Research, options, and a proposed path.
Reads loan agreements, appraisals, rent rolls, and operating statements in any format and turns them into structured, stress-testable data.
Property records, covenant tracking, lease rollover, counterparties, investor reports, single-asset stress testing.
Portfolio concentration, weighted LTV by property type, fixed vs floating vs hedged, maturity ladders, watch list and CRO reports, borrower action plans.
Two research passes. The first benchmarked four adjacent markets: independent loan review, bank-side CRE risk software, borrower-side debt management and brokerage, and loan portfolio trading. Who the incumbents are, what they charge, how they sell, what regulation applies.
The second was a skeptical test of the loan-trading idea, because it was the one we liked most.
Sources: FDIC, OCC, and Federal Reserve guidance; MBA, S&P Global, Trepp, MSCI, New York Fed; company filings, press, and funding databases.
| Figure | What it is | What it means |
|---|---|---|
| $875B | Commercial mortgages maturing in 2026, of $5.0T outstanding (MBA). Peak expected in 2027. | The maturity wall is real. Loans written at 3.5% to 4.7% are refinancing near 6% to 7%. |
| 3% | Distressed share of CRE transactions this cycle, vs about 20% after the GFC (MSCI). | Banks extended and modified instead of selling. The forced-seller wave didn't come. |
| 2025 | Exams refocused on material financial risk, higher bar for MRAs, examiner headcount down. | Regulatory pressure to shrink CRE books is easing. Anything that needs a crackdown is a shrinking market. |
| 1.47% | Multifamily delinquency at banks, Q1 2026, highest since 2013. CMBS is 7.35%. | Bank credit is softening, not breaking. About 30% of community banks are CRE-concentrated. |
| $1B | Exposure per loan reviewer today; reviewer experience fell double digits year over year (Abrigo). | Banks can't staff independent review either way. This is the durable driver. |
Anything that depends on distress is shrinking. Anything mandatory, recurring, and short on talent is not.
An operating layer for CRE owners.
Worth keeping: the confidence-routing design and the hospitality depth.
Build it cheaper and undercut EnFi.
Free owner platform, aggregate the data, broker the refinance.
The idea: banks over their CRE limits have to sell loans, pricing a pool takes weeks of analyst time, and if the engine can abstract and stress a pool in days we either bid tighter or charge a fee to run the process. Four things broke it.
On large pools nobody reads every file. Full abstraction adds a little information, not a new capability.
Bid-ask is driven by views on future rents and cash flow, cost of capital, and 20%-plus return hurdles. Not analyst hours.
Blackstone and Rialto paid $1.2B for 20% with 50% FDIC seller financing. Cheap leverage and servicing won.
Distress is about 3% of transactions. Banks extended, modified, raised capital. Supervision has loosened since.
What survives: pre-sale pool optimization sold to the bank as a service. Same engine, no fund, no conflict with being the bank's reviewer.
Tech-enabled services is where the capital is going right now, but it only works when a few conditions line up. Brokerage, the Lev model, fails most of them. Independent credit review passes all of them.
| Condition | CRE debt brokerage | Independent credit review |
|---|---|---|
| The service is mandatory and recurring | No | Yes |
| Labor is the cost the technology replaces | Partly | Yes |
| The buyer is credential-gated, so trust is scarce | No | Yes |
| Incumbents are fragmented, founder-owned, and aging | No | Yes |
| Private equity is already consolidating the tier above | No | Yes |
Every bank is required to have independent credit review under the 2020 interagency guidance. The incumbents are small firms of ex-bankers and ex-examiners with no technology, and PE has bought the CPA firms above them but hasn't touched the boutiques yet.
Audit committees and examiners buy the reviewer, not the tool. A former regional bank CFO with a stress framework in FDIC examiner coursework is a credential the incumbents can't match. The software never appears in the pitch.
Incumbents deliver with people at roughly 40% gross margin. The engine takes it to 65% or 70%. Abrigo's AI arms banks doing their own reviews. Nobody is running an AI-native independent firm yet.
The review is required regardless of the cycle, and the binding constraint is reviewer talent, which is getting scarcer either way. This works in a soft market.
Organic case: 60 to 80 institutions at $75k to $150k, $6M to $10M of revenue at 65%-plus margin, cash-flowing from the first engagement. Then the roll-up on slide 12.
Credit unions are growing commercial books with no CRE credit infrastructure. A CUSO can legally provide credit review to its members. One contract, many institutions, distribution nobody can copy.
Boutique review is a $15k to $75k cost line. A $250k engagement holds when it's framed as avoided capital and avoided enforcement, backed by a named methodology.
The engine shows which borrowers are rate-sensitive; the bank sells the swap. Hedged loans earn about 1.62% in fees vs 0.26% unhedged. Chatham and PCBB built this. Separate entity.
Deposit concentration, funding buckets, rate shocks. Post-SVB this is where examiners look, and it's the framework that was actually adopted into coursework.
Annual concentration stress test, board deck, and policy review under $50k for the $500M to $3B tier. Almost entirely software-delivered.
What's left of the trading idea. Help a bank structure and price a pool before it goes to market. Same engine, sold to the seller.
One paid engagement with a bank we already know. CRE portfolio stress test plus watch list, $150k to $250k, money-back structure. Nothing gets built before this.
Tells us whether a warm relationship converts at a real price off a prototype.
Three to five more engagements. We build the pipeline against real bank documents as paid work. SOC 2 starts immediately. One conversation on the CUSO.
Tells us what breaks in extraction, what examiners accept, whether the CUSO is real.
First acquisition. Launch the exam-readiness tier. Stand up the hedge desk in a separate entity.
Tells us whether the roll-up works.
Principal: the credential, the methodology, the face in front of CROs and examiners. Original: build, operate, go-to-market, data governance, SOC 2, capital, acquisitions. Family office: distribution, if the CUSO is available.
Two entities from day one: the review firm stays independent, transaction economics sit elsewhere with real walls. Data rights explicit and de-identified in every contract.
The independent loan review market is a handful of founder-owned boutiques, most started in the 80s and 90s, with sticky bank clients and no technology. Nobody has consolidated them.
| Firm | What we know |
|---|---|
| CEIS Review | Founded 1989. About $6.7M revenue, 39 staff, 200+ bank clients. Founder-owned. |
| Young & Associates | Founded 1978. About 50 consultants. Already through one succession. |
| Ardmore Banking Advisors | Founded 1991. Small senior team, management-owned. |
| ProBank Austin | The comp. Sold to Forvis in 2022 at about $20M revenue and 50+ professionals. |
Buyers at the end: Forvis Mazars, Crowe, Baker Tilly, or the software platforms with PE behind them, Ncontracts (Hg) and Abrigo (Accel-KKR, Carlyle). PE paid 10 to 15x EBITDA for the CPA platforms one tier up.
Illustrative. Boutique multiples and margins aren't published; validate on the first acquisition.