Case Study 30.1 — The 2020–2022 Rate Cycle and What It Did to Locked Pipelines

Sourcing note. This case is built from documented public record: Federal Reserve policy announcements, Federal Housing Finance Agency actions, Freddie Mac's long-running Primary Mortgage Market Survey, and contemporaneous public statements by industry trade associations. No statistic in this case is invented, and none is stated with more precision than the public record supports. Where a magnitude is described qualitatively ("roughly doubled," "widened sharply"), that is deliberate — verify current and historical figures at the sources named at the end. Analytical reconstruction of mechanism is labeled as such.

Background: why this cycle is the one to study

Most loan officers working today learned the lock decision during a period when it did not seem to matter very much. Rates drifted. A lock was a formality. You took the thirty-day, the file closed, and nobody thought about the secondary desk.

Then, in the space of about three years, the market ran the entire range of what a lock can be: it became worthless, then priceless, then a weapon. If you want to understand why your lock desk behaves the way it does — why the cutoff is hard, why extensions cost money, why worst-case pricing exists, why nobody will let you lock a shopper — this is the period that wrote most of those policies.

The cycle had three distinct phases, and each one broke a different part of the lock machinery.

Phase 1 — March 2020: the hedge becomes the emergency

What happened, on the record. In March 2020, as the COVID-19 pandemic reached the United States, the Federal Reserve cut its target for the federal funds rate to near zero and announced large-scale purchases of Treasury securities and agency mortgage-backed securities. Those purchase announcements are public and dated. The immediate effect on the mortgage-backed securities market was violent: after an initial period of severe dislocation in which spreads widened dramatically, Federal Reserve buying drove MBS prices sharply higher over a very short window.

Why that was a problem rather than good news — the mechanism. Recall §30.1 and §30.10. When a lender locks a borrower, its secondary desk hedges the commitment by selling forward in the to-be-announced market. A rising MBS price is a loss on that forward sale. Under ordinary conditions this is fine: the locked loans in the pipeline become correspondingly more valuable, and the two offset.

But forward positions are collateralized. When prices move against a short position, the counterparty calls for margin — cash, now. And the offsetting gain in the pipeline is not cash. It is a portfolio of loans that have not closed, cannot be sold today, and in many cases have not even been approved.

So originators found themselves in the specific bind this chapter has been describing in miniature: a hedge that must be funded in cash today against an asset that will not convert to cash for weeks. Industry trade associations publicly raised the issue with federal regulators at the time; those communications are a matter of public record and are worth reading in the original if you want to see what a liquidity squeeze sounds like in the industry's own words.

What lenders did in response, as widely and publicly reported. Within weeks, across much of the industry: products were suspended (jumbo and non-agency programs first), minimum credit score overlays appeared, loan-to-value maximums tightened, some lenders temporarily stopped accepting new locks in certain channels, and lock periods were shortened or re-priced. Borrowers who had been quoted a rate on a Tuesday found on Thursday that the program no longer existed.

The teaching point. Every one of those responses is a lock desk protecting itself against the option it has written. In a calm market, the option is cheap and the policies are invisible. In a violent one, the option is the whole business, and a loan officer who does not understand that will spend the crisis telling borrowers that their lender is being unreasonable.

Phase 2 — 2020 into 2021: the boom, and the second nightmare

What happened, on the record. Mortgage rates fell to the lowest levels in the history of Freddie Mac's Primary Mortgage Market Survey, with the 30-year fixed average moving below 3%. Refinance volume surged.

The lock-desk problem in a boom is not price. It is capacity. Locks are commitments with expiration dates, and expiration dates do not care that underwriting is running six weeks behind. Across the industry, turn times stretched — appraisals, underwriting queues, closing calendars — and the result was predictable and expensive: a very large volume of locks approaching expiration on files that were nowhere near ready.

Two things follow, and both are documented industry behavior rather than speculation:

Longer locks became the default. When a lender knows its own underwriting queue is running five weeks, a thirty-day lock is not a product. It is a promise to charge an extension later.

Lenders widened their margins. With capacity constrained, many lenders priced deliberately less competitively — not out of greed but because the binding constraint was operational, not commercial. A lender that cannot process the volume it already has does not want more of it. This is one of the genuinely counterintuitive facts about mortgage pricing: your rate sheet reflects your employer's capacity as well as the bond market's mood, and a loan officer who assumes the sheet is purely a market read will misjudge it repeatedly.

And fallout ran hot. In a falling market, every borrower with a lock is watching the rate improve after they took it. Some renegotiated. Some walked. Some lenders responded with discretionary market-improvement policies (§30.6) — a business decision to give back part of an improvement rather than lose the loan and eat the hedge unwind.

Phase 3 — 2022: the reversal, and the in-the-money lock

What happened, on the record. Through 2022, the Federal Open Market Committee raised its target for the federal funds rate repeatedly and rapidly. Long-term mortgage rates rose faster than the policy rate did, because the market was re-pricing inflation expectations over a thirty-year horizon, not an overnight one. The 30-year fixed average in Freddie Mac's survey roughly doubled over the course of the year, moving from around 3% to above 7%.

Note what that does to §30.4's argument. In 2022 the Fed was raising and mortgage rates rose. In other episodes the Fed has moved and mortgage rates have gone the other way. The relationship is not a mechanism; it is a correlation that holds when the underlying story is the same and breaks when it is not. A loan officer who taught borrowers "the Fed raises, your rate rises" in 2022 was right for the wrong reason and will be wrong later for exactly that reason.

