The tariff refund windfall that lifted second-quarter retail earnings is likely to show up next in the government’s price data, not just in corporate margins. The signals from the August and September earnings round point to US core goods inflation, the CPI series for commodities less food and energy, printing a negative year-over-year reading by the November 2026 report, due in mid-December. Three independent tells support the call: how the largest retailers say they are deploying more than $4 billion of refunded IEEPA duties, the scale and pace of the refund program itself, and a tariff schedule that has already stepped down from the peak of early 2026. The base case is modest goods deflation into the holiday quarter; the risk case is that fuel costs and margin repair absorb the money before it reaches the shelf edge.
In short
- Prediction: the CPI index for commodities less food and energy is likely to print at or below 0.0% year over year in the November 2026 report (released mid-December), down from +0.7% in August. Apparel, the most tariff-exposed line, is likely to slow from 3.6% to below 2%.
- Signal 1, earnings deployment language: Walmart is putting a $2.9 billion refund into price with more than 11,000 rollbacks and told investors the effect lands in Q3; Burlington reinvested its full $55 million; Target and Kohl’s list lower prices and customer value among the uses of $994 million and roughly $150 million respectively.
- Signal 2, refund scale and pace: CBP reported roughly $106.6 billion of IEEPA refunds completed by August 21, 2026, out of about $166 billion collected, with a congressional letter on September 9 pushing for the rest. The cash is arriving in the quarter that sets holiday prices.
- Signal 3, the rate step-down: the trade-weighted US tariff rate fell from about 15.3% under IEEPA to roughly 11.4% under Section 122, and the Section 301 replacement that took effect July 24 carries broad exemptions. Landed cost in H2 is lower than the cost embedded in H1 shelf prices.
- Main caveat: energy is up 16.3% year over year and gasoline 27.4%, and department stores are steering refunds to marketing rather than price. If fuel surcharges and margin repair win, core goods could stall near +0.5% instead of crossing zero.
Why this matters now
The second-quarter reporting season for US retail closed with an unusual pattern: strong profit beats, raised full-year guidance, and soft top lines. Goldman Sachs’ retail team noted on September 8 that 64% of hardlines and broadlines retailers raised their FY26 EPS guidance at the midpoint, and it attributed part of the improvement to tariff refund windfalls flowing from the Supreme Court’s IEEPA ruling. That combination, more profit on weaker sales, is the fingerprint of a one-off cost reversal rather than a demand recovery.
What matters for the next 90 days is where that reversal goes. A refund is not income in the ordinary sense; it is a return of duties that retailers either absorbed in margin or passed to shoppers during 2025. Companies can keep it, share it with vendors, spend it on marketing, or hand it back through price. The distribution across those four buckets is now visible in the earnings transcripts, and it tilts toward price at the mass-market end of the industry, which is exactly where the CPI basket carries the most weight.
The timing is also unusually clean. Refund cash arrived between May and August, the tariff schedule stepped down in February and again in July, and the holiday quarter starts in October. The three forces line up in the same direction for the first time since tariffs began lifting goods prices in mid-2025, which is why the prediction is about a specific print rather than a general direction.
Signal 1: earnings deployment language points to price, not buybacks
The clearest tell comes from Walmart’s August 20 release for the quarter ended July 31. The company received nearly $2.9 billion in IEEPA refunds and said it prioritized investment in price during the period. Walmart US delivered more than 11,000 rollbacks in the quarter, against roughly 7,200 in the prior quarter and a typical run rate near 5,000.
Management told investors the pricing impact would be visible in the third quarter, and CEO John Furner framed it plainly: the company is investing heavily in price because customers need it to. With Walmart’s grocery and general merchandise categories running in a 3% to 4% comp range, a rollback count more than double the norm is a material deflationary impulse for the largest US retailer’s basket.
Off-price followed the same logic. Burlington’s August 27 report showed the full $55 million refund reinvested in price cuts, and the company’s third-quarter guidance missed consensus as a result; the details are in our earlier coverage of how Burlington spent its $55 million refund on prices. That is the second-largest off-price chain choosing lower sticker prices over a cleaner earnings print at the moment its Q3 guide was under scrutiny.
