Case Study 30.2 — A Lock That Was Sized Correctly and Expired Anyway

⚠️ This is a labeled composite. The file below is constructed from documented industry patterns and does not describe any real borrower, lender, or transaction. It is not one of this book's four anchor files, and it does not recur in later chapters. Every dollar figure is derived arithmetically from the stated loan amount and is labeled where it depends on a hypothetical fee schedule. Extension pricing varies by lender and by day — verify your own lender's current schedule.

Case Study 30.1 looked at a whole market. This one looks at a single file, and it deliberately inverts the Linden Street lesson.

On Linden Street, the lock was under-sized and the expiration was arithmetically inevitable from day 12. That is the common failure, and §30.3 is built on it. But it is not the only way a lock dies, and if you take only that lesson away you will conclude that correct sizing solves the problem. It does not. It removes the failure you control. The ones you do not control are still out there, and when they hit, the interesting question is no longer "how did this happen" but "who pays?" — which is a harder question with a less comfortable answer.

The composite file

THE FILE                                                    [labeled composite]
  Purchase price ......................................... $412,000
  Conventional, 30-year fixed, primary residence
  Down payment, 10% ...................................... $41,200
  Loan amount ............................................ $370,800
  Locked ................................................. day 10
  Rate ................................................... 6.750%, par
  Lock period ............................................ 45 DAYS
  Lock expiration ........................................ day 55
  Contract's stated closing date ......................... day 42
  ─────────────────────────────────────────────────────────────────
  BUFFER .................................................. 13 days

Stop and appreciate that buffer, because it is what a correctly sized lock looks like. The loan officer opened the purchase contract, found day 42, subtracted day 10, got 32 days, and rounded up to the 45-day period rather than reaching for the cheaper 30. That is the §30.3 discipline executed properly. Thirteen days of margin on an ordinary conventional purchase is generous.

And on day 58 the loan closed, three days after the lock expired.

The payments, so the arithmetic below has something to bite on

Monthly principal and interest is linear in the loan amount at a fixed rate and term, so these scale exactly from the frozen \$365,750 grid by the factor $370{,}800 \div 365{,}750 = 1.01380725$:

| Rate | P&I on \$365,750 (frozen) | P&I on \$370,800 | |---|---|---| | 6.750% | \$2,372.25 | **\$2,405.00 | | 6.875% | \$2,402.72 | **\$2,435.89 | | 7.000% | \$2,433.34 | **\$2,466.94** |

Point costs on \$370,800: 0.125 point = **\$463.50 · 0.250 point = \$927.00 · 0.500 point = \$1,854.00**.

The extension. Assume a hypothetical lender schedule of 0.250 point per 15 days: the extension costs \$927.00.

Under a constructed market movement in which pricing on day 55 is an eighth worse, a relock at worst-case pricing delivers 6.875% — \$30.89 a month more**, which means the \$927.00 extension pays for itself in about 30 months. Under a constructed movement of a quarter worse, worst-case delivers 7.000% — \$61.94 a month more, and the extension pays for itself in about 15 months. The extension is the right call in both. The question was never whether to extend. The question is whose \$927.00 it is.

Three versions of the same expiration

The file above is identical in all three variants. Only the cause differs — and the cause is everything.

Variant A — the appraisal came back twelve days late

The appraisal was ordered on day 12 through the lender's appraisal management company, which quoted a seven-business-day turn time. It arrived on day 31. Nothing was wrong with the report, the value supported the contract, and there was no reconsideration. It was simply late, in a market where appraisers were backed up.

Who caused the delay? Not the borrower. Not, in any meaningful sense, the loan officer — the appraisal was ordered two days after the lock, which is prompt. The vendor was late, and the vendor was selected by the lender.

Who pays? The lender, almost certainly, and it should not be a difficult conversation internally. There is no borrower conduct to point at. Whether a charge of this kind can reach the borrower at all is a disclosure question governed by the tolerance rules and by what constitutes a documented valid changed circumstance — Chapter 22 owns that analysis and you should not attempt it from the sales side. What §30.8 tells you is the shape: an origination charge of this type is zero-tolerance, and a lender that cannot document a valid changed circumstance absorbs the increase.

The transferable lesson. Your vendors' turn times are your risk, not your borrower's. Which means your buffer has to be sized for the vendor you actually have, not the turn time on the vendor's marketing page. A loan officer who has watched their appraisal management company miss its quote three times this quarter and is still budgeting seven days is not sizing a lock. They are hoping.

Variant B — the borrower sat on a document for fifteen days

The underwriter's conditional approval on day 22 asked for two months of statements on a retirement account being used for reserves. The borrower was asked on day 23, reminded on day 27, reminded again on day 31, and delivered on day 38.

Who caused the delay? The borrower, unambiguously, and everybody involved knows it.

Who pays? This is where new loan officers reason their way into a violation. The instinct is: they caused it, so they pay. That instinct is a fairness argument, and fairness is not the operative test. Whether an increase in a zero-tolerance charge may be passed to a borrower depends on whether there is a documented valid changed circumstance and whether the re-disclosure requirements were met — a question with a specific legal answer, owned by Chapter 22, decided by your compliance department and not by your sense of justice.

In practice, a great many lenders absorb this one anyway, for two unglamorous reasons. First, the documentation bar is real and the penalty for getting it wrong is worse than \$927.00. Second, there is a commercial calculation: this borrower has a referral source behind them, and \$927.00 is cheap compared with the version of this story that borrower tells for the next ten years.

