Chapter 6: The Loan Process: Application → Processing → Underwriting → Closing

"A file rarely fails because somebody did the wrong thing. It fails because for nine days nobody did anything, and then the calendar did something." — constructed; the argument of this chapter

Overview

Five chapters in, you know what a mortgage is, where the money comes from, what license you need to touch it, how to compute the payment, and which program a borrower might fit. What you do not yet have is the thing every one of those chapters quietly assumed: a process. A sequence, with stages, owners, handoffs, and dates.

That sequence is the actual product a loan officer sells. Borrowers think they are buying a rate. Real estate agents think they are buying certainty. What both are actually buying is a forty-five-day sequence of events that ends with a wire, and the only person in the transaction whose job is to make that sequence happen on time is you.

This chapter is the map. It follows one file — 4412 Linden Street, the purchase you opened in Chapter 1 — from a phone call at 8:40 on a Wednesday to a recorded deed on day 51. Along the way it answers the questions a new loan officer is too embarrassed to ask out loud: what does a processor actually do all day? What is an underwriter looking at? Why does a file that is "approved" still have eleven things wrong with it? Why did that approval take five days and the eleven conditions take nineteen?

That last question is the chapter. The naïve mental model of origination is a relay race: you run your leg, hand off, and the next person runs theirs. The real shape is different and much less comfortable. A file spends most of its life waiting — on an appraiser, on a title examiner, on a borrower who has not opened your email, on an underwriting queue, on a condition that could have been cleared Tuesday and was cleared Friday. On the Linden Street file, the underwriter's actual decision consumed five of fifty-one days. Conditions consumed nineteen. And eleven of those nineteen days were a file sitting perfectly still, fully documented, waiting for a closing date that was still two weeks away — which is precisely the window in which the rate lock expired and the borrowers walked into a furniture store.

Learn the sequence and you can predict a file. Learn where the sequence stalls and you can compress one, which is the only competitive advantage in this business that a lower rate cannot beat.

In this chapter, you will learn to:

• Name the seven stages of a purchase file and the event that ends each one
• Distinguish lead, pre-qualification, pre-approval, and application — and say what legally counts as an application
• Describe what processing actually does, and find the two days where the calendar is won
• Explain what underwriting decides and why almost every approval is conditional
• Define turn time, measure it two ways, and know which one your partners use
• Name the one person who owes the next action on any file, at any moment
• Read a pipeline report and identify the file that needs today's first call

Learning Paths

🎓 Exam — §6.2 is the heavily testable section: the six items that constitute an application, and the pre-qualification/pre-approval distinction. §6.6's disclosure timing is tested too, though Chapter 22 owns the mechanics. 🏠 New LO — §6.3, §6.5, and §6.8. These three sections are your first ninety days. Read §6.5 twice; conditions are where new originators lose files they had already won. 🤝 Partner — §6.7 and §6.8. If you are an agent, §6.7 explains why the lender's "turn time" and your "days to close" are different numbers, and §6.8 tells you who to call when a file is quiet. 📊 Operations — §6.1, §6.7, and §6.9. The stage definitions in §6.1 are the ones your reports depend on; §6.7 is measurement discipline; §6.9 is the daily management artifact.

6.1 The pipeline as a physical thing

Everyone in this business says "pipeline," and almost nobody stops to notice that it is a plumbing metaphor doing real analytical work.

Your pipeline is the set of loan files you currently have in process, from first contact through funding. That is the operational definition, and it is the one this chapter uses. (Loan officers also use "pipeline" loosely to mean "future business" — the leads, the agents who might send something, the pre-approvals that have not found a house. §6.9 explains why mixing those two meanings on one report is the most common way a loan officer lies to themselves about their income.)

The metaphor is worth taking seriously, because pipes have four properties and so does this:

A pipe has a lag. What comes out the end today went in weeks ago. The loans funding this month were sold last month. If your phone stops ringing today, your income does not fall today — it falls in six weeks, at which point you have already lost six weeks you could have spent fixing it. This single property is the reason Chapters 38 and 39 exist, and it is why experienced originators panic at quiet weeks that look, to everyone else, like a light workload.

A pipe has capacity. You can carry a certain number of files at the standard of care you want to provide, and the number is smaller than you think. Past that point, the constraint is not your effort; it is your attention. Files do not fail loudly when you are over capacity. They fail by going quiet.

A pipe has pressure at the joints. Every handoff — you to processing, processing to underwriting, underwriting back to you, you to the closer — is a place where a file can sit with nobody's name on it. §6.8 is entirely about those joints.

And a pipe blocks. Not evenly. One stage backs up and everything behind it stops. On a purchase file, that stage is almost always the same one, and it is not the one people blame.

The seven stages

Here is the whole process, with the Linden Street file's real days on it. Study the fourth column — what ends the stage — harder than the others. A stage is defined by its exit event, not by its activity, and a shop that cannot say precisely what ends a stage cannot measure anything.

THE PIPELINE — seven stages of a purchase file                 [the Linden Street file]
LEGEND: the day numbers are this file's actual calendar (day 0 = the agent's 8:40 call).
        Percentages are that stage's share of the file's 51 days.

  #  STAGE            OWNER            DAYS        SPAN  %      WHAT ENDS THE STAGE
  ──────────────────────────────────────────────────────────────────────────────────────
  1  PROSPECT/LEAD    loan officer     day  0 → 5    5   9.8%   an application is taken
  2  APPLICATION      loan officer     day  5 → 7    2   3.9%   the file is turned over
                                                                 with third-party orders out
  3  PROCESSING       processor        day  7 → 23  16  31.4%   submission to underwriting
  4  UNDERWRITING     underwriter      day 23 → 28   5   9.8%   a decision is issued
  5  CONDITIONS       shared           day 28 → 47  19  37.3%   CLEAR TO CLOSE
  6  CLOSING          closer           day 47 → 51   4   7.8%   the borrowers sign
  7  FUNDING/RECORD   closing agent    day 51        —    —     the wire lands; documents record
  ──────────────────────────────────────────────────────────────────────────────────────
                                                    51 100.0%

  ┌──────────┐ ┌──────────┐ ┌────────────┐ ┌──────────────┐ ┌─────────┐ ┌─────────┐
  │  LEAD    │→│   APP    │→│ PROCESSING │→│ UNDERWRITING │→│ CLOSING │→│ FUNDING │
  └──────────┘ └──────────┘ └────────────┘ └──────┬───────┘ └─────────┘ └─────────┘
       LO           LO         processor          │  underwriter  closer   closing agent
                                                  ↓
                                    ┌─────────────────────────┐
                                    │      CONDITIONS         │  ←─────┐
                                    │   (the "stip sheet")    │        │ every re-submission
                                    │   19 of 51 days here    │ ───────┘ goes back in the queue
                                    └───────────┬─────────────┘
                                                ↓
                                       CLEAR TO CLOSE (CTC)

Walk that diagram once, slowly, because three things in it are counterintuitive.

First, the arrow that loops. Six of the seven stages are a line: the file moves forward and does not come back. Stage 5 is a loop. The underwriter issues conditions; you and the processor clear them; the file goes back to the underwriter; the underwriter reviews what came back and may issue more. Each pass through that loop costs queue time on both ends. This is the only stage in the process whose duration is not bounded by anything except how many times you go around, which is why it consumed 37.3% of this file's calendar and why Chapter 19 is devoted to it.

Second, notice how small underwriting is. Five days. Every borrower who has ever waited on a loan believes "underwriting" is where the time goes, and every loan officer who has ever been asked "what's taking so long?" has been tempted to let them believe it, because it is a comfortable answer that blames someone the borrower will never meet. It is 9.8% of the file. The honest answer is harder and better for your business.

Third, notice who owns stage 5. Nobody. "Shared" is a euphemism. The underwriter has issued a list and moved on to another file. The processor is chasing some of the items. You are chasing others. The borrower owes several. The title company owes one. There is no single throat to choke and no queue that automatically advances, which is exactly why files rot there.

