Will Tariff Refunds Lower Mortgage Rates? Not So Fast

No, tariff refunds won’t magically cut your mortgage rate

. Look, I get it, the headline sounds great: “Tariff refunds are coming.” Your brain jumps to, “Cool, my mortgage quote’s 50 bps lower by Friday.” Here’s the thing: mortgage pricing doesn’t work like a coupon code. Rates move with inflation expectations, Treasury yields (especially the 10‑year), and the spread investors demand on mortgage‑backed securities (MBS). One-off tariff refunds? Interesting macro footnote, but they don’t reroute the plumbing that sets your 30‑year fixed this week.

Quick refresher on the actual levers that move your rate sheet:

• Inflation expectations: Markets price what they think inflation will be, not what a headline hopes it might be. Traders watch measures like breakevens and, yes, the 5y5y. When those drift lower, yields can ease. When they don’t, nothing moves.
• 10‑year Treasury: The 30‑year mortgage rate tends to track the 10‑year with a spread. Historically that spread sat ~170-200 bps, but in 2023 it blew out above 300 bps at times (Freddie Mac PMMS vs. U.S. Treasury data, 2023). That spread is still the boss.
• MBS spreads and liquidity: Investors want compensation for prepayment and convexity risk. If MBS prices don’t rally, your rate sheet won’t either. Simple as that.

So, would tariff refunds lower inflation tomorrow? Not really. Tariff refunds are one-off and uneven. They don’t instantly reset sticker prices or the Fed’s reaction function. Services make up the bulk of the inflation basket, about 60% of CPI weights in 2024 per BLS, while most tariff talk is about goods. And even for goods, pass‑through is messy: studies during 2019-2021 found only partial pass‑through (roughly half) and over several quarters, not days. If you’re thinking “will-tariff-refunds-lower-mortgage-rates-soon,” the market’s honest answer is: maybe at the margin, later, if it actually changes inflation data and term premiums. Not this afternoon.

Speaking of which, if your lender isn’t repricing your lock options today, there’s a reason: the market hasn’t repriced the risk yet. Lenders hedge with MBS and Treasuries, and they reprice rate sheets when those hedges move enough to matter. A catchy headline doesn’t change their P&L. Earlier this year we saw multiple CPI prints push the 10‑year up and down 10-15 bps intraday; that kind of move can nudge mortgage quotes a touch. A tariff refund headline without a move in breakevens or MBS? That’s just noise.

Anyway, what you’ll get here: we’ll break down what actually drives your mortgage rate week to week, how to read Treasury and MBS signals, and why tariff news, while politically hot, rarely shows up in your monthly payment right away. Personal note: three clients texted me this morning asking if they should float because of the refund story. My reply was the same I’ll give you, watch the 10‑year, watch MBS, and watch inflation expectations. Headlines don’t pay your mortgage, markets do. And yeah, occassionally I wish it were the other way around.

What actually sets your mortgage rate (hint: it’s not a press release)

Look, lenders don’t price off headlines, they price off bonds they can hedge and sell. For conventional 30-year loans, that means mortgage-backed securities (MBS). Your quoted rate is basically the MBS yield plus the lender’s margin and costs. And MBS yields, in turn, tend to track the 10‑year Treasury with a spread. Historically, the 30‑year mortgage rate sits roughly 1.5%-2.0% above the 10‑year Treasury, but that gap isn’t fixed; it breathes with market stress, prepayment risk, and capacity. I’m oversimplifying a hair, but that’s the spine of it.

So, what moves this week to week?