Target and Kohl’s are more mixed but still lean toward the consumer. Target recognized a $994 million pretax refund in its second quarter, worth $1.65 per share and roughly 370 basis points of gross margin, and raised full-year net sales growth guidance to about 5%. Its stated uses include store remodels, merchandising, technology and lower prices, alongside the rollout of Target Beauty Studio to more than 600 stores. Kohl’s August 26 release described its roughly $150 million refund as partly recorded as a reduction of inventory, partly shared with vendor partners, and partly invested to deliver greater value to customers, with about $100 million running through cost of merchandise sold.
The department-store end of the spectrum is where the pattern weakens.
Macy’s September 10 call disclosed $116 million of refunds, all now received, and CFO Thomas Edwards said only about $20 million, or five cents a share, flows to the bottom line. The remainder goes to brand building, accelerated reimagine-store pilots, 2027 program expansion, strengthening the value proposition, and offsetting fuel headwinds. Bath & Body Works, per the Goldman Sachs note, is directing roughly $30 million of its $80 million refund to offset tariff and input cost inflation and about $35 million to marketing. Neither is a price cut in the CPI sense, and both are a reminder that the deflationary impulse is concentrated in mass and off-price, not in the mall.
The signals matrix
| Retailer | IEEPA refund disclosed | Stated deployment | Timing of consumer effect | Source (date) |
|---|---|---|---|---|
| Walmart | ~$2.9 billion | Price investment; 11,000+ rollbacks in Q2, grocery and general merchandise prioritized | Q3 FY27 (Aug to Oct 2026) | Q2 FY27 release and call (Aug 20) |
| Target | $994 million pretax ($1.65/share) | Remodels, merchandising, technology and lower prices; Beauty Studio in 600+ stores | H2 2026 | Q2 2026 release (Aug 19 to 20) |
| Kohl’s | ~$150 million (~$100m via COGS) | Inventory reduction, vendor sharing, customer value initiatives | Q3 to Q4 2026 | Q2 2026 8-K (Aug 26) |
| Burlington | $55 million | Fully reinvested in price cuts; Q3 guide missed | Q3 2026 | Q2 2026 release (Aug 27) |
| Macy’s | $116 million | ~$20m to bottom line; rest to brand building, store pilots, value, fuel offset | Marketing-led, limited price effect | Q2 2026 call (Sep 10) |
| Bath & Body Works | $80 million | ~$30m offsets input inflation, ~$35m marketing | Margin and marketing, not price | Goldman Sachs retail note (Sep 8) |
Signal 2: the refund program is large, fast, and landing in the holiday-planning quarter
The second tell is fiscal rather than corporate. According to the US Customs and Border Protection status report dated August 21, 2026, roughly $132.5 billion of potential and certified IEEPA refunds had been accepted into the CAPE processing system, and about $106.6 billion in duties and interest had been completed and sent to Treasury for payment. Against total IEEPA collections of about $166 billion, that means roughly two thirds of every dollar of invalidated duty had already been returned to importers before Labor Day. CBP publishes the program details on its IEEPA duty refunds page.
The cadence matters as much as the total. CAPE Phase 1 opened on April 20, Phase 2 added reconciliation-flagged entries on June 29, and CBP has held to a 60-to-90-day processing window after acceptance. That schedule explains why the refunds cluster in retail fiscal quarters ending in late July and early August, and it implies a further tranche landing in the quarters ending in October, when holiday price files are locked. Congress is pressing for that tranche to arrive on time: a September 9 letter from House members urged CBP to fix account-approval delays that have left some importers unable to meet refund deadlines, as covered in our report on the 90-day refund deadline.
Two features of the program push the money toward price rather than balance sheets. First, refunds go to the importer of record, which for the largest retailers is the retailer itself; there is no intermediary to capture the spread. Second, the refunds are of duties that were, at least partly, passed through to shelf prices in 2025, which means the competitive baseline has shifted: a retailer that pockets the refund while a rival reinvests it is visibly more expensive on like-for-like items during the most price-transparent quarter of the year. That is a difficult position to defend in mass merchandise, and Walmart’s rollback count suggests the largest player has already moved.
