The prediction: when Walmart reports its fiscal fourth quarter in the second half of February 2027, it is likely to guide fiscal 2028 capital expenditure at 3.5% of net sales or higher, rather than stepping back below the level its own chief financial officer described as the peak twelve months earlier. On a net sales base approaching $780bn, that implies roughly $27bn at the low end and comfortably above $30bn at the high end. The company has already moved this goalpost once inside a single fiscal year, and the pattern suggests it is likelier to move again than to snap back.
Three signals observed between early September and the end of September 2026 point the same way. Walmart lifted its own fiscal 2027 capex guide from approximately 3.5% of net sales to approximately 4.0% on 20 August 2026, five months after calling the peak. Costco, on 24 September 2026, guided fiscal 2027 capital spending up roughly 17% and attached precisely the same forward promise of a slowdown that Walmart made and then broke. Meanwhile the largest automation vendor serving this spend carries a contracted backlog that functions as a forward obligation rather than a forecast.
This piece is not a claim that retail capital spending runs away. It is a narrower claim about the reliability of peak-capex language in retail guidance, and about where inside the envelope the money is actually moving. The second part matters more than the headline number, and it is the part most readers of the quarterly releases have skipped.
In short
- The prediction: Walmart likely guides fiscal 2028 capex at 3.5% of net sales or above at its Q4 FY2027 release, expected in the second half of February 2027. No step-down below the originally promised ratio. Base case confidence around 55%.
- Signal 1: Walmart revised FY2027 capex guidance from approximately 3.5% to approximately 4.0% of net sales on 20 August 2026. First-half capex reached $14.181bn against $11.409bn a year earlier, up 24%, while free cash flow fell $1.414bn to $5.529bn.
- Signal 2: Costco guided FY2027 capex to approximately $7.5bn from $6.4bn in FY2026 on its 24 September 2026 call, with 33 planned openings, then promised a slowing rate of growth beyond FY2027. The promise is structurally the same one Walmart has already failed.
- Signal 3: Warehouse automation backlog at the sector’s dominant vendor stood at roughly $22.5bn, with systems in deployment rising from 46 to 77 year over year. Contracted deployment schedules make a sharp Walmart step-down in FY2028 mechanically awkward.
- The companion call that matters more: inside the rising envelope, spend is likely to keep rotating away from distribution-center automation toward store remodels, new stores and international markets, as it did across the first half of FY2027.
Why this matters now
Retail capital expenditure guidance has become a load-bearing part of the equity story, not a footnote. When a retailer tells investors it is at the peak of an investment cycle, it is implicitly promising an inflection in free cash flow, and increasingly promising a buyback or dividend trajectory that depends on that inflection. The guidance language is therefore a commitment with a cash consequence attached.
That makes the February 2026 statement and the August 2026 revision worth taking seriously as a pair. Walmart’s chief financial officer John David Rainey told investors on the 19 February 2026 call that capex would run at about 3.5% of sales and that the company was hitting the peak of annual spending levels on supply chain automation and store remodels. On 20 August 2026 the company told investors it now expected approximately 4.0%, citing its omnichannel growth strategy.
The arithmetic of that change is larger than it sounds. Walmart’s fiscal 2026 capex was $26.642bn on net sales of $706.413bn, a ratio of 3.77%. A 3.5% guide against a sales base growing 4% to 5% would have held dollar spend roughly flat; the 4.0% guide converts a flat year into an increase of roughly $3bn, or about 12%.
So the February guidance did not merely imply a peak. It implied a dollar plateau, and the August revision replaced that plateau with double-digit growth. That is a material change in the cash profile of the largest retailer in the world, delivered in one sentence inside a quarterly release.
For context on how the company has been funding its profit growth while this spend ran, our earlier read on Walmart’s reliance on advertising and membership rather than aisles through FY27 sets out the other half of the picture. The capex and the platform income are two sides of the same strategy, and the platform side is what has made the capex tolerable.
Signal 1: Walmart moved its own capex goalposts in six months
The primary document is the fiscal 2027 second quarter earnings release, covering the quarter ended 31 July 2026 and filed with the Securities and Exchange Commission on 20 August 2026. It carries both the revised guidance and the cash flow detail that explains why the revision was necessary rather than optional. Readers who want the source can consult the filed quarterly earnings release directly.
Three numbers in that release carry the signal. Six-month capital expenditure came in at $14.181bn against $11.409bn in the comparable prior period, an increase of $2.772bn or roughly 24%. Six-month free cash flow fell to $5.529bn from $6.943bn, a decline of $1.414bn, even though operating cash flow rose $1.4bn to $19.7bn.
