My Rate Lock Was 11 Days From Expiring When Rates Dropped 0.5 Points

Wednesday, 10:15 AM. Kitchen counter, phone on speaker propped in the fruit bowl so the loan officer's voice echoed off the ceramic — rates opened at six and a quarter this morning — and behind the coffee maker, a paper calendar with day 37 of my 45-day lock circled in green. I set my mug down hard enough to slosh. Six and a quarter. I was locked at six and three quarters.

The loan was $340,000 on a Boise house under contract at $425,000, principal and interest of $2,206 at the locked 6.75. At the mornin's 6.25, the same loan runs $2,093 — $113 a month, $40,680 across thirty years. But the lock expired in 11 days, closing sat on day 44, and October had taught everyone in this market what a rebound looks like: mortgage rates bounced 0.375 in a single week that month and stranded two buyers in my agent's office alone. Eleven days is not a long time in title-work years.

The way through was two pages I'd nearly crossed out at application in October: a float-down addendum, free at this lender, letting a borrower in a falling market re-lock once at the better rate. I'd nearly declined it because of its catch — the re-lock price isn't the billboard rate — and that catch is the difference between a half point and three eighths, which I'll unpack, cuz nobody unpacked it for me first.

day 37 of 45

A rate lock is the lender holding a price for a window — 30, 45, 60 days — with longer windows priced slightly worse, which is why my 45-day lock at 6.75 cost a quarter point more than a 30 would have. Day 37 meant the appraisal was done, underwriting was clear, and the only things standing between me and keys were the title work and a calendar. Nothin about a lock prevents the market from improving around you. It only prevents you from benefiting, unless the addendum behind it says otherwise.

the addendum I nearly crossed out

In October I'd compared two lenders. Lender B — bigger name, nicer lobby on Fairview Avenue — wanted $995 for a float-down, usable once, with a minimum improvement of a quarter point. Lender an included the same option at no fee and the same trigger, but the addendum waived my right to extend the lock if the float-down got exercised, a clause the size of a sunflower seed on page 2. I almost struck the clause on principle. My agent, who watched a float-down save a client $212 a month back in 2024, told me to initial it and stop reading like a lawyer.

6.25 on a Wednesday

The morning move was real: the 10-year Treasury had dropped on an inflation print, and mortgage rates — which track it loosely, the way a dog tracks its owner around a park — opened at six and a quarter at three separate lenders I checked from the counter before calling my loan officer. She confirmed it in ninety seconds, then said the sentence that matters: you have one float-down, this is a qualifying move, want me to price it? I wanted her to price it. The price came back before lunch, and lunch went cold while I read it.

the half point that became three eighths

The catch, in numbers: the billboard 6.25 is priced for a fresh 30-day lock with the full pricing sheet behind it. A float-down re-lock gets quoted off a reduced sheet, and mine came back at 6.375 — every bit of the qualifying half point consumed down to three eighths of grasped improvement. Payment at 6.375 on $340,000 is $2,121 against my locked $2,206, so the real number is $85 a month and $30,600 over the life of the loan. Eighty-five, not one hundred thirteen. Counting the theoretical saving instead of the real one is, I've decided, the fastest way to hate a good deal.

nine days of paperwork

Exercising the float-down re-dated everything: a new loan estimate, a fresh three-business-day waiting period, a re-disclosed Closing Disclosure, all of which quietly consumed five of my remaining eleven days with the pace of a DMV in a heat wave. The title company re-drew the settlement statement twice. Underwriting never re-underwrote — the file, the income, the appraisal all stood — but paper moves at the speed of paper. Day 44 held. Closing happened on a Thursday at 11 AM in a title office with a single-cup brewer that deserved better pods than it got.

$85 a month, thirty years of it

The final stack, signed: $340,000 at 6.375, principal and interest $2,121, first payment January 5. Eighty-five dollars a month doesn't sound like victory until you hold it against what the calendar almost did — the October rebound, repeated inside my eleven days, would have left me locked at the original number with no float-down left and a closing date sliding toward extension fees. I bank the $85 by auto-transfer into a savings account labeled lock dividend, a name my wife finds unbearable and accurate in equal measure.

what happens when a lock dies

For contrast, my coworker Dara let a 30-day lock die in September 2023 waiting for a market that never came. Extensions cost approximately an eighth of a point per eight days; hers ran three extensions, then expired, and her re-lock priced a quarter point worse than where she'd began. The whole saga cost her about $3,900 between fees and rate, and the lesson she handed me for free: a lock is not a prediction, its insurance. You dont cancel insurance on day 37 because the weather looks clear.

the calendar in the kitchen drawer

The paper calendar lives in the kitchen drawer with the takeout menus now, day 37 still circled in green, day 44 starred in gold. I stopped telling myself I'd fix the rate later with a refinance whenever the market got friendly — the float-down was the market gettin friendly, exactly once, on a Wednesday, at 10:15, into a fruit bowl. Rates do what they want. The addendum I nearly crossed out is the only reason one Wednesday of it did what I wanted. If the phone rings from the fruit bowl again, I'll answer it the same way. Price it.