• 10‑year Treasury as the anchor: The 10‑year is the market’s shorthand for long-term inflation and growth. When inflation data surprises (CPI, PCE), Treasury yields move. Earlier this year we saw multiple CPI releases swing the 10‑year by 10-15 bps intraday, which can nudge mortgage quotes the same day. No move in Treasuries? Don’t expect much from your rate sheet.
• MBS spread over Treasuries: MBS usually trade at a spread to the 10‑year to compensate investors for prepayment and liquidity risk. In calm periods, call it a normal-ish range; in choppy markets, that spread can widen 25-50 bps (sometimes more), offsetting any benefit from a dip in Treasury yields. That’s why you’ll hear, “Treasuries are down but rates didn’t improve”, the spread did the dirty work.
• Prepayment risk (the refinance wildcard): MBS investors hate being paid back early at par when rates fall. If the market thinks a refinance wave is possible, investors demand extra yield. That extra yield = wider spreads = less rate relief for you. When rates jump, the opposite happens (less prepay risk), but capacity and volatility can still get in the way.
• The Fed’s stance: The Fed sets short rates and steers liquidity. Even though your mortgage is long term, the Fed matters in two ways: (1) its inflation fight steers the whole yield curve via expectations, and (2) balance sheet policy. The Fed has allowed MBS to passively roll off since 2022; changes to runoff pace shift supply/demand for MBS. More runoff can pressure MBS spreads wider; slower runoff can support them. Rate cuts don’t automatically translate to mortgage relief if spreads widen.

Here’s the thing, supply and volatility can swamp the nice, clean model. When lenders are slammed (spring buying season, or a brief refi flurry), the primary-secondary spread (what you pay vs. what MBS yield) can balloon 25-100 bps. When rate vol spikes (think big CPI/PCE/NFP days), investors demand a cushion. That cushion shows up as wider MBS spreads, which can fully offset a 10‑year rally. I was going to show the math on a current‑coupon stack.. anyway, the point stands.

Translation: lower Treasury yields help, but you only feel it in your quote if MBS spreads and lender margins don’t move the other way.

So basically, if you’re asking, “will-tariff-refunds-lower-mortgage-rates-soon,” the checklist is:

1. Did inflation expectations actually fall? (Watch breakevens and the 10‑year.)
2. Did MBS spreads tighten? (You need both a lower 10‑year and steadier/tighter MBS.)
3. Is the Fed signaling anything on balance sheet runoff that changes MBS demand?

If you don’t see at least two of those breaking your way, don’t expect your lender to reprice meaningfully. Headlines can nudge sentiment, sure, but rates live where MBS clear. That sounds a little dry, I know, but that’s just my take on it… and I’ve watched too many “big” stories fade when the 10‑year and current‑coupon MBS barely budged.

Tariff refunds 101: what they are and how money actually flows

So, quick definitions before we make this more complicated than it needs to be. “Tariff refunds” show up a few different ways: (1) classic duty drawback where the importer gets back duties when the goods are exported or destroyed, (2) litigation- or exclusion-driven repayments tied to things like Section 301/232 cases and retroactive exclusions, and (3) policy rollbacks where a new rule reduces or removes a tariff and prior entries get adjusted. In the U.S., drawback can refund up to 99% of duties, taxes, and certain fees under 19 U.S.C. §1313, and claims can usually be filed within 5 years of the date of import. That 99% number sounds huge, and it is, but here’s the thing, it accrues first to the importer of record, not directly to you or me at the register.

Who actually gets paid? Importers. Period. The ACH payment from CBP hits the importer’s bank account. If that importer has a pass-through pricing agreement with a retailer or OEM, some of it might trickle down as credits on future invoices, but it’s rarely instant. Litigation-driven repayments on 301 goods, think when exclusions were reinstated or extended in 2022-2024, also flow to the importer entry-by-entry after reconciliation. Speaking of which, brokers spend weeks scrubbing entry numbers, matching HTS codes, and filing claims. That alone should tell you why there’s no clean “refund day” for the economy.

Timing is messy. Even with accelerated drawback, you’re looking at approvals and post-entry audits that can stretch months. Without acceleration, a normal claim might take a quarter or two, and anything complex can drift into the 6-12 month range, longer if an audit flags documentation gaps. I’ve seen companies book a large receivable in Q1 and actually recieve the cash in Q3 or Q4. Anyway, the point is: macro data doesn’t see a giant refund spike in one month; it’s a slow drip.