The second and third numbers together are the tell. Operating cash generation improved materially, and free cash flow still went backwards, because capital spending absorbed the entire improvement and more. That is the signature of a company spending ahead of its own plan rather than one managing down to a target.
Management nonetheless reaffirmed an expectation of double-digit free cash flow growth for the full year. That reaffirmation is doing a lot of work. It requires either an unusually capex-light second half or a working capital tailwind, and it is the single commitment most likely to be quietly revised if the trajectory holds.
| Walmart capex trail | Figure | Capex as % of net sales | Source and date |
|---|---|---|---|
| FY2025 actual | $23.783bn | Approximately 3.5% | Q4 FY2026 earnings release, 19 Feb 2026 |
| FY2026 actual | $26.642bn | 3.77% (on $706.413bn net sales) | Q4 FY2026 earnings release, 19 Feb 2026 |
| FY2027 original guide | Implied roughly $26bn, broadly flat | Approximately 3.5% | Q4 FY2026 call, 19 Feb 2026 |
| FY2027 revised guide | Implied roughly $29bn to $30bn | Approximately 4.0% | Q2 FY2027 release, 20 Aug 2026 |
| FY2027 first half actual | $14.181bn, up 24% year over year | Tracking at or above 3.9% | Q2 FY2027 release, 20 Aug 2026 |
| FY2028 predicted guide | Roughly $30bn or above at the top of the range | 3.5% or higher (base case) | Expected Q4 FY2027 release, second half Feb 2027 |
One honest qualification belongs here rather than in the caveats, because it shapes how the signal should be read. The February remark was scoped, on its face, to supply chain automation and store remodels specifically, not to total capital expenditure. A charitable reading is that the narrow peak was reached and the total rose for other reasons, which is close to what the first-half line detail actually shows.
Signal 2: Costco attached the same forward promise to a 17% increase
On 24 September 2026, on the fiscal fourth quarter call, Costco chief financial officer Gary Millerchip guided fiscal 2027 capital expenditure to approximately $7.5bn against $6.4bn of actual fiscal 2026 spend. That is an increase of roughly 17%, and it is the third consecutive year of what the company itself calls outsized growth. The stated drivers were a growing pipeline of new warehouses plus what Millerchip described as outsized spend on the supply chain.
The opening plan behind the number is concrete and therefore checkable. Costco guided to 33 planned warehouses in fiscal 2027 including five relocations, which nets to 28 new buildings, against a stated long-term run rate of 30 net new warehouses annually. The geographic split included four buildings in Europe, five in Canada and one in Mexico, with the company flagging a stronger Asia, Australia and other international pipeline for fiscal 2028.
The forward promise is the part that rhymes with Walmart. Millerchip told investors that beyond fiscal 2027 the company would expect to see a slowing in the rate of capital expenditure growth following three years of outsized growth that began in fiscal 2025. Structurally that is the same construction Walmart used in February 2026: a current-year number, plus a verbal commitment that the following year moderates.
It is worth being precise about how weak that particular promise is. A slowing in the rate of growth, following a 17% year, is satisfied by any increase below 17%. A fiscal 2028 guide of plus 5% would honor it exactly, which is why the scoreable Costco leg in this piece is whether the $7.5bn number holds rather than whether the slowdown arrives.
Our preview of the quarter in which this guidance landed, covering what the 24 September Costco print had to clear, sets out the sales and earnings bar the company was working against. The capex guide arrived alongside a result strong enough that the spending increase drew little pushback, which is itself informative about how much latitude these commitments carry.
Signal 3: the automation order book is a forward obligation, not a forecast
The third signal sits outside the retailers’ own disclosures, which is what makes it useful. Symbotic, the automation vendor whose revenue has been overwhelmingly concentrated in Walmart work, reported fiscal third quarter results for the period ended 27 June 2026 showing revenue of $721m, up 22% year over year, with net income of $55m against a prior-year loss of $21m. Contracted backlog stood at roughly $22.5bn, with around 15% expected to be realized over the following twelve months.
The deployment counts matter more than the backlog headline. Systems in deployment rose from 46 to 77 year over year, and operational systems generating recurring software and services revenue rose from 42 to 56. Deployment is the stage at which retailer capital actually converts into vendor revenue, so a near-doubling of systems in deployment is a direct read on committed retailer spend in flight.
