Appendix J — The Mortgage Timeline

Everything in this book that has a date on it, in order, on one page.

A mortgage file is not a list of tasks. It is a schedule with dependencies, and almost every loan that dies late dies because somebody was managing the list and not the schedule. The conditions were being worked. The borrower was responsive. The processor was diligent. And the file still closed six days late, because nobody looked at the calendar and asked the only question that matters on a timeline: what is everyone waiting for, and who is waiting on me?

This appendix answers that question four ways. §J.1 shows the clean file — a forty-five-day purchase in which nothing goes wrong — so you know what "on schedule" actually looks like. §J.2 isolates the deadlines that are not yours to move: the regulatory clocks, which run on their own definition of a day and do not care about your closing date. §J.3 reproduces the Linden Street calendar in full, all fifty-one days, with the weekday on every entry, because the weekday is where half the lessons live. §J.4 puts the clean file and the real one side by side and names the divergence points. §J.5 through §J.8 turn the whole thing into working tools: the dependency map, the variations by loan type, the day-by-day checklist, and a ranked list of where the days actually go.

⚠️ Every day range in this appendix is TYPICAL, not guaranteed, and not a promise you may repeat to a borrower. Turn times vary by lender, by market, by season, by appraiser availability, by county recording practice, and by whether the file crosses a holiday. An appraisal that takes six days in a slow February takes nineteen in a busy June in a market with four appraisers. State a range, state what drives it, and never convert a range into a date without checking the specific vendor and the specific week. The fastest way to lose a real estate agent's trust permanently is to promise a date you had no basis for.

J.1 The clean timeline — a 45-day purchase where nothing goes wrong

Start with the file you almost never get, because you cannot recognize a problem without knowing what the absence of a problem looks like.

THE CLEAN 45-DAY PURCHASE                     [constructed teaching example]
Day ranges are TYPICAL. Each column is one calendar day.

                         0    5    10   15   20   25   30   35   40   45
                         |----|----|----|----|----|----|----|----|----|
  lead, pre-approval     ###
  contract executed         ###
  application + LE            ####
  AUS run, orders out          ####
  APPRAISAL                     =============
  TITLE                         ===========
  VOE / VOD                     -------
  rate lock                          L
  insurance bound                      -------
  submission                                 ###
  underwriting                                 #####
  conditions                                       ##########
  final approval, docs                                      #####
  Closing Disclosure                                             C
  the three-day wait                                              ....
  PTF conditions                                                   +++
  CLOSE / FUND / REC.                                                 *

  LEGEND  ### lender-side work   === third-party vendor work
          --- verification in flight       ... a mandated WAIT, not work
          +++ prior-to-funding items   L rate lock   C CD received   * closing

Two features of that chart matter more than the bars themselves.

First, look at where the long bars are. The two longest runs on a clean purchase are the appraisal and the title work, and neither of them is being done by anyone at your company. They are vendor time. You can order them early or late, and that is the entire extent of your control. Everything the lender actually does — the application, the automated underwriting run, the underwriting review, the document preparation — is measured in days, not weeks. That single observation reorganizes how a loan officer spends attention, and §J.5 develops it.

Second, look at the block of dots near the end. That is not work. That is the three-business-day Closing Disclosure waiting period, and it is the only stretch of a mortgage file during which the correct action is to do nothing. It cannot be compressed, expedited, escalated, or charmed. It can only be started earlier, which means the real deadline on a purchase file is never the closing date — it is the date the Closing Disclosure has to be in the borrower's hands.

The milestone sequence, with the actor and the typical day

Reading the sequence as four phases

Phase one, days 0 to 6 — the front door. Everything here is the loan officer, and almost all of it can be done in a single afternoon if the borrower is prepared. The pre-approval on day 1 is not the loan; it is the borrower's permission slip to make an offer. The single most valuable thing that happens in this phase is not the letter — it is reading the contract on the day it is executed and writing down the three dates it contains: the financing contingency, the inspection deadline, and the closing date. Chapter 6 makes that a discipline and Chapter 20 makes it a contract-law obligation. Skip it and you inherit a schedule you have never read.

Phase two, days 6 to 22 — the vendors. Orders go out, and then the file mostly waits. The appraisal and the title commitment are the two long poles, and they run in parallel, which means the phase is as long as the longer of the two, not the sum. This is also the phase in which a loan officer feels least useful and is in fact most useful, because the pre-emptive work done here — a condition list anticipated, an insurance agent contacted, a gift letter drafted before anyone asks for it — is the only work that reliably shortens phase three.

Phase three, days 22 to 40 — underwriting and conditions. The underwriter's first look is fast. The conditions are not, and the reason is structural: a condition list is not a task list you can attack in parallel with your own effort. It is a set of requests directed at other people, most of whom do not work for you and none of whom have your closing date in mind. Chapter 19 is the whole craft of this phase. The pattern to internalize is that conditions do not take long because they are hard; they take long because they queue.

Phase four, days 40 to 45 — the disclosure and the table. This phase is governed by the clocks in §J.2 and by almost nothing else. Once the Closing Disclosure is received, the file is on rails. The only real work left is the prior-to-funding items, the walkthrough, and the wire — and the wire is the single highest-risk event in the entire transaction, because it is the moment a first-time buyer with every dollar they have is most vulnerable to a fraudulent instruction. Verify wire instructions by voice, using a number the borrower already had, every time, with no exceptions.