The two clocks

A purchase file runs on two clocks at once, and they are not synchronized.

The lender's clock starts at application and is measured in stages: how long from submission to decision, from condition receipt to review, from clear to close to docs. This is the clock your employer reports on and the one your operations team is graded on.

The contract's clock started before you existed on this file. On Linden Street, the buyers wrote their offer the night of day 0 and named a closing date forty-five days out — day 45. The offer was accepted on day 4. From the moment the contract was executed, the transaction had 41 days, not 45. Nobody sent a memo about the four missing days. They were simply gone, and they were gone before the loan officer had a single verified fact in hand.

This gap between the two clocks is the most reliable source of surprise in residential lending, and the discipline it demands is simple: the first thing you do with an executed contract is not read the price. It is read the dates, subtract today, and say the remaining number out loud.

6.2 Lead to application

The front of the pipeline has four distinct states, and new loan officers blur them constantly. The blurring is not a vocabulary problem. Each state carries different obligations, different risk, and a different answer to the question "what have I actually promised?"

Prospect / lead

A prospect — used interchangeably with lead — is a person who has expressed interest in financing and about whom you have no verified information. That is the whole definition. A referral from an agent is a lead. A form fill from your website is a lead. Your neighbor mentioning at a barbecue that they might move is a lead.

A lead is not a file. It has no loan number, no disclosures, and no obligations attached to it. It also has no value until it converts, which is why Chapter 7 spends a chapter on where leads come from and Chapter 38 turns conversion into arithmetic.

The Linden Street file was a lead for exactly one phone call. At 8:40 on day 0, the buyer's agent — someone the loan officer had closed four prior files with — called about two clients writing an offer that night.

Pre-qualification

A pre-qualification is an estimate of what a borrower can likely borrow, based on information the borrower has stated and you have not verified. It is a conversation with arithmetic in it.

Pre-qualification is fast, free, and worth precisely what the underlying facts are worth — which is nothing, if the borrower has told you their gross income including a bonus that has no two-year history, or has forgotten the student loan that went into repayment last month. Pre-qualification is not worthless; it is the right tool early, and Chapter 8 teaches the conversation properly. But it must be labeled honestly to the borrower and to the agent, every time.

Pre-approval

A pre-approval is a written statement that, based on verified information — a pulled credit report at minimum, and in a serious shop verified income and assets as well — a specified borrower qualifies for a specified loan amount and program, subject to a property and to conditions.

The distance between pre-qualification and pre-approval is the distance between "they told me" and "I looked." Chapter 1 named the failure mode already: a pre-approval letter is a statement of fact made to a third party who will rely on it, and you should be able to point at the document supporting every fact in it. Chapter 8 draws the line with a scalpel; for our purposes here, the process point is that the pre-approval is the first artifact in the pipeline that anybody outside your company relies on.

On Linden Street the pre-approval was issued on day 1, after a discovery call and a credit pull. That is fast. It is also the only stage in the entire fifty-one days that finished ahead of schedule, which tells you something about where loan officers put their energy.

Application

An application is the formal request for a specific loan on a specific property, and — unlike the three states above — it is a term with a legal definition that triggers obligations the moment it exists.

Under the integrated disclosure rules, an application consists of six pieces of information:

THE SIX ITEMS THAT MAKE AN APPLICATION

  1  the consumer's name
  2  the consumer's income
  3  the consumer's Social Security number (to obtain a credit report)
  4  the property address
  5  an estimate of the value of the property
  6  the mortgage loan amount sought
  ─────────────────────────────────────────────────────────────────────
  When all six exist, you have an application — whether or not you meant to
  take one, whether or not anything is signed, and whether or not you have
  opened the loan origination system.

That last line is the part that costs people money. An application is not an event you schedule. It is a threshold you cross, sometimes in the middle of a phone call, and the disclosure clock starts whether or not you noticed.

🎓 NMLS Exam Watch

Two items from this section are near-certainties on the SAFE MLO test.

The six items. Memorize them. The classic trap is a seventh distractor — older versions of the definition included a catch-all along the lines of "any other information the originator deems necessary," and the current integrated-disclosure definition does not. If an answer choice adds a seventh element, look hard at it.

Pre-qualification versus pre-approval. The exam tests the basis, not the paperwork: pre-qualification rests on stated information; pre-approval rests on verified information. A stem that says "the loan originator reviewed the borrower's credit report and paystubs" is describing a pre-approval no matter what the letter is titled.

A third, quieter distinction worth knowing: neither one is a commitment to lend. A commitment is an underwriting decision on a complete file with a property. The exam likes stems that describe a pre-approval and offer "commitment" as a plausible answer.

On Linden Street, the offer was accepted and the contract executed on day 4. The full application was taken on day 5, the day after there was a property to apply for. That ordering is the normal one on a purchase, and it is worth stating explicitly because it confuses new originators: on a purchase, the pre-approval comes first and the application comes second, because item 4 on that list — the property address — does not exist until there is a contract.

⚖️ Compliance Check

Crossing the application threshold starts several clocks at once. The three that matter most in the first week:

• The Loan Estimate. Once you have an application, a Loan Estimate (LE) must be delivered or placed in the mail within a defined, short window — three business days — and there is a second deadline tied to consummation. On the Linden Street file the application was taken day 5 and the LE issued inside that window.
• The fee restriction. Before the consumer has received the LE and indicated an intent to proceed, a lender generally may not impose fees other than a bona fide and reasonable fee for obtaining a credit report. This is the rule that makes "just put a card down to get started" a dangerous habit.
• Notice of action taken. The Equal Credit Opportunity Act (ECOA) and Regulation B require notice of action taken on an application within a defined period — thirty days for a completed application — and that obligation attaches to applications, including ones you would rather think of as informal.

Chapter 22 works the TILA-RESPA Integrated Disclosure (TRID) timing rules in full and Chapter 25 works ECOA. Requirements change and state law varies; verify current rules with your compliance department and your regulator before you rely on any timeline in this book.

What the application stage actually produces

Stage 2 is short — two days on this file — and it is the stage new originators most often treat as paperwork. It is not paperwork. It is the stage that determines whether stages 3 through 6 are possible.

By the end of stage 2 you should have produced four things:

1. A complete, accurate Uniform Residential Loan Application (Form 1003, the URLA) — accurate meaning the income figure on it is the one you calculated from documents, not the one the borrower said on the phone.
2. The initial disclosure package out and acknowledged, including the LE.
3. A run through automated underwriting — on Linden Street, day 6, returning Approve/Eligible. Chapter 15 explains what that recommendation is and, more importantly, what it is not.
4. Every third-party order placed — which on this file happened day 7 and is the subject of the next section.

Item 1 is where the file's whole future is decided. An application built on a number the borrower recited is a file that will be re-underwritten at day 25 with a different income, and everything downstream — the pre-approval letter, the ratios, possibly the program — moves with it.

6.3 What processing actually does

Processing is the assembly of a complete, internally consistent, submittable loan file: ordering the third-party reports, collecting and reviewing the borrower's documentation, verifying that the pieces agree with each other, and delivering the package to underwriting.

Chapter 1 said the processor has no approval authority, and that is true and important. It also tends to leave new loan officers with the impression that processing is clerical. It is not. A good processor is doing three cognitively distinct jobs at once, and the second one is the one that saves your files.

Job one: the orders

Third-party work is anything the lender must buy from someone outside the company. On a typical purchase file that means:

Look at the third column. Almost none of that elapsed time is under your control, and all of it starts when the order is placed. On the Linden Street file the appraisal, title, and employment verifications all went out on day 7. The appraisal came back day 16 — nine days. The title commitment came back day 19 — twelve days.