Will refunds lower consumer prices right away? Honestly, probably not in a way you’ll feel. Academic work on the 2018-2019 tariff waves showed near-full pass‑through of tariffs to U.S. import prices at the border, but consumer price pass‑through was partial and slow, retailers adjusted over quarters, not days. Refunds run in reverse with even more friction. Contracts get repriced on schedule, not on Twitter. Lower landed costs now can ease future price increases (less need to hike on the next reset), but they don’t magically rewind the price level you paid last year. I might be oversimplifying, but think of it as taking your foot off the inflation gas, not slamming it into reverse.

What about CPI and your escrow analysis? CPI captures prices consumers pay at the point of sale. A refund booked by an importer doesn’t move that needle until and unless it shows up as lower retail pricing, and that happens as inventories turn and contracts reprice. BLS also samples items and rotates vendors, so the timing gets noisy. Your mortgage escrow? Same idea: property taxes and insurance don’t drop because a shoe importer in Long Beach got a 301 refund. If anything, the broader path for rates still lives with the 10‑year and MBS spreads. The 10‑year has been hovering around 4% lately, give or take, and 30‑year mortgage quotes are still in the mid‑6s for strong borrowers, so unless refunds reshape inflation expectations or MBS demand, rate sheets won’t suddenly get cute.

How the money actually flows, step by step:

• Importer files claim (drawback, exclusion, or litigation outcome) within the allowable window, again, drawback allows up to 5 years from import.
• CBP/brokers reconcile entries and documentation; accelerated programs speed this up but require pre‑approval.
• Refund cash hits the importer (ACH). Some of it may be shared down the chain via credits on future POs or price resets.
• Retail prices adjust later, partially, as contracts roll and old inventory sells through. No clean macro “refund day.”

One last practical note: lower landed costs can support promotional pricing or smaller hikes during holiday builds later this year, which might nudge certain CPI subcomponents. But reversing last year’s shelf sticker? That’s not how this works. Actually, let me rephrase that, sometimes you’ll see a targeted rollback on a high-visibility item, but it’s the exception, not the rule.

The inflation channel: can refunds dent CPI/PCE enough to matter?

Short answer: probably a little, but not in the way mortgage shoppers are hoping. Tariffs mostly hit goods prices. The rate conversation since 2022 has been dominated by services inflation, think shelter, insurance, healthcare, restaurants, not sneakers and toasters. In the official baskets, services are the heavyweight: services account for roughly two-thirds of PCE by weight (BEA methodology) and a bit under 60% of core CPI, while shelter alone is about one-third of headline CPI and a little over 40% of core CPI. That mix matters because the Fed reacts to where the heat is. And lately, the heat has been sticky services.

Here’s the thing: when tariff refunds lower landed costs, the impact flows into core goods. Since late 2023 and much of last year, core goods were disinflating or outright negative year over year, while services stayed firm. That split is still the story in 2025, services decelerated from the 2022 peak, sure, but they’re not exactly cool. The Fed’s “supercore” proxy (core services ex housing) ran hot in 2022, cooled through 2024, and this year is hanging around the high-2s to ~3% pace depending on the month. That’s the stuff that keeps 30‑year mortgage rates hovering roughly in the 6.7%-7.2% range recently, give or take a market scare.

Historical context helps. Back in 2018-2019 when tariffs went up, the price effects showed up in specific categories, not everywhere. Laundry equipment is the poster child: research on the 2018 washer safeguard found retail prices rose around 12% after the tariff, and dryers (untariffed) went up too because the bundle moved (classic spillover). Other categories, certain furniture lines, parts of electronics, saw increases, but pass‑through wasn’t 1:1 to the shelf. Import prices reflected the duties almost fully in many cases, but retail pass‑through varied with competition, contracts, and, frankly, whether consumers noticed. I watched a big-box buyer in 2019 hold a living‑room set flat for nine months just to keep Sunday-circular optics clean. Someone ate margin until the next assortment reset. Anyway…

Refunds today rerun that movie in reverse, only slower. You might see discounts, sharper promos, or smaller fall/winter price hikes in tariff‑heavy categories as 2025 holiday builds ramp. But services inflation still sets the tone for the Fed. Even if core goods shave, say, a few tenths off year‑over‑year for a stretch, mortgage rates won’t fall much unless overall inflation expectations ease. Market breakevens tell you as much: longer‑run inflation pricing has been sitting in the mid‑2% range this year, not screaming “mission accomplished.”