J.2 The regulatory clocks — the deadlines that are not yours to move

Everything in §J.1 can be argued with. A vendor can be expedited, an underwriter can be asked for a rush, a borrower can be pushed. The clocks in this section cannot. They are set by federal regulation, they run whether or not you are watching them, and missing one is not a scheduling problem — it is a compliance violation with an audit trail.

Chapter 22 teaches these in depth. This is the reference card.

THE REGULATORY CLOCKS ON A PURCHASE FILE       [structure; verify current rules]

  APPLICATION (the six items are complete)
     |
     |--- 3 BUSINESS DAYS ---> LOAN ESTIMATE delivered or placed in the mail
     |                          (GENERAL definition of business day)
     |
     |--- 3 BUSINESS DAYS ---> written notice of the right to receive a copy
     |                          of appraisals and written valuations
     |
     |--- no fee may be charged (other than a bona fide and reasonable
     |    credit-report fee) until the consumer RECEIVES the Loan Estimate
     |    AND indicates INTENT TO PROCEED
     |
     v
  RATE LOCK, or any valid CHANGED CIRCUMSTANCE
     |
     |--- 3 BUSINESS DAYS ---> REVISED LOAN ESTIMATE
     |
     v
  APPRAISAL COMPLETED
     |
     |--- promptly upon completion, or a specified number of business days
     |    before consummation, whichever is earlier ---> COPY TO THE BORROWER
     |
     v
  CLOSING DISCLOSURE
     |
     |--- must be RECEIVED at least 3 BUSINESS DAYS before consummation
     |    (PRECISE definition of business day)
     |
     v
  CONSUMMATION
     |
     |--- on a QUALIFYING REFINANCE only: 3 BUSINESS DAYS of RESCISSION
     |    before funds may be disbursed. NOT on a purchase.
     v
  DISBURSEMENT / FUNDING / RECORDING

J.2.1 The Loan Estimate — three business days after application

The Loan Estimate must be delivered or placed in the mail no later than the third business day after the creditor receives the consumer's application. "Application" is not a feeling and not a conversation; it is a defined set of six items, and the clock starts the moment the sixth one arrives. Chapter 9 lists them and Chapter 22 counts them. The practical consequence is that a loan officer can start the clock accidentally — a borrower who volunteers the property address in a text message on a Friday afternoon may have completed the sixth item without either of you saying the word "application."

A second, quieter Loan Estimate rule sits behind it: the consumer must receive the Loan Estimate no later than the seventh business day before consummation. On a forty-five-day purchase this is invisible. On a fifteen-day closing it decides everything, and it uses a different definition of business day from the three-day delivery rule. See §J.2.7.

J.2.2 Intent to proceed — the gate on charging anything

Before the consumer has received the Loan Estimate and indicated an intent to proceed, the creditor may not impose a fee on the consumer in connection with the application. The one narrow exception is a bona fide and reasonable fee for obtaining the consumer's credit report. There is no deadline by which the consumer must indicate intent, and silence is never intent — it has to be an affirmative act, and you have to be able to document what it was and when.

This is why the appraisal order sits where it does in the sequence. The appraisal carries a fee. On the Linden Street file, the appraisal was ordered on day 7, which means the Loan Estimate had been received and the intent to proceed had been documented before that order went out. Nothing in the calendar says so explicitly, and nothing has to; the sequence requires it. If you cannot point to the intent-to-proceed record that precedes your appraisal order, you have a problem that no amount of good faith fixes after the fact.

The related restriction is easy to violate out of pure helpfulness: a creditor may not require the consumer to submit documents verifying information related to the application before providing the Loan Estimate. Asking for paystubs is fine. Requiring them as a condition of moving forward, before the Loan Estimate has gone out, is not.

J.2.3 Revised Loan Estimates and the changed-circumstance rule

The Loan Estimate is a good-faith estimate in a technical sense: the charges disclosed are measured against the charges at closing, and the creditor is on the hook for the difference within stated tolerances. A creditor may use a revised Loan Estimate to reset that baseline only if a permitted reason exists. The main ones:

Two timing rules govern the revision. The revised Loan Estimate must be delivered or placed in the mail no later than three business days after receiving the information sufficient to establish the changed circumstance — not three days after you get around to it, and not three days after your processor tells you about it. And there is an outer boundary at the other end: the consumer must receive a revised Loan Estimate no later than four business days before consummation, and a revised Loan Estimate may not be issued on or after the date the Closing Disclosure is provided. When there is not enough room left, the revision is carried onto the Closing Disclosure instead.

The rate lock is the trigger a loan officer meets most often and forgets most reliably, because it does not feel like a disclosure event. It feels like good news. On the Linden Street file the rate was locked on day 12, a Monday, which put the revised Loan Estimate deadline on day 15, a Thursday — three business days, no weekend inside it.

Where a charge exceeds its tolerance and no valid changed circumstance supports the increase, the creditor cures it: a refund to the consumer together with a corrected Closing Disclosure, delivered within the period the rule allows after consummation. That is the mechanism behind the lock extension on the Linden Street file, which the lender absorbed rather than passing through — see §J.3.

J.2.4 The appraisal-delivery requirement

Two obligations, and loan officers routinely remember the second and forget the first.

The notice. For a first-lien loan secured by a dwelling, the creditor must notify the applicant in writing, generally within three business days of application, of the right to receive a copy of all written appraisals and valuations developed in connection with the application. This lives in the initial disclosure package, which is exactly why it gets forgotten — it goes out automatically and nobody reads it, right up until an examiner asks whether it went out.