Signal 3: the tariff schedule has stepped down twice since February
The third tell is the rate itself. Global Trade Alert’s estimates put the trade-weighted average US tariff at about 15.3% under the IEEPA regime that ended February 20, 2026, and at roughly 11.4% under the flat 10% Section 122 surcharge that replaced it from February 24. Section 122 was explicitly temporary and expired after 150 days on July 24. The Section 301 forced-labor tariffs that took effect the same morning run at 10% or 12.5% depending on the partner economy, with a net-of-MFN structure for the EU, Taiwan, Japan, Korea and Switzerland, and a broad set of exemptions covering raw materials, pharmaceuticals, civil aircraft, Section 232 goods, and anything that qualifies for an existing free trade agreement, including USMCA-compliant goods from Canada and Mexico.
The implication for landed cost is straightforward. Goods ordered in the second half of 2025 arrived under IEEPA reciprocal and fentanyl-related rates that were in many cases 20% or higher; goods arriving now face a 10% to 12.5% additive rate on most partners, with meaningful carve-outs. China is the exception, at a 37.5% aggregate after the new 12.5% stacked on the existing 25%, and the pending overcapacity action could add more.
But China’s share of US retail sourcing in apparel, footwear and home has fallen for six years, and the Section 301 partner list rewards exactly the Vietnam, Bangladesh, India and Central America sourcing that replaced it. The apparel line is the cleanest test: USTR’s textile quota mechanism has kept the duty at 10% for most importers, as we explained when USTR missed its textile quota target, and apparel CPI has already slowed from 3.9% in July to 3.6% in August with a flat month-over-month reading.
The August CPI report, released September 11, shows the step-down beginning to register. Commodities less food and energy rose 0.1% on the month and 0.7% over 12 months, new vehicles were up just 0.6% year over year, used vehicles were down 2.3%, and several sell-side desks estimated that core goods excluding vehicles fell about 0.15% on the month, led by household goods and apparel. Those are the categories where tariff pass-through peaked in late 2025, which means the base effects turn favorable for the year-over-year comparison from September onward.
What the pattern suggests
Put the three signals together and the arithmetic of the prediction becomes tractable.
Core goods CPI stood at +0.7% year over year in August. For the November print to reach 0.0% or below, the index needs to underperform its September-to-November 2025 path by about 0.7 percentage points cumulatively. In the same three months of 2025, tariff pass-through was running at its strongest, with core goods rising on the order of 0.2% to 0.3% per month.
If the 2026 months print flat to slightly negative, which is roughly what August already delivered on an ex-vehicles basis, the year-over-year rate crosses zero by November through base effects alone. Any incremental price investment from the refund tranche pushes it lower.
The precedent also favors the call. Core goods CPI was negative year over year for most of the period from late 2023 through the first half of 2025, as pandemic-era freight and inventory distortions unwound; it was tariffs that lifted the series back above zero. A partial reversal of the tariff shock, combined with cash refunds of the earlier duties, is a return toward the pre-tariff regime rather than a novel deflation. The pattern suggests the market has been anchored on the 2025 pass-through story and is under-weighting the reversal now in motion.
Prior precedents
| Episode | Shock | Core goods CPI response | Read-across to Q4 2026 |
|---|---|---|---|
| 2015 to 2016 dollar strength | Import prices fell on a stronger dollar | Core goods negative year over year for most of the period, near -0.5% | Shows that a modest fall in landed cost is enough to hold the series below zero |
| February 2020 List 4A reduction | China tariff on ~$120bn of goods cut from 15% to 7.5% | Little visible effect before the pandemic overwhelmed the data | Cautionary: a single-partner cut can be swamped by other shocks |
| Late 2023 to mid-2025 freight normalization | Container rates and inventory gluts reversed | Core goods fell to roughly -1% year over year at the trough | The regime the series is likely reverting toward |
| 2025 IEEPA pass-through | Trade-weighted tariff rose to ~15% | Core goods turned positive; apparel to nearly 4% | The comparison base that now flips favorable from September |
Scenarios
| Scenario | Rough probability | Core goods CPI, November 2026 (y/y) | What drives it |
|---|---|---|---|
| Base case: reversion below zero | ~55% | -0.3% to 0.0% | Walmart and off-price price investment lands in Q3, base effects from 2025 pass-through, Section 301 exemptions lower landed cost |
| Stall near current level | ~30% | +0.3% to +0.7% | Fuel surcharges and freight offset refund-funded cuts; department stores hold price; memory-driven electronics inflation persists |
| Re-acceleration | ~15% | above +0.7% | China overcapacity tariff takes effect on retail categories, the September 15 court test unsettles Section 301, energy pass-through broadens into goods |
Wider context: a $1 trillion holiday built on price, not volume
The prediction sits awkwardly against the season’s headline forecast. Bain’s projection of the first $1 trillion US holiday attributes most of the growth to inflation rather than unit demand, a point we examined in our look at the Bain holiday forecast. Both views can be right at once: headline CPI was 3.4% in August with energy up 16.3%, so total nominal spending can rise on fuel, food and services while the goods basket that fills shopping carts gets cheaper. The distinction matters for retailers because comp-store sales reported in dollars will understate unit growth if goods deflation materializes, and it matters for investors because the market tends to read negative goods CPI as a demand problem even when it is a cost pass-back.