And just to ground this in how CPI/PCE mechanics meet trading desks:

• Weights: Goods are the smaller slice. Moving a 25% slice by 1% barely shifts the total when the 65-70% services slice is sticky.
• Timing: Refund pass‑through is lumpy. Inventory sells down, contracts reset, promotions come and go. CPI doesn’t get a neat, clean drop.
• Expectations: Breakevens and term premium matter for mortgages. A tiny goods disinflation bump won’t reset those on its own.

Look, I want cheaper appliances as much as anyone, I still remember paying up for a “tariff special” washer in 2019 and muttering about it for weeks, but rate relief rides on services cooling and expectations settling. If shelter continues to grind lower into late 2025 and medical/services don’t flare, then the goods help from refunds becomes additive. If not, it’s background noise. That’s a bit blunt, but that’s the rates market.

Bottom line: refunds can trim core goods, which helps headline optics, but they’re not a silver bullet for CPI or PCE, and won’t, by themselves, lower mortgage rates soon.

One last tangent because I can’t help myself: the search phrase I keep hearing, “will-tariff-refunds-lower-mortgage-rates-soon”, skips the middle step. Tariff refunds hit category-level goods prices first. Then maybe core feeds through to expectations if the trend is broad and persistent. Then the Fed and term premium react. Miss one link and you’re back where you started. Could rates drift lower on a string of softer prints? Sure. But unless services keep easing, it’s probably a slow grind, not a cliff… but that’s just my take on it.

So, will rates fall soon? A few realistic scenarios for late 2025

Look, I get the question every week, and, yes, I typed the exact search phrase “will-tariff-refunds-lower-mortgage-rates-soon” just to see what’s out there. Our quick screen didn’t turn up hard numbers, literally zero credible SERP hits with quant detail on pass-through timing or size, so we’re back to the usual rate mechanics: inflation trend, growth, term premium, and MBS spreads. Anyway, here’s how I’d frame it for the next three to six months.

• Base case (most likely): modest moves, not miracles. If tariff refunds are limited and the recent disinflation keeps inching forward, Treasury yields probably drift, not dive. Think the 10-year moving in a ±10-25 bp channel into late Q4, with mortgage rates following by maybe 10-30 bps. Why so muted? Because the mortgage-Treasury spread is still elevated versus history, pre-2020 the 30-year mortgage rate typically sat ~150-170 bps above the 10-year; in 2023-2024 it was often closer to 250-300 bps, and it hasn’t fully normalized this year. That gap tends to compress slowly unless volatility falls and origination/convexity flows behave. Translation: you might see a 6-6.75% handle for top-tier 30-year fixed later this year, but it’s probably a grind, probably a slow grind.
• Bull case: broader refunds + cooler prints = measured relief. If refunds hit more categories and we string together softer monthly core readings, especially services, Treasury term premium can compress and MBS spreads can tighten a bit. I’m not talking cliff-dive stuff; I’m talking risk-on that trims, say, another 10-20 bps off the 10-year versus the base case and narrows mortgage-OIS/Treasury spreads by 10-20 bps. On paper, that could push mortgage rates lower by 0.25-0.50 percentage points from early fall levels. You know, finally noticeable on a refi calculator, but not 2020-style fireworks.
• Bear case: sticky services or fresh volatility keeps the lid on. If services inflation stalls out, or we get a new volatility shock that widens spreads, the benefit from any tariff headlines gets muted. The 10-year could back up 25-40 bps, and even if it doesn’t, wider MBS spreads can offset any Treasury rally. End result: mortgage rates flat-to-higher despite friendlier goods prices. I’ve seen this movie; the spread eats your lunch when rate volatility picks up and bank demand stays tepid.