The copies. The creditor must provide copies of appraisals and other written valuations promptly upon completion, or a specified number of business days before consummation, whichever comes first — and must provide them whether or not the loan closes and whether or not the applicant asks. The applicant may waive the advance-delivery timing; the applicant may not be deprived of the copy, which must still be provided at or before consummation. You may charge a reasonable fee for the appraisal itself. You may not charge for the copy. Verify the current number of business days and the current waiver mechanics with your compliance department — this is a rule whose timing element has been revisited and is the kind of specific that ages badly in print.

The scheduling consequence is the one to hold on to: an appraisal that lands three days before closing has a delivery problem in addition to whatever value problem it may have. On a file with a short fuse, order the appraisal on the first day it is lawful to order it, and treat the delivery obligation as part of the appraisal's turn time rather than as an afterthought.

J.2.5 The Closing Disclosure — received at least three business days before consummation

The consumer must receive the Closing Disclosure no later than three business days before consummation. Four words in that sentence carry the whole rule.

"Receive." Not send. Not issue. Not upload. If the disclosure is hand-delivered or electronically delivered with a tracked acknowledgment, receipt is the date of delivery. If it is mailed, or delivered by a method that cannot document actual receipt, the consumer is presumed to receive it three business days after it is placed in the mail — and only then does the three-business-day waiting period begin. A mailed Closing Disclosure therefore consumes roughly six business days, not three. This is the entire reason lenders push electronic delivery with signed acknowledgment.

"Three business days." Counted under the precise definition (§J.2.7), with the day of receipt as day zero, and consummation permitted on the third business day, not after it.

"Before consummation." Consummation is when the consumer becomes contractually obligated on the loan — which, in most states, is the signing. It is not funding and it is not recording, and in some jurisdictions those happen on different days. Know which event your state treats as consummation, because the clock runs to that event and to no other.

Three changes, and only three, restart the three-day clock:

That table is the one to memorize, in both directions. Loan officers lose days by assuming any change restarts the clock, and lose licenses by assuming none of them do.

J.2.6 The right of rescission — and the fact that it does not apply here

On a loan secured by the consumer's principal dwelling that is not a purchase-money mortgage, each consumer with an ownership interest in the dwelling has the right to rescind until midnight of the third business day after the latest of: consummation; delivery of the material disclosures; or delivery of the required copies of the notice of the right to rescind. Funds may not be disbursed until that period has expired and the creditor is satisfied that no consumer has rescinded. In practice that means a refinance signs on one day and funds on another — which is a fact you tell the borrower at application, not at the table.

Note the ownership test rather than the borrower test: a spouse who is not on the loan but is on title to the principal dwelling generally holds a right to rescind and must receive the notice. Miss that and the rescission period does not close.

⚠️ Two errors, mirror images of each other, both expensive.

Error one: applying rescission to a purchase. There is no three-day right of rescission on the purchase of a home. A first-time buyer who has heard about "the three-day right to cancel" will sometimes ask about it at the table, and a loan officer who confirms it has just told them something false about the largest contract they will ever sign. What a purchase borrower has is the three-business-day Closing Disclosure review period before signing. That is a period to read, not a period to undo. The two protections sit on opposite sides of consummation, and confusing them is the most common timeline error in this book.

Error two: forgetting rescission on a refinance. A refinance that signs on Thursday does not fund on Thursday. If Sunday falls inside the count — Friday one, Saturday two, Sunday skipped, Monday three — rescission expires at midnight Monday and disbursement happens Tuesday. A loan officer who told that borrower "we'll have your money Friday" has misled them by three business days and will spend the weekend explaining it. Count the rescission period before you promise a disbursement date, and count it with Saturday included.

J.2.7 The two definitions of "business day" — the point this appendix exists to make

Regulation Z defines "business day" twice. The two definitions produce different answers on the same calendar, and which one applies depends entirely on which timing rule you are counting.

The general definition: a day on which the creditor's offices are open to the public for carrying on substantially all of its business functions. This is a fact about your employer. If your shop is open Monday through Friday, your business days are Monday through Friday.

The precise definition: all calendar days except Sundays and the legal public holidays named in the federal holiday statute — New Year's Day, the Birthday of Martin Luther King Jr., Washington's Birthday, Memorial Day, Juneteenth National Independence Day, Independence Day, Labor Day, Columbus Day, Veterans Day, Thanksgiving Day, and Christmas Day. This one has nothing to do with your employer.

Saturday is the only day where they disagree, and that is exactly why it decides so many files.

⚠️ This is the single most misunderstood point in this appendix, and it is worth a worked example in both directions.