What the reversal did to locks. Everything inverted:

• Fallout collapsed on locked files. A borrower holding a lock taken six weeks ago was holding something genuinely valuable and sprinted to close. Extension requests became a different conversation: borrowers wanted extensions and were happy to pay for them, because the alternative was current market.
• Worst-case pricing became a live and painful concept. A lock that expired in a rapidly rising market did not relock at anything close to the original number. This is the year in which a large number of loan officers learned §30.7 the expensive way.
• Purchase files broke on qualification. Look again at the §30.2 table. A file approved at 42.66% back-end has roughly one percentage point of rate movement before a 45% ceiling comes into view. In 2022 the market covered several times that distance. Borrowers who were floating were not merely paying more; a substantial number stopped qualifying for the house they were under contract on.
• Volume collapsed and the industry contracted. Refinance activity fell away almost entirely, and publicly reported layoffs, consolidations, and exits followed across the sector through 2022 and into 2023.

One more documented item worth knowing. In August 2020 the Federal Housing Finance Agency announced an "adverse market refinance fee" — a 0.50% price adjustment on most refinance loans delivered to Fannie Mae and Freddie Mac. After public objection the effective date was delayed, and the fee was eliminated in 2021. That sequence is public record and is a clean illustration of something §30.4 only gestures at: your rate can move because of a policy decision that has nothing to do with the bond market at all. Verify the dates and terms with the FHFA before quoting them.

What this case shows

1. A lock is an option, and in a volatile market its value is enormous. Everything in Phase 1 is the consequence of that sentence. The hedge exists because the option exists. The margin call exists because the hedge exists. The product suspensions exist because the lender is trying to stop writing options it cannot hedge.

2. Lock policy is written in the aftermath of a crisis and enforced during the calm. If your employer's lock policy feels excessively cautious, it is because somebody's employer was not cautious enough during one of these phases.

3. The correlation between policy rates and mortgage rates is not a mechanism. Phase 3 looks like the Fed driving mortgage rates. It is not. It is inflation expectations moving both, at different speeds, in a period where they happened to agree.

4. Capacity prices loans. In Phase 2, rate sheets reflected operational constraint as much as market level. This is invisible to borrowers and to most loan officers.

5. The direction of the market determines which way your problems come at you. Falling market: fallout, renegotiation, hedge losses, borrowers leaving. Rising market: expirations, worst-case relocks, qualification failures, and borrowers who cannot buy the house anymore. There is no market in which the lock decision is unimportant; there are only markets in which it fails differently.

Outcome

The industry that emerged from this cycle is measurably more conservative about locks than the one that entered it: longer default lock periods on purchase business, stricter stage requirements before a file may be locked, tighter cutoff discipline, and more explicit written lock policies. Those are observable, ordinary features of the business today, and they are the residue of the three phases above.

For an individual loan officer, the durable outcome is smaller and more useful. Three of the habits this book teaches were, for many practitioners, learned in this specific period:

• Size the lock against the contract, not against the pipeline you wish you had.
• Never explain a mortgage rate by reference to the federal funds rate.
• Never tell a borrower which way the market is going.

Lesson

A rate lock is the only place in a residential mortgage file where a household's thirty-year cost is decided by a market that moves in seconds. Everything else in origination is documentation: slow, checkable, and correctable. The lock is not. It is a real-time decision with an expiration date attached, made on behalf of people who cannot evaluate it, using an instrument the lender has to hedge in a market the borrower has never heard of.

That is why this chapter refuses to give you a rule, and why it spends its arithmetic on the one part of the decision that is knowable in advance: how many days you actually need.

Discussion questions

1. In Phase 1, rising MBS prices — normally good news for borrowers — created a cash emergency for originators. Walk the mechanism from a borrower's lock to a margin call, naming each step. Why does the offsetting gain not solve the problem?

2. In Phase 2, several lenders priced less competitively while demand was at a record. Explain how that can be rational, and say what it implies about reading a competitor's rate sheet as a pure market signal.

3. Phase 3 looks like proof that the Federal Reserve sets mortgage rates. Using §30.4, write the three sentences you would use to explain to a borrower why that reading is wrong even though the two moved together that year.

4. Suppose you had a purchase file under contract, floating, at the start of Phase 3, with a borrower at a 42.66% back-end ratio. Using the §30.2 table, describe the sequence of things that go wrong, in order, as the market moves — and identify the point at which the problem stops being about the payment.

5. The FHFA's adverse market refinance fee changed loan pricing by administrative action rather than by market movement. What does that episode imply about the sentence "your rate is set by the bond market"? How would you phrase that sentence more accurately to a borrower?

6. Every lock policy provision named in this case — stage requirements, cutoffs, extension fees, worst-case pricing — is a response to something that happened in one of the three phases. Pick three provisions from your own employer's lock policy and identify which failure each one is defending against.

Sources to verify

• Federal Reserve — FOMC statements and implementation notes for 2020 through 2022, including the announcements of large-scale Treasury and agency MBS purchases. (Tier 1)
• Freddie Mac, Primary Mortgage Market Survey — the long-running weekly series for the U.S. average 30-year fixed rate. The authoritative public source for the levels described here. (Tier 1)
• Federal Housing Finance Agency — the August 2020 announcement of the adverse market refinance fee, the delay of its effective date, and its 2021 elimination. (Tier 1)
• Mortgage Bankers Association — contemporaneous public statements to regulators regarding margin calls on originators' hedge positions in March 2020. (Tier 2 — real and public; verify the specific statements and dates.)
• Any specific figure not published by one of the above should be treated as unverified. Do not quote a fallout percentage, a reprice frequency, or a margin-call amount from memory or from a trade-press summary without checking the primary source.