Process · Green Peaks Capital · Dallas–Fort Worth
The most important number in a sponsor's track record isn't the deals they closed. It's the ratio, how many they looked at to get there.
A sponsor who underwrites forty deals and buys one is running a filter. A sponsor who underwrites three and buys one is running a search for justification. Both can show you the same closed transaction.
This article is the filter. Not a summary of it — the actual sequence, with the actual numbers, in the order they get applied.
The sequence
Intake
SAME DAY
Every sourced deal is logged and filed the day it arrives. Missing T-12, rent roll, or offering memorandum? We request them within 24 hours, along with the seller’s story, timeline, and call for offers. Deals get evaluated on documents, not narratives.
QuickScreen
24 HOURS · KILL FAST, KILL CHEAP
Hard stops applied first, then a rent-affordability check against median household income at one, three, and five miles. Go or Kill decision inside a day. Most deals end here.
Full underwriting
INCOME AND EXPENSES REBUILT FROM SOURCE
We do not use the seller’s proforma. We rebuild income from the unit-by-unit rent roll and actual T-12 collections, then normalize every expense line. Four return thresholds, all required. Then we stress it.
Letter of intent
PRICE IS AN OUTPUT
Our offer is whatever the underwriting produces, submitted whether or not it reaches the asking price. Walk-away price is set before the first counter and does not move without new information.
Due diligence
30–45 DAYS · WALK EVERY UNIT
Third-party condition assessment, unit-by-unit inspection, rent roll audited against actual leases, bank deposits verified against reported collections. Three possible outcomes, and all three are wins.
Close
EQUITY COMMITTED BEFORE CLOSING NOTICE
Settlement statement reviewed line by line by both partners and counsel 48 hours ahead. Takeover plan locked a week out: management contracted, banking live, tenant notices drafted.
Step one: the hard stops
Before anyone opens a model, a deal has to survive a list that requires no judgment at all. These are instant kills — no discussion, no exceptions, no “but look at the return.”
Automatic pass
- ✕Basis above renovated comps. If the purchase price exceeds what already-renovated comparable properties have sold for, the upside has been pre-spent by the seller.
- ✕Rent gap under 10%. No spread between in-place and market rent means no value to force. This is a momentum bet, not a value-add deal.
- ✕CapEx above $20,000 per door unless fully priced into the basis. Beyond that, we are a construction company that happens to own real estate.
- ✕Debt service coverage that fails our minimum threshold, tested at day-one actual income. Non-negotiable. This is the line between a bad year and a capital call.
- ✕Negative population or income growth in the submarket. We do not fight demographics.
- ✕Oversupplied submarkets — regardless of how attractive the price looks.
- ✕Outside the buy box: Class A, student housing, ground-up development, single-family, non-multifamily, or anything outside DFW.
- ✕Regulated-rent or subsidy-dependent housing — LIHTC, project-based Section 8, or any recorded rent/income restriction. Tenant-based vouchers are not part of this; those units are market-rate.
- ✕Flat roofs. Pitched construction only — a small rule that has saved us from a large recurring capital expense.
- ✕Flood zone or environmental flag on public record.
These rules exist because they were written when we were calm. A criterion set in advance is worth ten judgment calls made at 11pm on a deal you have already spent forty hours on.
Step two: rebuilding the numbers
Deals that survive the screen get fully underwritten. The core discipline is a single sentence:
Value what the property does, not what the broker says it could do.
Income comes from the current unit-by-unit rent roll and actual collections in the trailing twelve months, never the proforma. We quantify loss-to-lease, delinquency, concessions, and vacancy against the submarket.
Expenses are where most optimistic underwriting hides, so we normalize the seller's operating statement and add back what gets stripped out:
Expense normalization — what we add back
- A real management fee. Owner-operators often show none. We will be hiring third-party management.
- Market-rate payroll. Not the seller's cousin doing maintenance for free.
- Replacement reserves of at least $250 per unit per year. Buildings consume capital whether or not the operating statement says so.
- Property taxes at post-sale reassessment. In Texas the assessed value resets at the transaction price. The seller's historical tax line is not our tax line, and pretending otherwise is the most common way a DFW deal gets underwritten into failure.
- Insurance at today's quotes. Not the seller's expiring legacy policy. North Texas premiums have repriced sharply.
There is also a counterintuitive flag we watch for: an operating expense ratio below roughly 40% of effective gross income on a Class B or C asset is a warning, not a selling point. Real buildings cost real money to run. A ratio that low almost always means deferred maintenance, uncompensated owner labor, or expenses that simply were not recorded. When a deal looks too clean, it usually means the problems haven't been written down yet.