A Closing Disclosure is hand-delivered and acknowledged on a Thursday. No federal holiday falls in the following week.

```text PRECISE definition (the correct one for this rule): Thu receipt = day ZERO, not counted Fri business day 1 Sat business day 2 *** SATURDAY COUNTS *** Sun ----- not a business day, ever, under either definition ----- Mon business day 3 === EARLIEST CONSUMMATION: MONDAY

GENERAL definition (wrong for this rule; the lender is closed Saturday): Fri 1, Mon 2, Tue 3 === it would say TUESDAY ```

The cost runs both ways. Close on Monday having counted with the general definition in your head and you closed a day early under a rule you did not understand — a violation, discoverable on audit, and not curable by anyone's good intentions. Refuse to close until Tuesday, and you cost your borrower a day of interest, your lender a lock day, and your referral partner their confidence, in exchange for nothing at all.

The memory hook that survives pressure: the precise definition governs the rules that protect the borrower's time — waiting periods and rescission. Those run on the calendar. The general definition governs the rules about your production of a document, and those run on your office hours. When in doubt, look it up before you count it, and do not assume the rule next to it has the same answer.

One more wrinkle, named here so you know it exists: four of the eleven federal holidays are fixed to a date rather than to a weekday, and when one of them falls on a weekend and federal offices observe it on an adjacent weekday, the treatment of the observed day differs from the treatment of the named date. It arises two or three times a year. Take the specific answer from your compliance department, not from memory.

J.3 The actual Linden Street calendar — all 51 days

Below is the file's calendar exactly as it stands in the record, with the weekday and the date added to every entry. Day 0 is a Wednesday. The anchor at the other end is the closing: October 24, which makes day 0 September 3.

The full day-to-weekday grid

Every day of the file, so any day can be checked against any other.

DAY -> WEEKDAY -> DATE.   S = September, O = October.
Day 0 is WEDNESDAY, September 3.  Day 51 is FRIDAY, October 24.

        Sun      Mon      Tue      Wed      Thu      Fri      Sat
        -------------------------------------------------------------
  wk 1                             d0  S3   d1  S4   d2  S5   d3  S6
  wk 2  d4  S7   d5  S8   d6  S9   d7  S10  d8  S11  d9  S12  d10 S13
  wk 3  d11 S14  d12 S15  d13 S16  d14 S17  d15 S18  d16 S19  d17 S20
  wk 4  d18 S21  d19 S22  d20 S23  d21 S24  d22 S25  d23 S26  d24 S27
  wk 5  d25 S28  d26 S29  d27 S30  d28 O1   d29 O2   d30 O3   d31 O4
  wk 6  d32 O5   d33 O6   d34 O7   d35 O8   d36 O9   d37 O10  d38 O11
  wk 7  d39 O12  d40 O13  d41 O14  d42 O15  d43 O16  d44 O17  d45 O18
  wk 8  d46 O19  d47 O20  d48 O21  d49 O22  d50 O23  d51 O24

The calendar, all 51 days

J.3.1 The three weekend days, and why none of them is an error

Three entries on that calendar fall on a Saturday or a Sunday. All three are deliberate, and each teaches something a weekday entry could not.

⚠️ Read the weekends before you read the events.

Day 4 is a Sunday — the offer is accepted. Real estate happens on weekends. Offers are written Friday night, countered Saturday, and executed Sunday, while every lender in the county is closed. The transaction was three days old before anyone at the lender could act on it. Note the consequence: the contract was executed on day 4 naming a day-45 closing, so the file did not have forty-five days. It had 41. "Forty-five-day contract" is loose shorthand, and Chapter 6 corrects it.

Day 45 is a Saturday — the original closing date. The contract, written the night of day 0, named a date that fell on a Saturday. Nobody looked at a calendar. The date simply passed, unattended, because settlement tables do not convene on Saturday. This is not a defect in the file and it is not an error in the record — it is the reason a loan officer reads the contract's closing date against an actual calendar on the day the contract is executed, which Chapter 6 names as a discipline and Chapter 20 develops. The first real question about any closing date is not "can we make it?" It is "what day of the week is that?"

Day 46 is a Sunday — the borrowers pay off the furniture account online. Which is exactly how that happens. A borrower in a panic on a Sunday afternoon does not wait for Monday; they log in and pay. The lender could do nothing with it until Monday, day 47 — but the fact had already changed by then, and the file's job on Monday was only to document a change that had already occurred.

J.3.2 The TRID count at the end

This is the arithmetic the whole ending depends on.

THE THREE-BUSINESS-DAY CD RULE            [the Linden Street file]

  day 48  TUE  Oct 21   Closing Disclosure issued AND received  <- day ZERO
  day 49  WED  Oct 22   business day 1
  day 50  THU  Oct 23   business day 2
  day 51  FRI  Oct 24   business day 3  ===  CONSUMMATION

  Wednesday, Thursday, Friday. Exactly three business days, and not one spare.
  No Sunday and no federal holiday falls inside the window, so the precise and
  the general definitions happen to agree here. They will not on the next file.

The Closing Disclosure could not go out before the file was clear to close on day 47, and clear to close could not happen before the payoff documentation resolved the day-44 credit problem. Every piece of the ending is welded to the piece before it.

J.3.3 Where the delay entered — and where it did not

The file ran fifty-one days against a contract that named forty-five. It is tempting to attribute the overrun to the crisis, because the crisis is the memorable part. The calendar says otherwise.

Where the delay did NOT enter:

• Not underwriting. Submitted day 23, conditional approval day 28. Five days for a first look on a conventional file with two W-2 borrowers is a normal-to-good turn.
• Not the appraisal. Ordered day 7, returned day 16 — nine days, and it came in at \$385,000, supporting the contract exactly. No gap, no reconsideration of value, no re-negotiation.
• Not the title defect. A prior owner's mechanic's lien surfaced on Schedule B-II of the day-19 commitment, and it was cleared by day 30. Eleven days to find, chase, and cure a defect somebody else created is competent title work, not a delay. Note what "released and re-recorded" actually means, because Chapter 21 makes it the sharpest lesson in the chapter: the release had already been recorded, the title company's update search found it described the wrong lot number, and a corrected instrument had to be executed and re-recorded. A release that describes the wrong land releases nothing.
• Not the conditions themselves. Eleven conditions issued day 28; all nine prior-to-document items cleared between day 29 and day 33. That is five calendar days — but only three business days, because day 31 (Saturday, October 4) and day 32 (Sunday, October 5) were a weekend. Three business days to clear nine conditions across two borrowers, two third parties, and the lender's own MI department is excellent work.