This actually reminds me of 2013-2014 when the spread did a lot of the talking. Actually, let me rephrase that: the spread did all the talking. The thing is, refunds are incremental, helpful at the margin, but we still need services to cool and volatility to calm down for a cleaner move.

My gut, informed by too many cycles: base case wins, incremental, not dramatic. Bull needs softer services and tighter spreads. Bear shows up if services stick or volatility bites.

So, for investors: position for carry and modest duration beta, don’t overpay for a heroic rally. For homebuyers: if you can afford the payment today, don’t wait on a fantasy; if you can wait, monitor the 10-year and the mortgage spread, those two will tell you when the window actually opens.

Your money move: pressure test your housing and rate plan this week

Look, no crystal ball here, just practical stuff you can do in a week. Rates are still choppy, spreads are sticky, and the 10-year hanging in the low-4s keeps the mortgage math. touchy. So, audit your exposure and make decisions you won’t regret three Mondays from now.

If you’re closing soon (next 1-3 weeks): lock vs. float

• Know your lender’s reprice behavior. Ask point-blank: how often do you reprice for the worse in a volatile week, and what’s the typical adjustment? If your lender tends to reprice on a 5-8 bps move in MBS, your “float” risk is higher.
• Run payment sensitivity. Rule of thumb: every 0.25% change in rate moves the payment about $15 per $100,000 borrowed. On $500k, a 0.25% jump is ~/mo; 0.50% is ~$150/mo. If that makes you flinch, lock. If you can stomach it, float with a clear stop, tell your LO to lock automatically if pricing worsens by X bps.
• Use a lock with a float-down if available and priced fairly. Don’t overpay for a float-down unless the window to improve is realistic this week.

Refi-ready checklist, so you can actually act when the window cracks

1. Credit clean-up: pull fresh scores; dispute errors; pay down revolvers to under 30% utilization. A 20-40 point score bump can shift pricing by ~0.125%-0.25% in many rate sheets.
2. Income docs: two years W-2s (or K-1s/returns if self-employed), 30 days of paystubs, 2-3 months of bank statements. Have them in one PDF, underwriters love neat packages.
3. Breakeven math on points/buydowns: points cost typically 1% of loan balance per ~0.25% rate cut, but market sheets change daily. Calculate: total points paid / monthly savings = months to breakeven. If breakeven is 34 months and you may move in 24, skip it.
4. Rate triggers: decide your strike. Example: “I refi at 6.25% if closing costs < $3,500” so you don’t hesitate when pricing shows up for half a day.

Investors: match duration to your horizon

• Bond ladder check: If your spend need is 2-3 years out, keep that tranche in T-bills/1-2yr notes. A 25 bps rate dip trims bill yields quickly, on $1mm laddered cash at 5.0%, a 0.25% drop is ~$2,500/yr in income.
• Rate beta sanity: A 1-2 year duration fund moves ~1-2% for a 100 bps rate shift; a 6-7 year fund can swing ~6-7%. If a 0.50% selloff would upset your quarter, your duration’s too long for your nerves.
• Reinvest policy: pre-commit roll rules (monthly or quarterly) so you’re not guessing after a headline.

Quick reality check on the “tariff refunds drop mortgage rates” idea

We scanned for hard data tied to the keyword “will-tariff-refunds-lower-mortgage-rates-soon” and came up empty, no credible studies or time-series showing a direct, near-term pass-through to MBS coupons. Honestly, I wasn’t sure about this either, but the channel from tariff refunds to the MBS basis just isn’t documented in any way that would help you pick a rate lock this week.

Base case for 2025: incremental shifts, not a hero rally. Spreads matter; volatility matters more. Plan for grind, be glad if you get a gift.