Twelve of the file's fifty-one days were spent waiting on work that had already been ordered. There is nothing wrong with that. It is the cost of doing business, and no amount of energy compresses an appraiser's schedule. What there is something very wrong with is placing that order on day 12 instead of day 7, which moves every subsequent date by five days and costs exactly as much as a five-day delay at the end of the file — except that nobody notices it, because the damage is invisible for a month.

⚠️ Where Deals Die

Serial ordering. The single most expensive habit in processing is placing third-party orders one at a time, in sequence, waiting for each to come back before starting the next.

The reasoning always sounds prudent. "Let's not order the appraisal until we know the income works — if this falls apart, I've cost them \$650." Or: "Let's wait for the title commitment before we send the VOE." Every one of those sentences trades the borrower's money against the transaction's calendar, and on a purchase with an executed contract the calendar is usually worth more.

Run the shape of it. Appraisal nine days, title twelve days, VOE five days. Ordered in parallel on day 7, the last one lands on day 19. Ordered serially — appraisal, then title, then VOE — the last one lands on day 33. Same work, same vendors, same fees: fourteen days, which on this file is the difference between closing on the contract date and not. (Illustrative; elapsed times vary by market and vendor.)

The disciplined version has three rules. Order everything you can on the day you have a complete application. Where an order genuinely should wait — an appraisal on a file with a real, identified risk of not proceeding — say so out loud to the borrower and the agent, in writing, with the date you will order it. And never let an order wait on a document you could collect in an hour.

Job two: reading for consistency

This is the job that is invisible in a job description and decisive in a file.

Every document that arrives has to agree with every other document. The paystub's year-to-date has to support the income you calculated. The bank statement's ending balance has to match the next statement's beginning balance. The employer on the VOE has to be the employer on the paystub, at the same start date. The property address on the appraisal has to be the property address on the contract, the title commitment, and the flood certification. The borrower's name has to be spelled the same way on the credit report, the driver's license, and the deed.

None of that is glamorous, and all of it is what an underwriter will do at day 23 if the processor did not do it at day 12. The difference between those two dates is the difference between a fix and a condition — and the difference between a condition and a re-submission, which is a trip back through the queue.

A specific example from this file: Borrower 2 is an outside sales representative whose qualifying income is base plus a twenty-four-month commission average. That structure means the file needs commission history documentation and a verification that the commission is likely to continue. A processor who notices on day 12 that the commission documentation covers only twenty months has saved the file. An underwriter who notices it on day 26 has issued a condition, and the file has lost a week.

Job three: the shelf life of documents

Loan documents expire. This surprises borrowers, who reasonably assume that a paystub proving they were employed in March still proves it in July, and it is one of the genuinely counterintuitive costs of a slow file.

Guidelines generally require that credit and income documentation be current as of the note date — commonly measured in a window of about four months for credit documents on a conventional existing-construction file, with tighter windows for items like paystubs and much tighter ones for a verbal verification of employment. These windows change and vary by agency, program, and lender overlay; verify the current requirement in the applicable guide and with your underwriting department before you rely on a number.

The operational consequence is what matters here: a file that takes an extra thirty days does not merely close late. It may have to be re-documented. New paystubs. New bank statements. Possibly a new credit report — which, on a file where the borrowers have since financed furniture, is not a neutral event. This is one of the mechanisms by which delay converts into risk rather than just inconvenience, and it is a large part of why §6.7 insists that turn time is a risk measure and not a vanity metric.

6.4 What underwriting actually does

Underwriting is the evaluation of a complete loan file against the applicable guidelines and lender overlays, resulting in a decision. That is the whole function, and the operative word is decision — underwriting is the only stage in the pipeline where anyone has authority to say yes or no.

The Linden Street file was submitted to underwriting on day 23 and received a decision on day 28. Five days, which in most markets is an unremarkable first-look turn time. What happened in those five days is worth understanding precisely, because the entire craft of file preparation is the craft of anticipating it.

The question the underwriter is actually asking

Not "do I like these people." Not "can they afford it," at least not in the way a borrower means the phrase. The underwriter is asking a narrower and stranger question:

Is this file what we will represent it to be when we sell it?

Chapter 1 traced the money: your employer funds the loan with borrowed money and sells it within weeks, under representations and warranties that can require a repurchase years later if the file was not what it claimed. The underwriter is the person who signs that claim. Every condition on every approval traces back to a specific sentence in a guide that a specific buyer of loans wrote, and the underwriter is not permitted to be persuaded out of it by a compelling story about a nice family.

That reframing is worth adopting early, because it changes how you talk to underwriting. "They can afford it, look at their reserves" is an argument to a person. "The guide permits reserves as a compensating factor at this ratio, here is the documentation, and here is the AUS finding" is an argument to a file. Only the second one works.

The four buckets

Underwriting evaluates a file across four categories, which the trade has called the four C's for about as long as there have been mortgages:

• Capacity — can the household make the payment? Income, employment stability, and the ratios.
• Credit — have they repaid obligations as agreed? Score, history, derogatory events.
• Capital — do they have the funds to close, and something left afterward? Down payment, sourcing, reserves.
• Collateral — is the property worth the loan, and is it acceptable security? Appraisal, condition, title.

Chapter 14 works the conventional rulebook behind these in detail, Chapter 15 covers automated underwriting, and Chapters 10 through 12 take credit, income, and assets one at a time. For process purposes, note something structural: the four buckets do not arrive at the same time. Credit exists on day 1. Income and assets exist as soon as the borrower produces documents. Collateral does not exist until the appraisal returns — day 16 on this file — and title does not exist until the commitment returns, day 19. A file cannot be submitted meaningfully until the slowest bucket has landed, which is why submission was day 23 and not day 12.

AUS is not an approval

On day 6, before a single verification had come back, the file was run through automated underwriting and returned Approve/Eligible.

This is the single most misunderstood artifact in origination, and the misunderstanding is the source of an enormous amount of misplaced borrower confidence. An automated underwriting system (AUS) evaluates the data you entered against the agency's guidelines and returns a recommendation plus a list of what must be verified to rely on it. It does not know whether the data is true. It has not seen a paystub. It cannot see the property, because on day 6 there is no appraisal.

An Approve/Eligible is genuinely valuable — it tells you the structure works and it defines the documentation you need. It is not permission to tell anybody the loan is approved. Chapter 15 is emphatic about this and so is every underwriter you will ever work with.

The three outcomes

A human underwriter's decision comes back in one of three forms:

A conditional approval — approved subject to a list — is not a hedge or a soft no. It is what an approval is. In ordinary practice a completely unconditional approval essentially does not exist, because some items, by design, cannot be satisfied until just before closing: an employment verification that must be current as of the note date cannot be obtained a month early.

New loan officers hear "approved with conditions" and relay "approved" to the borrower and the agent. That is the beginning of most of the hard conversations in this business.

🔍 Check Your Understanding

1. The AUS returned Approve/Eligible on day 6. The underwriter did not issue a decision until day
2. Name three facts that existed on day 28 that did not exist on day 6.
3. A file is submitted on day 23 and suspended on day 27 because the commission documentation covers twenty months instead of twenty-four. Whose failure was that, and on which day should it have been caught?
4. Your borrower says: "We're approved, right? The system said approved." Answer them in two sentences, without using the word "just."

(For 1: a verified appraisal establishing collateral value, a title commitment establishing lien position, and verified income and asset documentation replacing the figures that had been entered by hand. For 2: processing's, on roughly day 12, when the documentation arrived — see §6.3, job two.)

6.5 Conditions: the loop that eats the calendar

A condition — universally called a stipulation or, in conversation, a stip — is a specific item the underwriter requires before the conditional approval becomes a final approval. The list of them is the stip sheet or condition sheet, and it is the working document of the second half of a loan file.

Conditions are classified two ways, and you need both classifications in your head at once because they answer different questions.

By source — who can actually produce this?