• Not the borrowers' responsiveness. Every borrower-sourced condition cleared on day 29 or day 30 — the first two business days after the list was issued.

Where the delay actually entered:

• Day 33 to day 44 — eleven days in which nothing happened. The file was documentation-complete on Monday, October 6, and then it sat. Nobody ordered the two prior-to-funding conditions. Nobody compared the lock expiration to the closing date. Nobody prepared the closing package. This dead window is the file's real failure and it is the single largest block on the entire calendar. It is also the reason the day-41 furniture purchase found an opening.
• Day 12 — the lock was under-sized the moment it was taken. A 30-day lock taken on day 12 expires on day 42. The contract named day 45. The lock was three days short of the file's own scheduled closing before a single thing went wrong. Chapter 30 takes that decision apart.

J.3.4 What the six-day overrun actually cost

Two hard-dollar consequences, running in opposite directions, plus a set of costs with no dollar figure at all.

Chapter 6 works that arithmetic in full and makes the honest qualification that belongs with it: the \$398.32 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.

⚠️ The \$914.38 was not caused by the six-day overrun, and saying so is the most common misreading of this file. The lock expired on day 42, which is three days before the contract's own closing date. Even if the file had closed exactly on schedule, it would have closed on day 45 with a lock that had already expired on day 42. The extension was unavoidable from the moment the lock was taken on day 12. The day-44 crisis did not create that cost; it merely arrived while the file was already paying it. Do not let a dramatic event at the end absorb the blame for a quiet decision at the beginning.

And the dollars were the cheap part. Set \$516.06 against 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. None of that has a dollar figure and all of it is larger than \$516.06.

There is one further cost that shows up nowhere on a settlement statement. Resolving the crisis required paying the \$5,200 furniture balance in full out of reserves. Reserves after closing therefore fell from \$12,623.66 — 4.16 months of PITI** — to **\$7,423.66, or 2.45 months. The borrowers kept the house and lost most of their cushion, and the cushion is the thing that would have absorbed the next surprise.

J.3.5 Why six calendar days produced so little movement

The contract named Saturday, day 45. A settlement table does not convene on a Saturday, so the earliest date that contract could ever realistically have produced was Monday, day 47. From Monday, day 47 to the actual closing on Friday, day 51 is four business days — Tuesday, Wednesday, Thursday, Friday. That is the cost the record assigns to the crisis, and it is smaller than the six-calendar-day headline because the weekend of day 45 and day 46 absorbed two of the six days before any working time was consumed at all.

That is the teaching point rather than an accident. The weekend is not free time. It is time that costs money — locks age, contracts run, rates move — and produces nothing. A loan officer who builds a schedule without looking at which days are weekends is building a schedule that is already wrong by two days.

⚠️ One holiday sits inside this file's fifty-one days, and it is worth knowing where. The second Monday in October is a federal legal public holiday, and on this calendar the second Monday in October is day 40, October 13. Under the precise definition it is not a business day. It cost this file nothing, because day 40 falls inside the dead window when no clock was running. But move the Closing Disclosure one week earlier — into the week of day 37 — and that holiday adds a day to the count, in a week when the lock had four days left. Check for holidays before you promise a closing date, not after. And note the second-order trap: many lenders and title companies work that Monday even though it is a federal holiday, so the general definition may treat it as a business day while the precise definition never does.

J.4 The clean timeline and the real one, side by side

PLANNED vs. ACTUAL — the Linden Street file        [the Linden Street file]

                         0    5    10   15   20   25   30   35   40   45   50   55
                         |----|----|----|----|----|----|----|----|----|----|----|--
  PLANNED  appraisal            =============
  PLANNED  title                ===========
  PLANNED  submission                        ###
  PLANNED  underwriting                        #####
  PLANNED  conditions                              ##########
  PLANNED  CD received                                           C
  PLANNED  3-day count                                            .....
  PLANNED  CLOSE                                                      *

  ACTUAL   appraisal            ==========
  ACTUAL   title                =============
  ACTUAL   lien cure                        !!!!!!!!!!!!
  ACTUAL   submission                       #####
  ACTUAL   underwriting                         ######
  ACTUAL   conditions                                ######
  ACTUAL   DEAD WINDOW                                    ............
  ACTUAL   crisis                                                    X
  ACTUAL   weekend                                                    ww
  ACTUAL   CTC issued                                                   K
  ACTUAL   CD received                                                   C
  ACTUAL   3-day count                                                    ...
  ACTUAL   CLOSE                                                            *

  LOCK  30-day original              ooooooooooooooooooooooooooooooo
  LOCK  15-day extension                                            xxxxxxxxxxxxxxx

  Note where the 'o' bar ends: day 42. Note where the PLANNED close sits: day 45.
  The lock was short of the file's own closing date before anything went wrong.

The divergence points, named

Read D5 and D8 together, because they are the same failure at two scales. The dead window cost eleven days and the closing-department turn cost one, and both are the same species of loss: a file that is ready and is not moving because no one is pushing it. The eleven-day version is visible in hindsight to everybody. The one-day version happens on nearly every file, is invisible, and is recovered only by a loan officer who is watching for it.

And read D2 against D6. The crisis on day 44 is the dramatic event, and it is the one everyone remembers. It cost four business days and it was caught by a control that was designed to catch it — condition 11, the pre-closing credit refresh, written on day 28, sixteen days before the event it existed to detect. The system worked. The lock decision on day 12 was quiet, felt like nothing at the time, and cost \$914.38 with certainty. The loud problem was handled. The quiet one was not.