Our working assumption is that operating expenses land at or below 52% of effective gross income after we take over. If we cannot plausibly get there, the deal dies.
Step three: four thresholds, all required
A deal has to clear every one of these. Not three of four, not “close enough on coverage because the IRR is strong.”
We test four measures independently: an internal rate of return floor, a cash-on-cash return floor, an equity multiple floor, and a debt service coverage floor. All four have to clear — a deal that's strong on three and short on the fourth is a pass, not a deliberation.
The coverage floor is tested against actual, day-one income, not stabilized, not post-renovation, not proforma. The property has to cover its debt on the rent it collects the month we buy it. That is the difference between a plan that has room to be wrong and one that requires everything to go right.
We also target loan-to-value around 65% rather than the 75–80% the market will often lend. That is a deliberately more expensive choice: it means raising more equity for the same asset, which dilutes returns. What it buys is survivability. Leverage is what turns a difficult year into a lost investment, and the equity that gets wiped out in a downturn is almost never the equity in a conservatively levered deal.
Step four: breaking our own model
Before any deal goes to the partners, we try to break it. Every underwrite is stressed on three axes simultaneously:
- Rents down 5% from underwritten levels
- Exit cap up 50 basis points from our already-conservative assumption
- Renovation costs up 20% over budget — on top of the 10% contingency already built in
If the deal only works in the base case, it doesn't work.
This is where most otherwise-attractive deals die, and it is the single highest-value thing we do for the capital we steward. The base case is the outcome that happens if nothing surprises you. Nothing has ever not surprised us.
One related rule governs the offer itself: maximum offer price is an output of the underwriting, not an input. We never start from the asking price and work backward to justify it. If our number lands below the ask, we submit our number anyway. Motivated sellers move. Unmotivated ones were never our deal.
What a pass actually looks like
Two recent examples from our pipeline, anonymized. Both looked reasonable at first glance. Both died, for different reasons, at different stages.
Mid-1990s garden-style, ~80 units
Killed at underwriting
Debt service coverage came in negative on day-one actual income, and pricing guidance was well above what the in-place rent roll supported. The broker’s proforma got there; the actual T-12 did not. We passed and told the broker exactly which two lines killed it.
Early-1980s asset, infill submarket
Killed on renovation cost
Galvanized supply plumbing throughout. Replacement scope pushed all-in cost per door past our ceiling, and phased repiping in an occupied building means downtime and displacement the returns couldn’t absorb. Good location, wrong bones.
Neither of these was a bad property. Both will trade to someone. They were wrong for a strategy built to protect capital first.
Due diligence: three outcomes, all wins
Even after a signed contract, the filter keeps running. Due diligence is 30 to 45 days, and the standard we hold to hardest is this: we walk every unit. No exceptions. The units a seller steers you away from are the ones that change the price.
Alongside the physical work — condition assessment, roof, HVAC, plumbing scoped on anything built before 2000, electrical panel types, environmental — we audit the rent roll against actual signed leases unit by unit, and check twenty-four months of operating statements against bank deposits to confirm the collections are real.
Due diligence ends one of three ways, and we treat all three as successful outcomes:
- Pass — findings confirm the underwrite. Proceed to close.
- Re-price — material findings, documented with contractor bids attached, renegotiated as credits or price. We never re-trade without documented findings; that reputation is worth more than any single deal.
- Kill — terminate in writing before the inspection deadline. Walking away from a bad deal costs pride. Closing one costs years.
Why the fast no matters
Every deal a broker sends us gets an answer within 48 hours: yes, no with a specific reason, or the questions we need answered. Not “we'll circle back.” Not silence.
That is partly courtesy, and mostly self-interest. A broker who gets a real reason — “your OpEx assumption doesn't survive reassessment,” “the rent gap is 6%” — learns what we buy and sends better deals next time. The best properties are sold before they are listed, and they go to the buyer who responds fast, gives honest feedback, and closes what they tie up.
Discipline in the screening process is what earns the deal flow. And deal flow is what makes discipline affordable, because you can only say no to most things if you are seeing enough things.
For an investor, that loop is the thing worth diligencing. Not the pitch deck on the one deal that made it through. The forty that didn't, and the reasons why.
Deal examples described in this article have been anonymized and details altered; they are illustrative of process and do not describe any specific transaction. This article is provided for educational and informational purposes only. It is not an offer to sell or a solicitation of an offer to buy any security, and it is not investment, legal, or tax advice. Stated criteria and thresholds are current guidelines and are subject to change. Meeting internal thresholds does not guarantee any result. Real estate investments involve substantial risk, including the possible loss of principal. Past performance does not guarantee future results. Prospective investors should consult their own financial, legal, and tax advisors.