• Borrower-supplied. Paystubs, letters of explanation, gift letters, signatures. You have real leverage here: you can call, text, and drive to a house.
• Third-party. A title release, an employer's verification, an insurance binder, a revised appraisal. You have influence, not leverage. Someone else's calendar governs.
• Lender-internal. A mortgage insurance certificate, a compliance test, a verbal verification of employment (VVOE) pulled by the closing department. You cannot touch these at all; you can only make sure nobody forgot them.

By timing — when must this be satisfied?

• Prior to document preparation (PTD). Must be cleared before closing documents can be drawn. These gate the clear to close.
• Prior to funding (PTF). Must be satisfied after documents but before the wire — typically dated items that must be current as of the note date.
• Prior to purchase. Investor-level items that arise after closing, when the loan is delivered and sold. These do not delay your closing; they can delay your employer's sale of the loan, and Chapter 28 explains why that matters to you anyway.

The reason both classifications matter is that they tell you two different things: the source tells you who to call, and the timing tells you whether calling today would even help. A PTF condition cannot be cleared early no matter how motivated everyone is. Recognizing that on day 28 instead of day 40 is the difference between a plan and a panic.

📄 Read the File

```text FIGURE 6.1 — "Eleven conditions" [the Linden Street file]

CONDITIONAL APPROVAL — issued day 28 Loan L-2214 · conventional 30-yr fixed · $365,750 · LTV 95.00% · rep score 706 AUS: Approve/Eligible (run day 6) Underwriter decision: APPROVED WITH CONDITIONS

# CONDITION SOURCE WHEN STATUS ───────────────────────────────────────────────────────────────────────────────── 1 Signed, dated initial URLA (Form 1003) plus borrower PTD cleared d29 complete initial disclosure package returned 2 Most recent 30 days' paystubs, both borrowers, borrower PTD cleared d29 current as of the note date 3 Written VOE, Borrower 2, confirming commission 3rd party PTD cleared d30 continuity, plus most recent commission statement 4 Signed IRS Form 4506-C, both borrowers borrower PTD cleared d29 5 Letter of explanation and source documentation borrower PTD cleared d33 for the $4,900 deposit, Borrower 2's checking 6 Gift letter for the $10,000 gift, signed by borrower PTD cleared d29 donor and recipients, plus evidence of transfer 7 Title commitment free of the prior owner's 3rd party PTD cleared d30 mechanic's lien shown at Schedule B-II 8 Evidence of homeowners insurance, one year paid, borrower PTD cleared d30 lender named as mortgagee 9 Mortgage insurance certificate at the approved lender PTD cleared d33 coverage and factor 10 Verbal VOE, both borrowers, within the required lender PTF OPEN window before the note date 11 Pre-closing credit refresh / undisclosed-debt lender PTF OPEN report; DTI may not exceed the approved ratio ───────────────────────────────────────────────────────────────────────────────── 11 conditions: 6 borrower · 2 third party · 3 lender 9 PTD · 2 PTF PTD = prior to document preparation PTF = prior to funding

THE DOCUMENT The underwriter's conditional approval and condition list, issued day 28, five days after submission. One page in the loan origination system; this is the version the loan officer and processor work from. THE CONTEXT A $385,000 purchase, 5% down, conventional, 95% LTV, representative score 706. Rate locked day 12. Contract closing date is day 45 — seventeen days from the date on this document. WHAT IT SHOWS An approval, subject to eleven items. Nine are prior-to-document and therefore gate the clear to close; two are prior-to-funding and by design cannot be satisfied early. Six of the eleven are things the borrowers themselves can produce. Two belong to a title company and an employer. Three belong to the lender's own departments. Every one of the nine PTD conditions cleared between day 29 and day 33 — five calendar days, but only THREE business days, because days 31 and 32 were a weekend. Nine items, three working days, and nobody had to be heroic about it. WHAT IT DOESN'T It does not tell you how long any of this will take, because a condition list has no dates on it. It does not show what happens after the last PTD item clears on day 33 — which on this file is nothing, for eleven days. And condition 11 does not say what it is really for: it is the tripwire that will detect a debt the borrowers have not incurred yet. THE DECISION On day 28: sort the list by source, send the borrowers ONE message containing only their six items in plain language with a single due date, call the title company about item 7 the same hour, and put items 10 and 11 on the calendar for the closing week — then ask operations for the earliest date this file can be cleared to close if everything lands. THE LESSON A condition list is not a to-do list; it is three to-do lists with three different owners and two different deadlines, printed in one column. The loan officer's job is to un-merge it before anyone reads it. ```

Constructed for this book's running file. Real condition lists vary enormously by lender and program; the taxonomy — source and timing — is what transfers.

Why the loop eats the calendar

Nineteen days. Day 28 to day 47. Here is where they went:

Read the bolded row again — twice, because it contains two different disasters. Eleven of the nineteen days in the conditions stage — 21.6% of the entire fifty-one-day file — were a fully documented loan sitting perfectly still.

Nobody was blocked. Nobody was slow. Nine of eleven conditions were satisfied by day 33. The two that remained were prior-to-funding items that had to be pulled close to the note date, and the closing was scheduled for day 45, so there was no reason to pull them yet. Every individual decision was defensible. The result was eleven days of slack that everyone treated as free.

It was not free, and it billed twice.

The first bill arrived on day 42, and nobody was watching for it. The rate lock taken on day 12 was a 30-day lock, and 12 + 30 = 42. It expired in the middle of the dead window, three days before a closing that was not going to happen anyway, and a 15-day extension was purchased at 0.250 point — \$914.38. Not one person on this file did anything wrong on day 42. They simply paid \$914.38 for days that had already been spent doing nothing. §6.7 prices it and Chapter 30 takes apart the day-12 decision that made a 30-day lock the wrong instrument for a day-45 contract.

The second bill arrived on day 44, and it arrived on a Friday. On day 41, in the middle of the same window, the borrowers financed \$5,200 of furniture at \$611 a month. On day 44 the credit refresh in condition 11 found it, the back-end debt-to-income ratio jumped from 42.66% to 48.48%, the approval's ratio condition was blown, and the day-45 closing date — a Saturday — passed with the file dead. The borrowers paid the account off online on day 46, a Sunday, because that was the only channel open to anybody that weekend; the payoff was documented and the findings re-run on day 47, the next business day. Chapter 19 works the resolution in full — it is that chapter's central worked example, and the details of how a blown DTI gets cleared belong there.

The process lesson is the one this chapter owns, and it is the most important sentence in the chapter:

Slack in a pipeline is not spare time. It is exposure.

Every day a file sits between "documented" and "closed" is a day in which a borrower can open a credit account, an employer can restructure, a paystub can go stale, a hurricane can suspend insurance binding, and a rate lock can expire. The file that closes on day 38 and the file that closes on day 51 are not the same file with different dates. The second one has been exposed to thirteen more days of the world.

The counterfactual on this file is uncomfortable and instructive, and it now settles both bills at once. Had the team pushed for a clear to close on day 35 and asked the seller to move the closing to day 38, two things follow arithmetically. The credit refresh in condition 11 would have been pulled on roughly day 37 — four days before the furniture was financed — so the condition would have cleared and the crisis would not have happened. And day 38 is inside day 42, so the lock would never have expired and the \$914.38 would never have been spent. Neither outcome requires anyone to have managed anything better. Both follow from the window in which trouble could happen not existing.

Which does not mean "always close early" is the lesson. Sellers have calendars, movers are booked, and you cannot accelerate a transaction unilaterally. The transferable rules are narrower and they both belong to the loan officer:

1. Drive to clear to close, not to the closing date. These are different targets. The closing date belongs to the contract. The CTC date belongs to you, and every day you pull it forward is a day of exposure removed — even if the closing itself does not move.
2. Say the furniture sentence more than once. "Do not open credit, do not finance anything, do not change jobs, do not move money" is the cheapest risk control in origination and it has a half-life of about three weeks. Say it at application. Say it again when the approval comes in. Say it again the day before the refresh is pulled. Chapter 12 explains what the underwriter is actually looking for; the process point is simply frequency.

6.6 Clear to close, closing, funding

Three words that borrowers use interchangeably, that mean three different events on three different days, and that involve three different sets of hands.

Clear to close

Clear to close (CTC) means the underwriter has signed off on every prior-to-document condition and has authorized the closing department to prepare the closing documents.

It does not mean you can close today. It means the document process may begin. On the Linden Street file, CTC came on day 47 and the closing was day 51 — four days later, and not one of those four days was slack.

CTC is nevertheless the emotional finish line for everyone on a file, and it is the single most useful milestone to communicate to a borrower and an agent, because it is the first moment when the remaining steps are procedural rather than evaluative. Before CTC, the answer to "will this close?" is a judgment. After CTC, it is a schedule — with two real exceptions, which is why the two prior-to-funding conditions still sat open.

Closing

Closing — the event lawyers call consummation — is the signing: the borrowers execute the note and the security instrument and the rest of the closing package, the seller executes the deed, and funds are collected.