The broader macro setting reinforces the divergence. The August retail sales report due September 16 lands on the same day as a Federal Reserve decision, and the consumer-facing picture is one of persistent inflation in energy and services alongside a cautious, value-seeking shopper. Retailers that cut goods prices in that environment are not being generous; they are responding to a shopper whose fuel bill has crowded out discretionary spend. That dynamic is why Walmart pairs its rollbacks with grocery, where the volume response to price is fastest.
There is also a policy feedback loop worth watching. Negative core goods inflation in the fourth quarter would land as the administration weighs further Section 301 actions and as the Section 122 litigation and forced-labor tariff challenges move through the courts. A visible consumer dividend from tariff reversal is a political argument as much as an economic one, and it is likely to feature in how both retailers and trade groups frame the next round of tariff decisions.
Implications for retailers, brands and investors
For mass and off-price retailers, the signals point to a holiday in which price leadership is set by the two or three largest players and everyone else follows. Walmart’s rollback count and Burlington’s full reinvestment establish a competitive floor that mid-tier general merchants will find hard to ignore in the same categories. The practical question for a regional chain or a mid-market department store is whether to match on a narrow basket of high-visibility items and protect margin elsewhere, or to hold price and accept share loss in the price-transparent categories.
For brands and vendors, the Kohl’s disclosure is the one to study. A refund shared with vendor partners is, in effect, a retroactive renegotiation of 2025 cost increases, and it sets a precedent for how the next tariff shock will be split. Brands that raised wholesale prices in 2025 on the strength of tariff pass-through should expect retailers to come back for a share of any refund the brand itself received as importer of record. The vendor-terms conversation in Q4 is likely to be about refunds, not just markdowns.
For investors, the earnings implication runs opposite to the CPI implication. Refund-funded price cuts flatter unit volume and share but depress gross margin in the quarters after the refund is recognized, which is why Burlington’s third-quarter guide missed and why Walmart’s third-quarter outlook explicitly carries the pricing actions.
The likely pattern in the November reporting round is gross margin rates that step down sequentially even as comps hold, and a wave of questions about how much of the refund benefit is durable. Companies that used the money for marketing or store investment, like Macy’s, will show cleaner margins and weaker price competitiveness. Neither profile is obviously better; they are different bets on the same shopper.
Caveats: what could go wrong
The most serious counter-signal is energy. Gasoline is up 27.4% year over year and the energy index 16.3%, and Macy’s explicitly earmarked part of its refund to mitigate fuel headwinds. If diesel and parcel surcharges keep rising into peak season, retailers may use refund cash to absorb logistics costs rather than cut shelf prices, and some of that cost could pass through to goods. That is the core of the case that fuel and energy likely outrank tariffs in November retail guidance, and if it plays out, core goods could stall in the +0.3% to +0.7% range rather than crossing zero.
The second caveat is category composition. The CPI goods basket is not evenly exposed to the refund story.
Electronics, a large holiday category, is facing a memory cost shock that points to shallower rather than deeper discounts; our analysis of why holiday electronics discounts are likely to shrink in 2026 is a direct counterweight to this piece. New vehicles, which carry heavy weight in core goods, are governed by Section 232 rather than IEEPA and saw no refund. If apparel and household goods fall while electronics and vehicles hold, the aggregate could land just above zero even if the retail price cuts happen exactly as described.
