trigger on: Same operator, similar bones, different location and deal size.
As a reminder, an entitlement deal (in the land space) usually means putting up the soft costs (earnest money deposits or extensions, engineering reports, surveys, environmental work, planning and zoning applications, plat approvals, and legal back and forth) to take raw land that’s under contract, but not yet purchased, and get it approved and platted.
Then a developer or builder buys the finished paper lots (lots that are fully approved and subdivided on paper, but not yet physically built out).
As always, our job was to strip out as much risk as humanly possible before a SINGLE dollar left our account.
Why most capital runs from it
You’re spending real money on approvals for an asset you (typically) don’t own yet. So if the end buyer walks (or you are unable to identify one), or costs balloon, or the seller bails, or the local planning authority says no…your soft-cost capital can go to ZERO, with no asset to fall back on. Caveat emptor.
Many senior lenders won’t touch these (no hard collateral, often too binary an outcome, each deal has significant characteristic variance that makes it difficult to build a standard playbook), and why it falls to boutique capital with the appetite and the discipline to do the work.
So we keep one hard requirement before we fund a dollar of soft costs…the takeout buyer (ideally more than one) has to already be lined up, under a letter of intent (LOI) at minimum. No identified dev/builder, no deal. We don’t fund speculation.
The shape of the deal
Rough numbers: an ~8 acre parcel becoming ~36 platted home sites in a growing metro. ~$2M to buy the underlying land (we’re not the ones buying the land, just funding the approvals), and a national builder contracted at roughly $4M on the back end.
All-in soft costs land somewhere in the ~$350K to $450K range (~$200K+ of that related to EMD and seller extension costs), give or take whatever the process throws at us (always plan for the worst, hope for the best).
Our capital entered the picture at the top of June, with the builder expected to close by the start of September (though they have the ability to extend through next April).
=====
The de-risking here was the real work, and the order matters.
First, the builder went from LOI to a signed purchase agreement (intent became obligation).
Then the Phase 2 environmental (the deeper contamination study) came back fully cleared, and that contingency was struck from the builder’s final purchase agreement amendment, so the risk was retired rather than left pending.
On top of that, the builder’s corporate investment committee and final due diligence both signed off before our first wire (no small thing, getting a multi-billion-dollar builder’s internal committee to commit).
Their only contractual out from here is if the preliminary plat isn’t approved. Short of that, they either proceed or they forfeit ~$400K in hard cash (their 10% deposit, now non-refundable).
=====
Our capital goes out in tranches tied to specific milestones, never in a lump (I’ll spare you the inside baseball), so we’re never further out over our skis than the next checkpoint justifies.
The plat sits with a second-generation local civil engineer who has never had a subdivision in this area fail to get approved under his watch (~150 under his belt).
We pushed him HARD on the prelim plat approval process, especially since we did not have previous experience in that region.
His read is ~95%+ that the plat clears as presented, and the residual ~5% is the planning committee asking for minor modifications, not a denial. A flat no, in his estimate, is virtually implausible here.
The real risk here is time, not principal
Even in the extremely unlikely world where the builder walked away entirely (this builder has a rep of never walking from deals), we are not staring at a zero (the scenario everyone fears with these). We already have verbal backup buyer interest from another large regional builder, because the underlying parcel is simply that attractive (“location, location, location”).
I stress-tested the submarket through ReVenture, and the fundamentals point to a stable home market skewed toward higher-income buyers (critical in this historically dispo-challenged market), and the individual-lot comps in the area conservatively support the value of what we’re funding.
So the honest worst case is a longer road to disposition while we line up the next buyer, not lost principal.
Generally, the best operators in this industry convert principal risk into timeline risk, and don’t pretend they can avoid risk altogether. Do that, and you’ve won most of the battle before the deal even plays out.
=====
Nevertheless, timeline risk is no joke in this deal, especially in light of a fragile macro with rates expected to rise again, which might encourage the takeout builder to slow-play the close.
Additionally, there’s a small window for the plat to be approved before our seller extension expires, and if we are forced to close on the property before the builder’s takeout terms are met, a bridge loan is expected to be a very expensive proposition (again, mindful of a rising-rate environment).
If a bridge loan is required, their lender’s capital will sit senior to ours, and if combined with an extended close scenario, the pressure to exit will compound daily.
Finally, regardless of how well we underwrote this deal, it’s still the first of its ilk we’ve pulled the trigger on, in a geography we’re unfamiliar with, and run by an operator (one of the most sophisticated and communicative folks we’ve had the pleasure to work with so far) we’ve never worked with before (whose expertise we are heavily leaning on…naturally incurring “key-man risk”).
With all of that in mind, the terms we requested (the absolute minimum to enable us to pull the trigger), were a 20% annualized preferred return that accelerates to a 2x (a 100% return on our capital) at close, plus a slice of the promote (the upside above our preferred return).
To note, the operator also contributed significant capital to this deal, in addition to sweat equity, and his capital sits junior to ours in the waterfall, yet on this structure he stands to make somewhere in the neighborhood of ~5x to 7.3x.
That is the entitlement game in one sentence…rare, outsized returns paired with high variance, reliant on deep, local relationships, which is exactly why most big senior lenders won’t fund the soft costs, which is why it falls to boutique capital to underwrite (FYI it’s entirely possible we underpriced the risk here, TBD).
AI is the only reason we could underwrite this
Without the tooling we’ve built inside Claude Cowork, I don’t think we do this deal (not at this speed, anyway).
The underwriting back and forth alone was enormous, and we ran almost all of it (plus the 100+ pages of legal drafting) in-house.
The operator had his own attorney firm reviewing and redlining our documents, and they came back treating ours as sophisticated, attorney-to-attorney work, with targeted redlines instead of scoffing, “Who on earth wrote this?” Honestly, that landed as a remarkable bit of validation.
My read, after living in these docs for weeks, is that drafting legalese well is basically an engineering and coding problem combined, if you understand which questions to ask and where to push versus where to leave well enough alone.
I think attorney review is getting less valuable by the day (and clearly I’m putting my money where my mouth is, considering ~$200K+ in capital is committed with AI-driven legal docs)…unless the case law is genuinely unprecedented, in which case you absolutely still pay for a specialist.
Remember, AI is only as good as the context you give it (our many years in land are what let us ask the right questions in the first place). I still utilize my own judgment, don’t take anything for granted, push back when something is off, and watch for hallucinations on long threads (they still happen).
What used to take a room of specialists weeks, we now do in days…and this capability is reshaping our core strategy: an extremely capital-efficient business, low overhead, and near-universal underwriting capability focused on low-volume, high-value deals.
(I’ve already built out supplementary Claude skills that allow us to rapidly DD real estate deals that are outside our typical wheelhouse, utilizing the learnings from this entitlement deal as a standard…never repeat valuable work if you don’t have to.)
=====
To be clear, this deal isn’t closed yet. A project with this many moving parts can certainly get more complex before the finish line, and if it does, you’ll read about it here.
Nevertheless, this is about as de-risked an opportunity as we’ve ever encountered, and things are going according to plan…so far.
=====
If you’re an experienced operator with routine deal flow, especially if you’ve got an entitlement deal with a takeout already under LOI (or frankly any deal that needs serious, reliable capital), def reach out.
We write checks from $50K+. We close 100% of the deals we commit to. And we bring national underwriting built across thousands of reviewed transactions, now sharpened by the most carefully tuned AI workflows we’ve built.