Who conducts it depends entirely on where the property is. In some states a title or escrow company runs the signing; in others an attorney must. In some markets buyer and seller sit at the same table; in others they never meet, sign days apart, and the escrow officer assembles the pieces. Chapter 23 covers closing day properly, including what actually happens in the room. Chapter 1's distinction is the one to hold onto here: the closer works for the lender; the closing agent works for the transaction.

Funding

Funding is the lender's disbursement of the loan proceeds — the wire. It is a separate event from signing, and in some states it happens the same day and in others it does not.

Practice varies by state and by lender. In "wet funding" markets, funds disburse at or immediately after signing. In "dry funding" markets, the lender reviews the executed package before releasing funds, which can put a day or more between the signing and the money. Neither is better; they are different legal and market traditions. Verify your state's practice and your lender's policy — this is exactly the kind of thing that varies and that nobody tells a new loan officer until the first time an agent asks when the seller gets paid.

Then the documents record — the security instrument goes into the county land records, creating the lien Chapter 1 described, and the deed transfers title. On the Linden Street file, day 51 was closing, funding, and recording.

THE LAST FIVE DAYS                                            [the Linden Street file]

  day 47   CLEAR TO CLOSE        underwriter releases the file to the closing dept.
              │                  9 PTD conditions satisfied; 2 PTF conditions remain
              ↓
  day 48   CLOSING DISCLOSURE    closer prepares figures with the settlement agent;
    (Tue)     │                  the CD is issued to and received by the borrowers, and
              │                  the required waiting period before consummation runs
              ↓
  day 51   CLOSING               borrowers sign the note and security instrument
    (Fri)  FUNDING               the lender wires; the settlement agent disburses
           RECORDING             deed and security instrument go to the county
              │
              ↓
           the loan exists. Servicing, delivery, and sale follow (Ch. 23, 28).

  Tuesday receipt -> Wednesday, Thursday, Friday. Three business days, exactly.
  These are the two days in this file whose weekday you must know, because the
  disclosure rule counts in business days and the contract does not.

⚖️ Compliance Check

The last few days of a file are the most heavily regulated stretch in origination, and the rules are timing rules — which means they convert directly into calendar.

• The Closing Disclosure (CD) must be received by the consumer a defined number of business days before consummation — three, under the integrated disclosure rule. The clock runs from receipt, not from the moment your system generated the document, and the definition of "business day" for this purpose is technical: it is not the same definition used elsewhere in the regulation. Chapter 22 works the counting rules and the delivery-versus-receipt presumptions in full.
• Certain changes after the CD restart the waiting period. A change in the annual percentage rate (APR) beyond tolerance, a change in loan product, or the addition of a prepayment penalty trigger a new three-business-day period. Most other changes require a corrected CD but not a new waiting period. Knowing which is which is what keeps a closing on the calendar.
• Revised Loan Estimates require a valid changed circumstance and have their own timing. You cannot re-disclose because you would like to.

The operational translation: build the waiting period into your closing schedule from day one, and treat the CD date as a hard milestone rather than a formality. The Linden Street file cleared to close on day 47 and issued its CD on day 48 for a day-51 closing — exactly three business days, with nothing in reserve. It worked. Had one figure moved beyond tolerance on day 49, it would not have.

Requirements change, interpretations evolve, and state law adds its own layers. Verify current rules with your compliance department and your regulator before relying on any timeline here.

⚠️ Where Deals Die

Clear to close is not the end of underwriting risk, and everyone behaves as though it is.

The two prior-to-funding conditions on the Linden Street file existed precisely because the lender does not trust the world to hold still between CTC and the wire. A borrower who buys a car on day 49, accepts a new job on day 50, or moves \$20,000 between accounts "to make the wire easier" has undone the file after everyone stopped watching — including you, because CTC feels like the end.

The mechanism is always attention. Between clear to close and funding, the loan officer's mental model of the file changes from at risk to done, and the borrower's changes from anxious to celebrating. Both changes are premature by four days, and the second one is what buys the couch.

The disciplined version is two sentences, said on the day the CTC comes in — not emailed, said: "Nothing changes until the money moves. No new credit, no new job, no moving money between accounts." And: "Wire instructions never change by email. If you get an email changing them, it is fraud. Call me at the number you already have, not the number in the email."

That second sentence is a thirty-second conversation that prevents the most financially devastating thing that can happen to a first-time buyer. Chapter 27 covers closing-wire fraud in full; there is no reason to wait for it.

One last calendar fact about this stretch, and it is the kind of thing that makes an experienced loan officer's answer to "why did that take six days?" sound so different from a new one's. The problem surfaced on a Friday and the contract's closing date was the next day, a Saturday.

Look at what that did. The credit refresh fired on day 44, a Friday. Day 45 — the date two families, an agent, and a moving company had been building toward — was a Saturday, and it simply went by. Day 46 was a Sunday. Of the three days between discovering the problem and clearing it, two were a weekend, and the only action anybody could take in them was the one that required no counterparty: the borrowers paid the furniture account off online on day 46. Everything else — documenting the payoff, re-running the findings, getting a human underwriter to look again — waited for Monday.

This is the divergence that makes turn time so easy to argue about and so hard to manage. A file is measured in calendar days, because that is what a borrower lives and what a contract counts. It is consumed in business days, because that is when anyone is at a desk. The two drift apart constantly, and they drift furthest at exactly the moment a file is in trouble — because a problem found late on a Friday does not get worked until Monday no matter how urgent it is.

So: when you promise a date, count the weekends. When you set a closing date, look at what day of the week it falls on before you agree to it. And when a condition is going to need somebody else's signature, find that out on a Wednesday, not at four o'clock on a Friday.

6.7 Turn times, and why they are the number your partners judge you on

Turn time is the elapsed time between two defined milestones on a loan file, measured the same way every time.

That definition has three load-bearing parts, and every argument about turn times is really an argument about one of them. Elapsed — calendar days or business days, and you must say which. Between two defined milestones — which two, exactly. Measured the same way every time — which is where most reported turn times quietly fail.

There is no such thing as "the" turn time

There are at least six turn times on any file, and they are not interchangeable:

Look at the last three rows. Same file, three legitimate numbers, and they differ by four days. If your lender advertises a twenty-one-day close and your agent experiences a forty-seven-day transaction, both of you may be telling the truth about different measurements — and the agent is the one who will describe the experience to their next client.

Your partners use contract-to-close. They are not measuring your underwriting department. They are measuring the interval between the moment their client's offer was accepted and the moment their client got keys, because that is the interval they lived through and the one they promised. Chapter 7 uses this number when it talks about earning referral relationships, and Chapter 38 turns it into a business development argument. Own it now: when an agent asks "how fast do you close," the honest answer is a contract-to-close number, and it should be the number you hit nine times out of ten, not your best file.

There is a widely quoted industry benchmark for average days to close, published monthly for years in the trade press from lender-system data. Use it as a directional reference and look up the current figure rather than quoting a remembered one — it moves with rate cycles, volume, and channel mix, and a stale benchmark quoted confidently is worse than no benchmark.

Measurement discipline: three ways the number lies

Moving the start line. The most common distortion in the business is beginning the count at "complete application" or "file submitted" rather than at the borrower's first contact or the contract date. It shortens the number, it is defensible in a report, and it describes an experience nobody had.

Averaging away the tail. A pipeline's mean is dragged by the two files that took ninety days. Report the median and the 90th percentile together. The median is your typical experience; the tail is your reputation. An agent who has one file blow up remembers that file, not your average.

Excluding fallout. A shop that reports turn times only on loans that closed is describing survivors. The loans that died at day forty took the longest and are not in the denominator.

What turn time does not measure

Speed is not quality, and the two are not even reliably correlated. A shop with a fast turn time and a high fallout rate is not fast — it is triaging, and the files it drops are files somebody else would have closed. A shop that reaches CTC quickly by pushing marginal items to prior-to-funding has moved risk to the day when it is most expensive to discover.

