Case files
Representative engagements.
Composites drawn from the principals' engagements and operating careers, anonymized by size band and role. Named references available in conversation.
01
HMBS issuer · $2.1B outstanding
IssuerBuyout liquidity
The wave, funded eleven months early.
Situation
Fixed-rate pools issued 2014–2017 were compounding toward mandatory buyout in a cluster. Treasury was sizing its warehouse line off last year's buyout run-rate, and the board had never seen the crossing laid out by month.
The work
A loan-level projection of the buyout date for every loan, then an assignability screen — tax-and-insurance defaults, non-occupancy, due-and-payable events — to separate loans that would leave the balance sheet through assignment from loans that would not. Facility sizing and covenant headroom followed from the two populations, not one blended number.
Outcome
The facility was upsized before the wave arrived. Cure work on the non-assignable population began while the loans were still in the pools, and a portion assigned that would otherwise have sat on balance sheet.
02
FHA servicer · 140,000 loans
ServicerClaims & curtailment
Nine percent of every claim was leaking.
Situation
Conveyance claims were paying, so nobody was alarmed. But interest and expense curtailments for missed first-legal and conveyance deadlines were running at 9.1% of claim amount, and the servicer could not say which states, which attorney networks or which process steps were responsible.
The work
Every curtailment was traced to the timeline it breached and the step that breached it. Two attorney networks and one internal hand-off accounted for most of the leakage. Claim filing was re-sequenced, supplemental claims were filed where the record supported them, and a monthly leakage report replaced the annual look-back.
Outcome
Leakage fell to 3.4% of claim amount within three quarters, and prior-period supplemental filings recovered money the servicer had written off.
03
Regional audit firm · HECM issuer client
AuditorValuation support
The fair value that could be defended.
Situation
The client's HMSR and bought-out loan balances were material, and the assumptions behind them — assignment timing, claim payment lag, non-reimbursable cost, discount rate — came from a spreadsheet no one on the engagement team could independently test.
The work
An independent assumption set built from the client's own claim history and the wider loan population, documented line by line with the evidence behind each input. Sensitivity tables the audit committee could read without a glossary. Both practices contributed: market inputs from capital markets, timing and cost inputs from servicing.
Outcome
The audit closed without a valuation adjustment. The assumption set and its documentation were reused the following year with a one-page update.
04
Public-sector housing finance analytics team
Guarantor sideSurveillance build
Seeing the issuer before the issuer fails.
Situation
The team could see issuer-level balances but not the shape of what was coming: how much of each issuer's book would have to be bought out, how much of that could assign, and how fast tax-and-insurance defaults were building underneath.
The work
A quarterly surveillance pack built on loan-level data: buyout pipeline against issuer size, non-assignable share, default trend and concentration — delivered as a PDF on a schedule, with the model behind it available in working sessions.
Outcome
The pack replaced an ad-hoc spreadsheet process and gave the team a consistent early-warning view across issuers, two quarters ahead of the balance-sheet signals they had relied on before.