J.5 Who is waiting on whom — the dependency map

A Gantt chart shows when. A dependency map shows why, and it is the more useful of the two, because it tells you which delay matters.

THE DEPENDENCY MAP — what blocks what

  CONTRACT EXECUTED
        |
        +---> APPLICATION (six items) ---> LOAN ESTIMATE ---> INTENT TO PROCEED
        |            |                                              |
        |            +---> AUS RUN                                  |
        |                                                           v
        |                                            +--------------+--------------+
        |                                            |                             |
        |                                     APPRAISAL ORDER              (fee may be charged)
        |                                            |
        +---> TITLE ORDER ---> SEARCH ---> COMMITMENT ---> [DEFECTS?] ---> CLEAR TITLE
        |                                       |                              |
        +---> VOE / VOD --------------+         |                              |
        |                             |         |                              |
        +---> INSURANCE BOUND ------+ |         v                              |
                                    | |    APPRAISAL REPORT                    |
                                    | |         |                              |
                                    v v         v                              |
                              +-----------------------------+                  |
                              |   SUBMISSION TO UNDERWRITING|                  |
                              +-----------------------------+                  |
                                            |                                  |
                                            v                                  |
                              +-----------------------------+                  |
                              |  CONDITIONAL APPROVAL (PTD) |                  |
                              +-----------------------------+                  |
                                            |                                  |
                                            v                                  |
                                   CONDITIONS CLEARED <---------------------- -+
                                            |
                                            v
                                    DOCS ORDERED  ---> SETTLEMENT FIGURES
                                            |                  |
                                            +--------+---------+
                                                     v
                                          CLOSING DISCLOSURE ISSUED
                                                     |
                                                     v
                                          *** RECEIVED BY BORROWER ***
                                                     |
                                       3 BUSINESS DAYS -- a WAIT, not work
                                                     |
                              PTF CONDITIONS ------->  |
                              (verbal VOE, refresh)    v
                                              CONSUMMATION -> FUND -> RECORD

What is strictly sequential

These cannot overlap, no matter how much pressure is applied:

1. Application → Loan Estimate → intent to proceed → any fee may be charged.
2. Title order → search → commitment → cure of any defect → clear title.
3. Appraisal order → inspection → report → delivery to the borrower.
4. Submission → underwriting review → the condition list exists.
5. All prior-to-document conditions cleared → documents may be ordered.
6. Documents and settlement figures → the Closing Disclosure can be accurate.
7. Closing Disclosure received → three business days → consummation.
8. Consummation → funding → recording. (And on a qualifying refinance, consummation → rescission period → disbursement.)

What runs in parallel — and is routinely run in series anyway

The last row of that table is the Linden Street lesson in one line. The two prior-to-funding conditions could have been ordered any time after day 33. They were not ordered until the very end, and the eleven-day gap that created is where the file broke.

⚠️ The critical path on a purchase file is almost never underwriting. Everyone blames underwriting, because underwriting is where the file is when the bad news arrives. Look at how the Linden Street file's fifty-one days actually decomposed:

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

Underwriting was 9.8% of the file. Third-party vendor time was 23.5%, and conditions were 37.3% — of which eleven days, 21.6% of the entire file, was a documented file sitting still.

The practical instruction that follows is unglamorous and worth more than any other habit in this appendix: order the appraisal and the title work on the first day it is lawful and possible to order them, and never let a condition list queue behind a single owner. Everything else is optimization at the margin. Those two items are the critical path.

Float — how much slack each step has

"Float" is the number of days a step can slip without moving the closing date. Steps with zero float are the critical path.

J.6 The timeline by loan type

The skeleton in §J.1 is a purchase. Every other transaction type is that skeleton with steps added, removed, or re-ordered. Every day range below is typical and must be verified against your lender's current turn times and your market.

Purchase

The baseline. Its defining feature is that the deadline is external and contractual: a date somebody else wrote, that carries earnest money and a financing contingency behind it. That is why a purchase file has urgency built in and a refinance does not. It is also why the single most valuable five minutes of a purchase file happens on the day the contract is executed, reading the three dates it contains against a real calendar. Day 45 of the Linden Street file was a Saturday, and nobody noticed until it arrived.

Rate-and-term refinance

Remove the agents, the seller, the earnest money, the inspection, and the walkthrough. Then add three things.

The payoff statement. The existing servicer must produce a payoff figure good through a specified date, and servicers produce them on their own schedule. This is the item that most often becomes a refinance's critical path, and it is invisible to loan officers who have only done purchases. Order it early, watch the good-through date, and re-order it if the closing moves past that date.

Subordination, if a second lien or a HELOC exists. The junior lienholder must agree in writing to remain junior to the new first. Subordination departments are slow, the request has its own package requirements, and a subordination that comes back on day 38 can end a transaction. Order it in the first week.