Turn time is best understood as a risk measure, for the reason §6.5 established: every extra day is a day of exposure to a world that does not hold still. That framing is more useful than the sales-oriented one, and it happens to be the one that produces better sales results anyway.

🧮 Run the Numbers

Where fifty-one days went, and what the last six cost.

Decompose the file to a finer grain than §6.1's seven stages. The cumulative column is the file's actual day number, which is a useful check that nothing is double-counted.

Span	What was happening	Days	Cum.	% of file
day 0 → 5	lead, pre-approval, contract, application	5	5	9.8%
day 5 → 7	application worked; AUS run; orders placed	2	7	3.9%
day 7 → 19	waiting on appraisal (d16) and title (d19)	12	19	23.5%
day 19 → 23	file assembled, reviewed, submitted	4	23	7.8%
day 23 → 28	underwriting first look	5	28	9.8%
day 28 → 47	conditions	19	47	37.3%
day 47 → 51	CTC, CD, closing, funding	4	51	7.8%
Total		51		100.0%

(Percentages are rounded to one decimal.)

The headline: underwriting — the stage everyone blames — was 9.8%. Conditions were 37.3%, and eleven of those nineteen days (21.6% of the whole file) were a documented file sitting still.

Now price the overrun. The contract closing date was day 45; the file funded day 51, six days late. Two hard-dollar consequences, running in opposite directions:

• The lock extension — and note when it was bought. The lock taken on day 12 was a 30-day lock, so it expired on day 42: three days short of the file's own scheduled closing, from the moment it was taken. Chapter 30 takes apart that day-12 decision. On day 42 it expired and a 15-day extension was purchased at 0.250 point: $$\$365{,}750 \times 0.00250 = \$914.38$$ or $\$914.38 \div 15 = \$60.96$ per day. Two days later the credit refresh fired and fifteen days stopped looking generous.
• Prepaid interest, which moved the other way. Per-diem interest on this loan is $\$365{,}750 \times 0.06625 \div 365 = \$66.3861$ per day. Closing day 51 left 8 days remaining in the month: $8 \times \$66.3861 = \$531.09$, the figure that appears in the borrowers' cash to close. Closing six days earlier, on day 45, would have left 14 days: $14 \times \$66.3861 = \$929.41$. Closing late therefore reduced prepaid interest by $\$929.41 - \$531.09 = \$398.32$.

Net cash cost of the six-day overrun:

$$\$914.38 - \$398.32 = \$516.06$$

(The \$929.41 and \$398.32 figures are a counterfactual computed here for teaching; the file's actual prepaid interest is \$531.09.)

Put \$516.06 in scale, because a number with no yardstick teaches nothing. Chapter 1 priced a quarter point on this loan at \$60.14 a month — the difference the borrowers were shopping for. \$516.06 is roughly eight and a half months of that advantage, spent on nothing at all. It is also about 28% of the \$1,828.75 discount point they paid to buy their rate down: better than a quarter of the money they spent purchasing the rate, handed back to keep the rate they had already purchased.

Now be honest about the \$398.32, because it is the weakest number on this page. It is not a saving. It is a timing shift — the borrowers prepaid six fewer days of interest because they owned the house six fewer days. Measured as cash on a settlement statement, the overrun cost \$516.06. Measured as economic cost, it is closer to the full **\$914.38**.

And the cash was still the cheap part. Set \$516.06 beside what those six days actually produced: a back-end ratio at 48.48%, a closing date gone, a seller who had to be asked for patience, a referral relationship with four prior closings behind it put at risk, and two first-time buyers who spent a weekend believing they had lost the house. Not one of those items has a dollar figure, and every one of them is larger than \$516.06.

Delay bills you in dollars and charges you in exposure. The invoice you can see is never the expensive one.

6.8 Who owns the file at each stage

Here is a rule that will do more for your first year than any amount of product knowledge:

At every moment, on every file, exactly one person owes the next action. If you cannot name that person and what they owe, the file is stalled — and it has probably been stalled longer than you think.

Files do not usually die from a decision. They die from ambiguity: the borrower thinks the processor has it, the processor is waiting on the underwriter, the underwriter reviewed it Tuesday and issued a condition nobody opened, and four days evaporate with everyone believing the file is moving.

Ownership versus possession

Two different ideas that get conflated.

Possession is who is physically working the file right now — whose queue it is in. Ownership is who is accountable for the file moving, which is always, on every file, at every stage, you. The loan officer is the only party who is present for all seven stages. The processor is not on the file at day 2. The underwriter is not on it at day 15. The closer appears at day 47. You are there the entire time, and you are the only one the borrower chose.

That is not a motivational point; it is an operational one. If the appraiser is late, the underwriter does not know and the borrower does not know. You know, or nobody does.

WHO OWES THE NEXT ACTION                                       [the Linden Street file]

  STAGE / DAYS         POSSESSION          THE LO OWES                   ESCALATE TO
  ────────────────────────────────────────────────────────────────────────────────────
  lead      d0–5       loan officer        everything                    —
  app       d5–7       loan officer        1003 accuracy, disclosures,   —
                                           AUS, orders placed
  processing d7–23     processor           borrower docs, the agent's    processing mgr
                                           weekly update, order chasing
  underwriting d23–28  underwriter         nothing but visibility        UW manager
                                           (do NOT "check in" daily)
  conditions d28–47    SHARED — the risk   un-merging the list; the 6    UW manager for
                       zone                borrower items; a date for    condition disputes;
                                           the third-party items;        mgmt for aging;
                                           WATCHING THE LOCK (d42)       lock desk for d42
  closing   d47–51     closer + settlement CD review, wire-fraud         closing manager
                       agent               warning, borrower prep
  funding   d51        closing agent       confirmation, and the         —
                                           first post-close call
  ────────────────────────────────────────────────────────────────────────────────────
  Ownership never moves. Possession moves six times. Every one of those six
  handoffs is a place a file can go quiet.

Two lines in that table deserve comment.

"Do not check in daily" during underwriting. New loan officers, anxious and eager to look attentive, message the underwriter every morning. It does not accelerate anything, it consumes the attention of the one person whose attention you actually need, and it damages a relationship you will need on a harder file. Know your shop's published turn time, add a day, and inquire then — with a specific question, not "any update?"

The escalation column. Have it filled in before you need it. The moment to find out who the underwriting manager is, is not the moment you need them. Chapter 39 builds the daily discipline around this; the habit to start now is knowing, for every stage, whose desk is one level up.

📞 On the Phone

The agent's status call, day 36. This is the most common conversation in the business and most loan officers handle it badly in the same three ways.

Agent: "Hey — just checking on Linden Street. Everything good?"

The bad answer: "Yep, all good!" It is content-free, it is unfalsifiable, and if anything goes wrong later the agent will remember that you said everything was good on day 36.

The other bad answer: "It's in underwriting." It is not, it has not been since day 28, and the agent cannot tell the difference — which is exactly why saying it is corrosive. Vague status becomes a habit, and the habit is how you end up telling an agent on day 44 that a file you described as fine has a problem.

The third bad answer: a four-minute recital of all eleven conditions. The agent does not want your file; they want to know whether to worry and whether to warn their client.

What actually works — four facts, in this order:

"Nine of eleven conditions cleared as of day 33. The two left are the employment verification and the credit refresh, and those get pulled the week of closing by design — that's normal, not a problem. Right now the file is waiting on the calendar, not on a person, and I don't love that, because our rate lock expires on day 42. So I'm pushing for a clear to close ahead of the date rather than on it, and I'll call you the day it happens. One thing you can help with: remind them not to buy anything on credit until we fund. Not a car, not a couch, not a store card."

Where it is, who owes the next action, when it changes, and one specific thing they can do. Every good status call has those four elements, and none of them is "everything's good."

Price that call. The last sentence, said on day 36, would have been worth the closing date, the 48.48% ratio, and the week that followed. And the clause before it — our lock expires on day 42 — is the one nobody on this file said out loud to anybody, which is why \$914.38 went out the door on a day when everyone believed the file was fine. A status update that does not name the next hard date is not a status update. It is reassurance.

6.9 Reading a pipeline report

The loan origination system (LOS) is the software of record for a mortgage file: the application data, the documents, the conditions, the disclosures, the audit trail, and the dates. Chapter 36 covers the technology stack; what matters here is the artifact the LOS produces every morning.

A milestone is a defined, dated event in the LOS that a file has either reached or not: application taken, submitted to underwriting, approved with conditions, clear to close, docs out, funded. Milestones are the vocabulary of every report, every turn-time calculation, and every conversation with your operations manager. They are also, quietly, a definitional minefield: two shops that both report "days to close" may be starting the count at different milestones, which is §6.7's measurement problem in institutional form.

The pipeline report is a list of your active files with their milestones and dates. It is the single most useful piece of paper in a loan officer's day, and almost nobody reads it correctly, because the eye goes to the loan amounts.

📄 Read the File

