Part III — Underwriting

Chapters 14–19

Underwriting is where new loan officers discover that their job is not sales.

You have a borrower who trusts you, an application that looks clean, and a set of numbers you calculated correctly. None of that is an approval. An approval is a specific person, who has never met your borrower, certifying in writing that this file satisfies a rulebook written by an entity several steps removed — and doing so knowing that if they are wrong, the loan can be sent back to their employer to be bought at par.

Six chapters on how that decision actually gets made, and what you can do about it.

Chapter 14 is the conventional rulebook: who writes it, where to read it for free, and what it actually asks. Eligibility versus creditworthiness. Waiting periods after a credit event. Layered risk, which is the single most useful concept in underwriting and the one most often explained badly. And overlays — your employer's stricter rules stacked on top of the agency's, which are not in the agency guide and which are responsible for a large share of the "but I thought this was allowed" conversations in a loan officer's first year.

Chapter 15 is the automated underwriting system, and it deserves its own chapter because it is the single most consequential piece of software in the process. Desktop Underwriter, Loan Product Advisor, and the TOTAL Scorecard do not merely approve or decline. They write the document list. Learning to read a findings report — the recommendation, the verification messages, the observations — is the fastest available upgrade to a new loan officer's competence, and most never do it.

Chapters 16 and 17 take the government programs seriously. FHA gets a full chapter because it is the program that serves borrowers conventional lending will not, because its rules genuinely differ, and because the mortgage insurance structure is the most misunderstood thing in residential lending. VA and USDA share a chapter: entitlement, the Certificate of Eligibility, the funding fee and its exemptions, residual income — a qualifying test the other programs do not have and arguably should — and the geography and income limits that govern USDA.

Chapter 18 is the appraisal. What it is for, who orders it and why you may not, how the sales comparison approach works, and then the part every loan officer eventually needs: what to do when the value comes in below the contract price. Five options, priced. The reconsideration of value, done properly rather than angrily. And the conversation with a borrower whose transaction just changed shape eleven days before closing.

Chapter 19 is conditions — the stip sheet, prior-to-doc versus prior-to-funding, the letter of explanation, re-verification, and the pre-closing credit refresh that catches the debt your borrower took on and did not mention. This is the chapter where the Linden Street file nearly dies, and it is the most operationally useful chapter in the book. Conditions are where the calendar goes, and the calendar is where the money goes.

The theme running through Part III is the book's second: the file is approved when it is documented, not when it is promised. Every chapter here is an elaboration of that sentence from a different angle — the guideline's angle, the machine's angle, the government's angle, the appraiser's angle, and finally the condition list's angle.

The Loan File is submitted, conditionally approved with eleven conditions, and then, on day forty-four, blown apart by a \$611 monthly payment that did not exist when the approval was issued. How that is discovered, and how it is fixed in four business days, is Chapter 19.

Chapters in This Part