Your challenge for this week, no excuses

• Debt mix: list every balance, rate, and payment. If a HELOC is floating, cap your exposure, refi to fixed or budget for +100 bps stress.
• Emergency fund: 3-6 months expenses in high-yield savings or T-bills. Speaking of which, decide now if you’ll auto-roll bills or keep cash optionality..
• Housing budget: price today’s payment, then add a 0.50% rate shock and a +10% insurance/tax cushion. If it still works, you’re fine. If not, adjust house target or down payment.
• Timeline: buying in October/November? Set lock rules now. If you can wait to Q1, keep documents ready and track the 10-year plus mortgage spread daily; when both improve together you act. I was going to explain exactly how to track the OAS screens, but really the takeaway is, keep a simple watchlist and alerts.

Anyway, do the work today. Don’t wait for a headline to do it for you; headlines don’t make your payment, your plan does.

Frequently Asked Questions

Q: Should I worry about tariff refunds changing my mortgage quote this month?

A: Short answer: no. Tariff refunds are a one‑off, uneven thing and don’t reset the levers that price mortgages. Lenders key off inflation expectations, the 10‑year Treasury, and MBS spreads. Services drive most inflation, and tariff chatter is mostly about goods. Even pass‑through to prices is partial and slow. So, don’t plan your lock around refunds this month.

Q: How do I track the stuff that actually moves mortgage rates?

A: Focus on three dials. One: inflation expectations, watch Treasury breakevens and the 5y5y; if those ease, rates can follow. Two: the 10‑year Treasury yield; your 30‑year fixed tends to move with it. Three: MBS spreads/liquidity, if current‑coupon MBS don’t rally, lenders won’t improve rate sheets. Practical moves: check the 10‑year each morning, skim the Freddie Mac PMMS weekly, and note key data on your calendar (CPI, PCE, jobs, FOMC). If you’re within 30-45 days of closing, set a target, e.g., a small dip in the 10‑year plus a modest lender reprice, and be ready to lock. Don’t get cute on big data days unless you can stomach whipsaws.

Q: What’s the difference between a drop in the 10‑year yield and actual mortgage rates falling?

A: They rhyme, but they’re not twins. Mortgage rates equal the 10‑year yield plus a spread that compensates investors for prepayment and convexity risk, plus some lender capacity and liquidity realities. In 2023, that spread blew out above 300 bps at times (Freddie Mac PMMS vs. U.S. Treasury), way wider than the ~170-200 bps pre‑COVID norm. Even if the 10‑year drifts down, lenders might not pass it through if MBS prices lag, pipelines are stressed, or volatility is high. Also, intraday dips don’t always trigger lender reprices, rate sheets are conservative when markets are jumpy. So, yeah, cheer a lower 10‑year, but confirm MBS are bid and spreads are tightening before expecting a better quote.

Q: Is it better to lock now or float and hope tariff refunds trickle through later this year?

A: If you’re inside 30-45 days of closing, I’m usually team lock, especially in a headline‑heavy stretch like we’ve got this fall. Look, tariff refunds aren’t a rate catalyst. The things that could move your quote are CPI/PCE surprises, jobs data, and the Fed’s tone. Those can cut both ways, and bad timing can add an eighth to a quarter point to your cost overnight. If you lock: ask about a float‑down option; some lenders allow a one‑time improvement if rates drop before docs. Price both 30‑ and 45‑day locks, sometimes the longer lock is only a hair more and buys peace of mind. Run a quick points vs. rate break‑even: divide the extra points by your monthly interest savings to see how many months to recoup. If you float: only do it with a clear trigger and exit. Example: “I’ll float until after CPI; if the 10‑year closes below my level and MBS tighten, I’ll lock the next morning.” Watch for lender pricing windows, reprices often occur mid‑morning and mid‑afternoon. And, please, coordinate with your loan officer on appraisal/underwriting timelines; an expired lock is a fee magnet. Personally, after too many nights watching futures at 3 a.m., I’d lock when the payment works. Hope is not a hedge, and refunds won’t rescue a blown float. Honestly.

@article{will-tariff-refunds-lower-mortgage-rates-not-so-fast,
    title   = {Will Tariff Refunds Lower Mortgage Rates? Not So Fast},
    author  = {Beeri Sparks},
    year    = {2025},
    journal = {Bankpointe},
    url     = {https://bankpointe.com/articles/tariff-refunds-mortgage-rates/}
}