```text FIGURE 6.2 — "Ten files, one screen" [constructed teaching example]

PIPELINE — LOAN OFFICER VIEW · exported 7:15 a.m. · day 33 of the Linden Street file AGE = days since the file was created in the LOS. IN-ST = days in the current stage.

LOAN # FILE / TYPE AMOUNT STAGE IN-ST AGE FLAG ───────────────────────────────────────────────────────────────────────────────────── L-2231 Harlow St · FHA purch 207,475 pre-approval 12 12 shopping L-2247 conv purchase 198,000 application 1 1 NOT LOCKED L-2244 conv purchase 268,000 processing 2 6 — L-2238 VA purchase 341,500 processing 3 9 COE pending L-2205 conv refinance 289,000 underwriting 4 21 — L-2214 Linden St · conv purch 365,750 conditions 5 33 2 of 11 open L-2190 Cypress Ct · conv purch 432,000 conditions 6 39 appraisal low L-2226 FHA purchase 244,000 conditions 9 26 no contact 4d L-2201 conv cash-out refi 176,500 conditions 14 30 AGING L-2172 conv purchase 512,000 clear to close 1 44 CD out ───────────────────────────────────────────────────────────────────────────────────── 10 files 3,034,225 avg file 303,422.50

BY STAGE: pre-approval 1 file 207,475 app/process 3 files 807,500 underwriting 1 file 289,000 conditions 4 files 1,218,250 ← 40% of files, 40.2% of dollars clear to close 1 file 512,000 ────────────────── 10 files 3,034,225

THE DOCUMENT Daily pipeline export from the loan origination system, loan-officer view, run before the desk opens on day 33. Ten active files. THE CONTEXT A working purchase-market pipeline in a retail shop. Three of these files are this book's constructed anchors; the rest are unnamed. WHAT IT SHOWS $3,034,225 in ten files, averaging $303,422.50. Four files — 40% of the count and 40.2% of the dollars — are sitting in conditions, which is exactly what §6.5 predicts and is normal rather than alarming. Sorted by days IN STAGE rather than by amount, the report says something the dollar column hides: L-2201 has been in conditions for 14 days on a 30-day-old file, and L-2226 has had no borrower contact in 4 days. WHAT IT DOESN'T It does not show whether any of this will close. A pre-approval with no contract (L-2231) is not comparable to a file at clear to close (L-2172), and adding their loan amounts together produces a number that means nothing. It does not show WHY a file is aging — L-2201's 14 days could be a title defect or an unopened email, and the report cannot tell you which. And it does not show a single hard date. There is no LOCK EXPIRES column, no CONTRACT CLOSING column, and no DAYS SINCE LAST CONTACT column. On the morning this report was run, the Linden Street lock had NINE DAYS LEFT and nothing on this screen says so. THE DECISION The first call of the day is not Linden Street and not the $512,000 file. It is L-2201, 14 days in conditions, and the question is not "any update?" but "which of these conditions is actually blocked, and by whom?" Second call: L-2226's borrower, because four days of silence on a file in conditions is a file that has quietly stopped. Third: L-2247, which took an application yesterday and is not locked — that is a pricing exposure, and Chapter 30 explains what it costs. Then, before the desk opens, add the three missing columns to this report by hand for all ten files. That exercise takes eleven minutes and would have saved $914.38 on one of them. THE LESSON The report sorts by dollars. The risk sorts by days. Read the IN-STAGE column first, every morning, and start at the bottom — then go find the dates the report left out, because the expensive ones are always the dates nobody printed. ```

Constructed for teaching. Loan numbers, amounts, and non-anchor files are illustrative; the Harlow Street, Linden Street, and Cypress Court files are this book's anchors.

The three questions a pipeline report answers

1. What is stuck? Sort descending by days-in-stage. The top of that sort is your morning.
2. What is at a deadline? Lock expirations and contract closing dates, which a good report shows and this one does not. Add the columns if your system allows it, and hand-maintain them if it does not — this is the question the Linden Street file failed. A lock expiration is the only date on a file that costs money the instant it passes, with no warning, no email, and nobody's signature required.
3. What is quiet? Days since last borrower contact is the most predictive column on any pipeline report and the one most often missing. A file nobody has spoken to in four days is not calm; it is unobserved.

The two questions it does not answer

Will these close? The pipeline report shows volume in process, not volume that will fund. The share of applications that actually fund — pull-through — is the number that turns a pipeline into an income forecast, and Chapter 39 makes it a management discipline. A loan officer who reads \$3,034,225 on a screen and mentally spends the commission on all of it is making a beginner's error, and the pre-approval row alone should tell you why.

Why is a file stuck? No report knows. Reports produce questions. Answering them is the job, and the answer always arrives by telephone.

🗂️ The Loan File

Chapter 6 contribution: map the fifty-one days onto the pipeline.

This is the first chapter that can show the whole shape of the Linden Street file. Chapters 1 through 5 contributed pieces — the parties, the history, the license, the arithmetic, the program. Chapter 6 contributes the spine that holds them: a dated sequence with an owner on every day.

The complete calendar, with ownership and control:

Where the days actually went:

• 12 days (23.5%) waiting on third parties who had already been given the order — irreducible, and the reason day 7 matters so much.
• 19 days (37.3%) in conditions — of which 11 days (21.6% of the entire file) were a fully documented loan sitting still.
• 5 days (9.8%) in underwriting, the stage everyone blames.
• 5 days (9.8%) at the front, before there was an application to take.
• 4 days (7.8%) from clear to close to the wire, of which three were the disclosure waiting period and therefore not anyone's to compress.

Which delays were controllable: exactly one block of them, and it is the biggest. Days 34 through 43 were not caused by an appraiser, a title examiner, an underwriter, or a borrower. They were caused by a team correctly observing that the closing was day 45 and that the two remaining conditions could not be cleared early — and drawing the reasonable, wrong conclusion that there was therefore nothing to do.