The right of rescission. Signing and funding are on different days. Tell the borrower at application, in the same breath as the rate. See §J.2.6.

And subtract one thing that turns out to matter enormously: there is no contract. No agent is calling, no earnest money is at risk, nobody loses a house. Refinances drift for exactly this reason, and the drift is expensive because the lock is running the whole time. A refinance needs a manufactured deadline — usually the lock expiration, stated out loud to the borrower at application as a real date.

Cash-out refinance

Everything in rate-and-term, plus: tighter maximum loan-to-value, meaningfully worse pricing, program-specific seasoning requirements on how long the borrower must have owned the property, and in some programs restrictions or documentation around the use of proceeds. Appraisal waivers are far less commonly granted, so plan on a full appraisal and its turn time. The right of rescission still applies on a principal dwelling. Program limits and seasoning rules on cash-out are revised periodically — verify the current requirements with the agency guide or your lender's matrix rather than from memory.

New construction

The one loan type where the lender is not the long pole. Two structural variants exist — two separate closings (a construction loan, then a permanent refinance) or a single-close construction-to-permanent — and the added steps are the same either way:

CONSTRUCTION -- WHAT GETS ADDED

  builder approval / review        weeks, before anything else can happen
  plans and specifications         must be complete enough to appraise
  cost breakdown and budget        line by line; the underwriter reads it
  appraisal "subject to completion per plans and specs"
  draw schedule                    each draw = an inspection = a delay window
  inspections at each draw         a third party, on their schedule
  certificate of occupancy         the municipality, on their schedule
  final inspection / completion report (Form 1004D)
  extended rate lock or float-down long locks price worse -- budget for it

Two timeline consequences worth stating plainly. First, the completion date will move, and the rate-lock strategy has to assume it will; this is the one context in which a long, expensive lock or an extended lock with a float-down is usually the right instrument rather than an admission of defeat. Second, TRID contains a specific accommodation for construction loans whose settlement is delayed beyond a stated period — but the creditor generally must have disclosed on the original Loan Estimate that the settlement date might be delayed. Verify the current mechanics before relying on it. Chapter 35 covers construction, renovation, and reverse lending in full.

Condominium

A condominium purchase is a purchase plus project review, and project review is a different kind of risk from anything else on this list: it is the only common step that can render a file ineligible after everything about the borrower has already been approved. The borrower can be perfect. The project can fail.

CONDO -- WHAT GETS ADDED, AND WHO CONTROLS IT

  project questionnaire      -> the HOA or management company (NOT you, NOT the seller)
  budget and reserve study   -> reserve adequacy is a pass/fail test
  master insurance policy    -> plus walls-in (HO-6) coverage for the unit
  litigation review          -> pending litigation can be disqualifying
  owner-occupancy ratio      -> investor concentration in the project
  single-entity concentration-> how many units one owner holds
  delinquency ratio          -> owners behind on HOA dues
  deferred maintenance and special assessments

The questionnaire is the item to fear. It goes to a management company that has no interest in your closing date, charges a fee to complete it, and commonly takes one to three weeks. Order it on the day the contract is executed, before the appraisal, before anything. It is the longest-running open item on most condo files and it is the one nobody orders first.

In the wake of a widely reported 2021 condominium tower collapse, the agencies added project-eligibility questions concerning critical repairs, deferred maintenance, and special assessments. Those requirements have been revised more than once since, and lenders layer their own overlays on top. Verify the current project-review standards, forms, and any lender overlays before you quote a condo timeline to an agent.

J.7 What the loan officer does on each day

The dependency map turned into a checklist. Use it as a template, not as gospel; the days shift with the contract.

DAY 0-2  -- THE FRONT DOOR
  [ ] Take the call. Ask what they are trying to do before you answer what your rate is.
  [ ] Establish the four facts you need before quoting: representative score, LTV,
      occupancy and property type, lock period. Missing one = you have a range, not a quote.
  [ ] Pull credit with authorization. Establish the representative score.
  [ ] Ask the questions that do not come back later: military service, prior ownership,
      pending debt, anything about to change at work.
  [ ] Issue the pre-approval, and say out loud what it is and what it is not.
  [ ] Give the buyer's agent your cell number and your turn times, honestly.

DAY 3-5  -- THE CONTRACT
  [ ] READ THE CONTRACT the day it is executed.
  [ ] Write down THREE dates: inspection deadline, financing contingency, closing date.
  [ ] *** LOOK UP WHAT DAY OF THE WEEK THE CLOSING DATE FALLS ON. ***
  [ ] Count the days actually remaining -- from EXECUTION, not from the offer.
  [ ] Complete the six application items. The Loan Estimate clock starts here.
  [ ] Confirm earnest money was delivered per the contract's own deadline.

DAY 5-8  -- DISCLOSE AND ORDER
  [ ] Loan Estimate out. Same day if you can; the deadline is not the target.
  [ ] Document intent to proceed. Nothing with a fee moves until it exists.
  [ ] Run automated underwriting. READ the findings; do not skim the recommendation.
  [ ] Order the APPRAISAL. First day it is lawful and possible.
  [ ] Order TITLE. Same day.
  [ ] Send VOE and VOD requests. Same day.
  [ ] Send the borrower the document list, sorted by how long each item takes them.

DAY 8-20 -- THE VENDOR WINDOW (the phase that feels idle and is not)
  [ ] Track the appraisal by name and date, not by "it's ordered."
  [ ] Read the title commitment the day it arrives. Schedule B-II is where the news is.
  [ ] Lock the rate -- and size the lock against the CONTRACT'S closing date,
      plus a buffer, not against your optimism. Then issue the revised Loan Estimate.
  [ ] Confirm the homeowners insurance agent is engaged and knows the closing date.
  [ ] Pre-empt the condition list: gift letter, large deposits, LOEs -- before anyone asks.
  [ ] Call the borrower weekly with a fact, even when the fact is "nothing has changed."

DAY 20-25 -- SUBMISSION
