Buyback demands do not usually come from exotic fraud. They come from a short list of defect categories, and most of them were sitting in the file at origination.
Below is that list, what each one actually is, and how a check before funding finds it. All five, it does — four on the first pass, and the fifth by reconfirming the file before it closes.
A lien or mortgage the borrower did not put on the application. It understates debt-to-income from the moment the file opens, so every ratio downstream is wrong.
Before funding
A MERS search by Social Security number reports active and inactive liens, including a mortgage the application omits. Civil court records add judgments and bankruptcies that bureau data misses.
A fabricated employer, an inflated salary, or a job that ended before closing. One of the most common defect categories in buyback populations.
Before funding
The stated employer is verified against business records — that the company exists, at that address, with that phone number — plus how far it sits from the subject property.
An investment property presented as a primary residence to secure better terms. It changes the risk profile the investor priced.
Before funding
Ownership history, current and undisclosed owners, recent addresses and foreclosure history are tested against the stated occupancy, along with days since the last transfer.
A transposed Social Security number, an alias, or a synthetic identity. The first is clerical; the last is fraud; both look the same in a pass-or-fail report.
Before funding
Social Security number, date of birth, name, address, ZIP and phone are verified and graded individually, so a miskey is separated from a genuine discrepancy.
A borrower who finances a car between application and closing. The file was accurate when it was underwritten and inaccurate when it funded.
Before funding
Reconfirm before closing. A reconfirm re-runs the workflow against changed data, so a lien or judgment recorded after the original order appears while you can still act on it. Lenders run as many reconfirms as their milestone workflow requires.
A defect found before funding costs a condition and a few days. The same defect found after the loan is sold costs the loan, plus the servicing, plus the time your team spends assembling a response.
PitchPoint’s Application Data Verification workflow runs these checks on every application, inside your loan origination system, and grades what it finds so your underwriters work the real problems first. Lenders converting to PitchPoint see roughly 30% lower verification cost, largely because fewer files come back.
A short list of defect categories rather than rare fraud: undisclosed liabilities that understate debt-to-income, income and employment misrepresentation, occupancy misrepresentation, identity and data integrity problems, and missing or non-compliant documentation. Most are present in the file at origination.
They catch different things. Monitoring watches the credit file through the quiet period and alerts on new tradelines. An undisclosed existing mortgage was already there and was left off the application, which misstates debt-to-income from the start. A MERS search by Social Security number reports it when the verification runs.
Reconfirm before closing. A reconfirm re-runs the ADV workflow against changed data, so a lien or judgment recorded after the original order shows up while the loan is still yours to fix. Lenders run as many reconfirms as their milestone workflow requires.
Inside your loan origination system. Application Data Verification runs natively in ICE Encompass, MeridianLink Mortgage, Calyx Point, Calyx Path and ByteLink, and is also ordered through Mortgage Credit Link, now XedaLink. See the integrations page.