Those ten days produced both of the file's failures. The rate lock expired inside them, on day 42, costing \$914.38. The furniture was financed inside them, on day 41, costing a closing date. Neither event required anybody to make a mistake; both required only that a file with nothing wrong with it be allowed to sit.

There is one further controllable item, and it sits earlier and quieter: the 30-day lock taken on day 12 expired on day 42, three days before the file's own scheduled closing date. That was knowable on day 12 with one subtraction. Chapter 30 owns the pricing judgment; Chapter 6's version is simpler — when you lock, do the arithmetic out to the expiration and compare it to the contract date, out loud, before you accept the price.

What this settles: the shape of a purchase file, the owner of every stage, and a defensible answer to "why does this take so long." You can now tell a borrower on day 5 approximately what days 9, 23, and 45 will look like, which is most of what "great communication" turns out to be.

What it does not settle: how to actually clear eleven conditions efficiently (Chapter 19), what the appraisal on day 16 really established (Chapter 18), what the title commitment's Schedule B-II means (Chapter 21), why a 30-day lock was the wrong instrument on day 12 and what the day-42 extension was really buying (Chapter 30), and how to run ten of these at once without dropping one (Chapter 39).

Open questions carried forward:

• Q6.1. If the two prior-to-funding conditions cannot be cleared early, what can be compressed in a nineteen-day condition stage? (Chapter 19)
• Q6.2. The file reached CTC on day 47 and closed on day 51. Three of those four days were the disclosure waiting period. What determined the fourth, and could it have been avoided? (Chapter 22)
• Q6.3. What would this file's contract-to-close number have been if the appraisal had come back low instead of at value — and is that number the loan officer's fault? (Chapters 18, 20)

Your task. In Appendix C's workbook, build the calendar. Take the seven stages from §6.1, write in your own file's dates (or Linden Street's), and add three columns the pipeline report in FIGURE 6.2 does not have: lock expiration, contract closing date, and days since last borrower contact. Then answer one question in writing: on the day this file reached its last cleared condition, what were the next three dated events, and who owned each one? If you cannot answer that question about a live file in under thirty seconds, the file is not being managed. It is being watched.

Conclusion

A loan file passes through seven stages. Six of them are a line and one of them is a loop, and the loop is where files die.

The stages have owners, and the owners change six times — but ownership never moves. The underwriter has the file for five days and the authority for all fifty-one. The processor assembles. The closer executes. The loan officer is present the entire time and is the only person who can see the whole calendar at once, which is not a burden the job description mentions and is the actual job.

The arithmetic of this file is the argument. Fifty-one days: five at the front, twelve waiting on orders already placed, five in underwriting, nineteen in conditions, four at the end. Underwriting — the stage that gets blamed for everything — was under a tenth of it. The stage nobody has a name for took better than a third, and well over half of that was a fully documented loan sitting motionless because its closing date was still two weeks out.

Those eleven motionless days sent two bills. The rate lock expired inside them and cost \$914.38 to extend. The borrowers financed furniture inside them and cost the file its closing date, a debt-to-income ratio at 48.48%, and very nearly the house. Netted against the prepaid interest the delay gave back, the cash came to \$516.06 — real money, roughly eight and a half months of the rate advantage the borrowers had been shopping for, and still the smallest item on the list.

That is the thing to carry out of this chapter. A file that is still open is a file that is still exposed — to a credit inquiry, a stale document, an expiring lock, an employer's reorganization, and a furniture store with a financing desk. Delay bills you in dollars, and you will pay that bill. It charges you in exposure, and that is the invoice that takes the house.

Which gives you the operating rule that the rest of Part II and Part III will keep proving: drive to clear to close, not to the closing date. Every day you pull that milestone forward is a day of exposure you have removed from a family's largest financial transaction, whether or not the closing itself moves an inch.

Next: Part II opens at the front of the pipeline, where the whole thing starts and where most loan officers are weakest. Chapter 7 asks the question this chapter took for granted — where does a file come from? — and it will use §6.7's contract-to-close number as currency, because turn time is not just an operations metric. It is the thing a referral partner is actually buying.

Key Terms

Pipeline — the set of loan files a loan officer or a lender currently has in process, from first contact through funding; also used loosely for future business not yet in process. (Ch.6)

Prospect / lead — a person who has expressed interest in financing and about whom nothing has been verified; not yet a file. (Ch.6)

Pre-qualification — an estimate of what a borrower can likely borrow, based on information the borrower has stated and the loan officer has not verified. (Ch.6)

Pre-approval — a written statement that, based on verified information, a specified borrower qualifies for a specified loan amount and program, subject to a property and to conditions. (Ch.6)

Application — a formal request for a specific loan on a specific property; under the integrated disclosure rule it exists once six items are present: name, income, Social Security number, property address, estimated property value, and loan amount sought. Its existence triggers disclosure obligations. (Ch.6)

Processing — the assembly of a complete, internally consistent, submittable loan file: ordering third-party reports, collecting documentation, verifying internal consistency, and submitting to underwriting. (Ch.6)

Underwriting — the evaluation of a complete file against applicable guidelines and lender overlays, resulting in a decision: approved with conditions, suspended, or denied. (Ch.6)

Conditional approval — an approval granted subject to a specified list of items being delivered and found satisfactory; in ordinary practice, what an approval is. (Ch.6)

Condition / stipulation ("stip") — a specific item the underwriter requires before an approval becomes final; classified by source (borrower, third party, lender) and by timing (prior to document, prior to funding, prior to purchase). (Ch.6)

Clear to close (CTC) — the underwriter's sign-off on all prior-to-document conditions, authorizing preparation of closing documents; not the same as permission to close today. (Ch.6)

Closing (consummation) — the signing event at which the borrowers execute the note and security instrument and funds are collected; conducted by a title company, escrow company, or attorney depending on state practice. (Ch.6)

Funding — the lender's disbursement of loan proceeds; a separate event from signing, and in some states on a different day. (Ch.6)

Turn time — the elapsed time between two defined milestones on a loan file, measured consistently; there is no single turn time, and the one referral partners use is contract-to-close. (Ch.6)

Milestone — a defined, dated event in the loan origination system that a file has either reached or not; the vocabulary of every pipeline report and turn-time calculation. (Ch.6)

Loan origination system (LOS) — the software of record for a loan file: application data, documents, conditions, disclosures, dates, and audit trail. (Ch.6)

Spaced Review

1. (Ch. 6) Name the event that ends each of the seven stages. Then say which stage has no defined duration and why.

2. (Ch. 3 + Ch. 6) A processor at your company, unlicensed, receives a borrower's call while you are out: "Can I switch to the 15-year? What would the payment be?" Using Chapter 3's definition of a mortgage loan originator and this chapter's description of the processor's role, say exactly what the processor may and may not do — and what they should say.

3. (Ch. 5 + Ch. 6) The Linden Street borrowers are conventional at 95% loan-to-value with monthly mortgage insurance. Suppose on day 30 the underwriter had suspended the file for an income shortfall and the only viable path was FHA. Using Chapter 5's program map, name three things about the calendar that would change — not three things about the payment.

4. (Ch. 6) Your agent asks on day 36 whether the file will close on time. Nine of eleven conditions are cleared; the two open ones are prior-to-funding. Write the two-sentence answer that is both honest and useful, and then name the one thing you would ask the agent to do. Note what you now know that nobody on this file knew on day 36: the rate lock expires in six days.

5. (Ch. 6) The six-day overrun cost \$914.38 in a lock extension, less \$398.32 of prepaid interest the borrowers no longer owed at closing, for a net of \$516.06. Reconstruct that arithmetic. Then answer two questions the chapter raises about it: why is the \$398.32 not really a saving, and why does the chapter still insist the \$516.06 is the cheap part of what those six days cost?