  [ ] Submit COMPLETE. An incomplete submission buys a condition list, not a head start.
  [ ] Tell the agent the file is in underwriting and what the turn time actually is.

DAY 25-35 -- CONDITIONS
  [ ] Read the condition list the hour it is issued.
  [ ] Sort by OWNER (borrower / third party / lender) and by LEAD TIME, not top to bottom.
  [ ] Send every third-party request the same day.
  [ ] *** ORDER THE PRIOR-TO-FUNDING ITEMS AS SOON AS THE FILE IS DOC-COMPLETE. ***
  [ ] Compare the LOCK EXPIRATION to the CLOSING DATE. Write both on the file.
  [ ] Warn the borrower in writing: no new credit, no new debt, no job changes,
      no large deposits, until the loan funds. Then warn them again at day 35.

DAY 35-42 -- DOCS AND THE DISCLOSURE
  [ ] Push for final approval and docs. This is where files silently stop.
  [ ] Reconcile settlement figures with the title company BEFORE the CD is built.
  [ ] Work BACKWARD from the closing date to the CD receipt date, using the
      PRECISE definition, and check for a federal holiday inside the window.
  [ ] Get the CD out. Delivered and ACKNOWLEDGED, not mailed.
  [ ] Confirm the borrower's cash to close and where it is coming from.

DAY 42-45 -- THE TABLE
  [ ] Verify wire instructions BY VOICE, on a number the borrower already had.
      Every time. This is the highest-risk hour in the transaction.
  [ ] Confirm the walkthrough happened.
  [ ] Confirm the time, the place, and what identification to bring.
  [ ] Be reachable at the closing. Answer the phone.

AFTER CLOSING
  [ ] Confirm funding and recording.
  [ ] Tell the borrower when the first payment is due and to whom -- and that a
      servicing transfer is normal and does not change the terms.
  [ ] Thank the agent. Then ask what the file taught you.

J.8 Where time is actually lost — ranked, with the fix

Ordered by how many days each one typically costs across a book of business, not by how dramatic it feels at the time.

1. The file that is ready and is not moving. The single largest loss on the Linden Street file was eleven days — 21.6% of the whole transaction — in which a documentation-complete file sat untouched. Nothing was wrong. Nobody was blocked. The fix: a daily pass over every file by days since last activity, not by closing date. A file with four days of silence goes to the top of the list regardless of when it is scheduled to close. Chapter 39 builds the routine.

⚠️ The dead window is the most dangerous condition a loan file can be in, precisely because it generates no alarm. A file with a problem produces email, phone calls, and a condition somebody is chasing. A file that is ready and idle produces nothing at all. It is invisible on a pipeline report sorted by closing date, invisible in your inbox, and invisible in your memory — and it is where the Linden Street borrowers walked into a furniture store on day 41 with a file that should have closed a week earlier. Sort by silence, not by urgency.

2. Ordering the appraisal late. It is the longest single vendor bar on a clean file and the one whose failure modes take the most time to resolve: a short value, a subject-to repair, a reconsideration of value. Every day of delay in ordering is a day added to the end. The fix: order it the first day it is lawful and possible — meaning the day intent to proceed is documented — and never wait for anything else to come back first.

3. Ordering title late, or reading the commitment late. Title is the only step that can surface a problem requiring weeks, and the problem is always somebody else's. A mechanic's lien from a prior owner. A vesting error. A release that describes the wrong lot. The fix: order title the day the contract is executed, and read Schedule B-II the hour the commitment arrives. Eleven days to cure the Linden Street lien was good work — but only because it started on day 19 and ran in parallel with underwriting rather than after it.

4. Working the condition list top to bottom. A condition list is not a to-do list; it is a set of requests aimed at different people with wildly different lead times. Working it in order means the third-party items — the ones with a week of latency — get sent last. The fix: sort by owner and by lead time. Every third-party request goes out the same hour the list is issued, before you touch a single item you can do yourself.

5. Leaving prior-to-funding conditions to the end. They cannot be cleared early — a verbal verification of employment and a credit refresh are deliberately performed close to the note date, and that design is correct — but they can be ordered, scheduled, and calendared early. On Linden Street they were neither, and the discovery on day 44 landed with no runway. The fix: the day a file goes documentation-complete, calendar the prior-to-funding items with a named owner and a date.

6. Mailing the Closing Disclosure instead of delivering it. A mailed disclosure is presumed received three business days after mailing, and only then does the three-business-day waiting period begin. That is roughly six business days consumed instead of three. The fix: electronic delivery with tracked acknowledgment, or hand delivery with a signature. Then confirm the acknowledgment actually happened — an unopened e-sign invitation is not receipt.

7. Counting business days with the wrong definition. Costs a day when you are cautious and a violation when you are not. The fix: §J.2.7. Ask which rule you are counting before you count, and check for a federal holiday inside every window.

8. Under-sizing the rate lock. A short lock prices better, which is exactly what makes it tempting. The Linden Street 30-day lock taken on day 12 could never have reached the day-45 closing the contract named, and the \$914.38 extension was certain from the moment it was taken. The fix: measure the lock against the contract's closing date plus a buffer — not against your best case, and not against a closing date nobody has checked the weekday of.

9. Not reading the contract on day one. Three dates and a day of the week. The Linden Street contract named a Saturday and gave the file 41 days rather than 45, and both facts were readable on day 4 by anyone who opened it. The fix: read it the day it is executed and write the dates on the file where the whole team can see them.

10. Silence with the borrower and the agent. This one costs no days on the calendar and is on the list anyway, because it costs files. A borrower who has not heard from you in four days starts calling other lenders, and an agent who has not heard from you stops referring. The fix: a weekly call with a fact in it, including the weeks when the fact is that nothing has changed. Especially then.

The one sentence to leave with. A mortgage timeline is not a countdown to a closing date; it is a set of dependencies with three immovable clocks buried in it, and the loan officer's job is to know which of today's idle steps is the one everything else is waiting for. On the Linden Street file the answer, for eleven straight days, was all of them — and nobody asked.