Third, the deployment evidence is skewed toward the companies that have reported. Department stores and specialty apparel chains that report in September and later may lean further toward marketing and margin repair than Walmart and Burlington did, and the aggregate could look less price-led once the full sample is in. Fourth, the tariff schedule is not settled: the China overcapacity action and the pending court challenge to Section 301 forced-labor tariffs could raise rates on retail categories before year-end. And finally, statistical noise is real; a single month of methodology quirks or seasonal adjustment can move core goods by 0.2 points, which is a large share of the gap this prediction needs to close.
FAQ
What exactly is the prediction, and how would a reader check it?
The prediction is that the BLS CPI series for commodities less food and energy prints a 12-month change at or below 0.0% in the November 2026 report, which the BLS is scheduled to release in mid-December. A secondary expectation is that the apparel index slows to below 2% year over year in the same report, from 3.6% in August. Both figures appear in the BLS CPI summary table, so the check is a two-line lookup.
Why not predict headline or core CPI instead?
Because the signals are about goods, not services or energy. Headline CPI is being driven by a 27.4% rise in gasoline and 16.3% in energy overall, and core CPI at 2.4% is dominated by shelter and services. The refund and tariff signals bear only on the goods basket, so the falsifiable claim has to be about that series.
Is the tariff refund really being passed to shoppers, or is that corporate messaging?
The best evidence is behavioral rather than verbal. Walmart’s rollback count more than doubled from its usual run rate and its Q3 outlook explicitly carries pricing actions; Burlington’s Q3 guide missed because it reinvested the full refund. Those are costly signals that are hard to fake. Macy’s and Bath & Body Works, by contrast, are being explicit that most of their refund goes to marketing and cost offsets, so the pass-through is real but uneven.
Could new tariffs reverse the effect before November?
Yes, and it is the main upside risk to goods prices. The Section 301 forced-labor tariffs stack with Section 232 duties, a China overcapacity action is pending, and courts are testing the July 24 regime. However, most retail goods now face 10% to 12.5% rates with broad exemptions versus the higher IEEPA rates embedded in 2025 costs, so a reversal would need a large and fast new action to show up in a November print.
Doesn’t Bain’s $1 trillion holiday forecast contradict goods deflation?
Not necessarily. Bain attributes most of the projected growth to inflation, and inflation in the total retail sales figure includes food, fuel and restaurant spending. Nominal spending can rise on those lines while the general-merchandise basket measured in core goods CPI falls. The two views are about different aggregates.
What about the argument that fuel costs will absorb the refund?
It is the strongest counter-argument and the reason the base-case probability is only around 55%. Macy’s has already earmarked part of its refund for fuel headwinds. If parcel and diesel surcharges keep rising, retailers could spend the refund defending margin against logistics costs, leaving less for price. The signal to watch is whether Walmart’s Q3 gross margin commentary in November still references price investment or shifts to freight.
How much of the $166 billion in IEEPA duties went to retailers?
CBP does not publish a sector split. The disclosed retail refunds in this earnings round, including Walmart’s $2.9 billion, Target’s $994 million, Kohl’s roughly $150 million, Macy’s $116 million, Bath & Body Works’ $80 million and Burlington’s $55 million, sum to about $4.2 billion, which is a small share of the total. Most refunds went to industrial importers, distributors and brands acting as importer of record, and how those firms deploy the money is harder to observe.
Why does the timing point to the November CPI rather than October or December?
Walmart said its price investment lands in the quarter ending October 31, CBP’s 60-to-90-day processing cadence puts the next refund tranche in September and October, and holiday price files are typically locked by late October. November is the first full month in which all three effects are in shelf prices, and its year-over-year comparison is against the strongest months of 2025 tariff pass-through.
What would make this prediction fail cleanly?
A November core goods print above +0.3% year over year would indicate that the refund-funded price cuts were either smaller or more concentrated than the earnings language implies, or that offsetting cost pressures dominated. A print between 0.0% and +0.3% would count as directionally right but not confirmed. Anything at or below zero confirms the call.