The transferable lesson, and it is the useful one: the fifteen days were not lost on day 38. They were lost on day 23, when the request went out as a line item on a stip sheet instead of as a conversation with a deadline and a stated consequence. Chapter 19 is about condition clearing as a discipline; this is what that discipline is for. A condition request that does not include a date and a reason is not a request. It is a notification.

Compare the two sentences:

  WHAT WENT OUT ON DAY 23
    "Underwriting needs 2 months statements on the retirement account. Thanks!"

  WHAT SHOULD HAVE GONE OUT ON DAY 23
    "I need two months of statements on the retirement account by Friday the
     26th. Here's why the date matters: our rate lock expires on day 55 and
     every day this sits pushes the closing. If I don't have it by Friday I'll
     call you Monday morning, because at that point it starts costing money."

Same request. One of them has a date, a reason, and a named consequence. The other one is a wish.

Variant C — the sellers asked to move the closing

On day 30 the sellers asked to delay closing by sixteen days because the home they were purchasing was not ready. The buyers, who wanted this house, agreed. A signed amendment moved the closing from day 42 to day 58.

Who caused the delay? The sellers.

Who pays? Almost always the buyer's lender, or the buyer — and almost never the sellers, because nobody asked them to.

That is the point of Variant C and it is the most immediately useful thing in this case study. The amendment is the moment when the cost of the delay gets allocated, and it is the moment when nearly everyone forgets that the delay has a cost. The sellers had something they wanted. The buyers had something to trade. And the lock extension — a real, quantifiable, foreseeable \$927.00 — was simply absorbed by whoever happened to be standing closest when it came due.

What a good loan officer does on day 30, the moment they hear the words "the sellers want to push closing":

1. Compute the exposure immediately. Open the lock, find the expiration, find the new closing date, and price the extension from your lender's current schedule. On this file: sixteen days past a day-55 expiration, so an extension is required, and it costs \$927.00.
2. Call the buyer's agent before the amendment is signed, not after. Give them the number.
3. Say the sentence almost nobody says: "Your buyers are being asked to do the sellers a favor that costs nine hundred twenty-seven dollars. If they're willing to do it, that's fine — but it should be in the amendment, and the sellers should be asked to cover the extension."
4. Stay out of the drafting. Allocating costs in a purchase contract amendment is the agents' and the parties' business, and in many states it touches on the practice of law. Your job is to supply the number and the deadline, accurately and early, and then be quiet.

Sometimes the sellers say no and the buyers agree anyway, because they want the house. That is a fine outcome — it was a decision made with information rather than a bill that appeared later. And sometimes the sellers say yes, and the loan officer who made one phone call on day 30 has saved their borrower \$927.00 without touching the rate.

What the three variants have in common

Line them up:

Every one of them was visible before it became expensive. That is the finding, and it is not a coincidence. Lock extensions are almost never caused by genuine surprises; they are caused by known risks that nobody priced at the moment they became known.

Which produces the discipline this chapter has been building toward, stated as an operating rule:

The day you learn a fact that could push the closing date, price the lock extension it implies and tell somebody the number. Day 12 for a short lock. Day 23 for a slow condition. Day 30 for a seller's amendment. Not day 55.

The variant that is not in this composite

There is a fourth cause, and the Linden Street file is it: the lock was never long enough, and no event caused anything.

It belongs in a different category from Variants A, B, and C for a specific reason. In each of those, something changed after the lock was taken — a vendor missed, a borrower stalled, a seller asked. In Linden Street, nothing changed. The contract said day 45 on day 12 and it still said day 45 on day 42. The lock was three days short at the moment it was signed.

That distinction matters practically, not just morally. A changed circumstance is, at minimum, something you can point at. An arithmetic error made at the outset is not an event. There is nothing to document, nothing to re-disclose against, and nothing to negotiate with a seller. The cost has one place to go.

That is the honest answer to the question §30.8 asks, and it is the reason this chapter exists in the shape it does. When you get the sizing wrong, the answer to "who pays?" is not complicated. It is you.

Discussion questions

1. The composite file carried thirteen days of buffer and still expired. Does that argue against the §30.3 sizing rule? Defend your answer using all three variants.

2. In Variant B the borrower unambiguously caused the delay, and the chapter still says the lender usually absorbs it. Explain why "they caused it, so they pay" is a fairness argument rather than an operative test, and name the chapter that supplies the operative test.

3. Rewrite the day-23 condition request from Variant B for a borrower who is anxious rather than negligent. Keep the date, the reason, and the consequence, but change the register. What did you have to alter, and what did you refuse to soften?

4. In Variant C, the loan officer's most valuable act took about four minutes and happened on day 30. Describe what makes that call hard to actually place, and what a loan officer has to believe about their role in order to place it.

5. A buyer's agent tells you that raising the extension cost during the amendment negotiation will "make the buyers look difficult and could cost them the house." Respond. Is there a version of your answer that is honest and still leaves the decision with the buyers?

6. Rank the four causes — vendor, borrower, seller, and under-sized lock — by how much control a loan officer actually has over each. Then rank them by how often each one ends up being paid by the lender. Explain the gap between the two rankings.

7. Using the frozen grid and the composite's \$370,800 loan, verify the \$2,405.00, \$2,435.89, and \$2,466.94 figures by scaling. Then state why scaling works here and one situation in which it would not.