Per the vendor’s filings, the underlying arrangement covers automation of Walmart’s full set of 42 US regional distribution centers on a contract running to 2037, and Walmart represented roughly 85% of vendor revenue in fiscal 2025. The filings also describe a contingent commitment relating to several hundred automated systems for online pickup and delivery at Walmart stores, reported as capable of adding more than $5.0bn to remaining performance obligation if triggered.
None of that is a spending schedule, and the piece does not treat it as one. What it does establish is that a large slice of the retailer’s automation spend is contractually sequenced across multiple years with a counterparty whose own guidance depends on the cadence. Sharp single-year step-downs are harder to execute against that structure than against a discretionary budget line.
| Signal | Primary source | Date observed | What it shows | What would falsify it |
|---|---|---|---|---|
| Walmart guidance revision | Q2 FY2027 earnings release filed with the SEC | 20 Aug 2026 | Peak language replaced by a 50 basis point ratio increase inside six months | A FY2027 outturn at or below 3.5% of net sales |
| Walmart cash profile | Same release, six-month cash flow statement | 20 Aug 2026 | Capex up 24%, free cash flow down $1.414bn despite higher operating cash flow | A capex-light second half delivering double-digit full-year free cash flow growth |
| Costco FY2027 guide | Q4 FY2026 earnings call remarks | 24 Sep 2026 | Capex up roughly 17% to approximately $7.5bn, plus a verbal slowdown promise for FY2028 | FY2027 capex landing materially below $7.5bn |
| Costco opening pipeline | Same call, FY2027 opening plan | 24 Sep 2026 | 33 openings including 5 relocations, with a larger international pipeline flagged for FY2028 | A cut to the FY2027 opening plan at the Q1 or Q2 FY2027 calls |
| Automation order book | Vendor Q3 FY2026 results and filings | Reported in the period to early Sep 2026 | Roughly $22.5bn backlog, systems in deployment 46 to 77 year over year | Backlog decline or a disclosed deferral of Walmart deployment cadence |
| Divergence check | Kroger Q2 FY2026 results | 11 Sep 2026 | Capex reaffirmed at $3.8bn–$4.0bn while identical sales guidance was cut | Kroger raising its capex envelope above $4.0bn for fiscal 2027 |
What the pattern suggests
Read together, the signals suggest that peak-capex language in US retail currently functions as an aspiration rather than a plan. It is issued at the start of a fiscal year, when the capital committee’s pipeline is least visible to investors, and it is revised mid-year when the pipeline asserts itself. Two of the three largest capital spenders in US physical retail have now issued that language within seven months of each other.
The mechanism is not mysterious. Store remodels, new builds and international openings run on multi-year real estate and construction pipelines that are approved long before the guidance sentence is written. Once a pipeline is committed, the only lever left is timing, and timing slips forward rather than backward when sales are growing.
That is why the base case here is a plateau rather than either a step-down or an acceleration. A plateau at 3.5% to 4.0% of net sales satisfies the internal narrative of discipline while funding a pipeline that has already been approved. It is the path of least resistance for a management team that wants to keep both the growth story and the free cash flow story alive.
| Scenario for Walmart’s FY2028 capex guide | What it looks like | Rough confidence |
|---|---|---|
| Base case: plateau | Guided at 3.5%–4.0% of net sales, rotation away from distribution centers continues | 55% |
| Step-down honored | Guided below 3.5%, framed as the automation build completing | 20% |
| Escalation | Guided above 4.0%, most plausibly on an AI or international push | 15% |
| Unscoreable | Company stops giving a ratio, moves to dollars only or a multi-year range | 10% |
How this prediction is scored
The primary claim scores on the ratio language Walmart uses for fiscal 2028 in its Q4 FY2027 release and call, expected in the second half of February 2027. A guide of 3.5% or higher is a hit; a guide below 3.5% is a miss. If the company abandons ratio guidance entirely, the claim is recorded as unscoreable rather than as a hit.
The companion claims score separately: Walmart’s FY2027 outturn at or above 3.9% of net sales, Costco’s FY2027 capex at or above $7.5bn at its Q4 FY2027 call expected late September 2027, and the rotation claim across the FY2027 line detail.
Wider context: the spend is rotating out of the distribution center
The more interesting finding is not that the envelope grew. It is which lines grew fastest. Across the first half of fiscal 2027, the slowest-growing major line in Walmart’s capital spend was the one most people assume is driving the cycle.
On the reported first-half detail, US supply chain, customer-facing projects and technology accounted for roughly $7.7bn, up about 15%. Store and club remodels came in near $3.6bn, up about 24%. New stores and clubs reached roughly $1.1bn, up about 88%, and international spend was near $1.8bn, up about 48%.
| Walmart H1 FY2027 capex line | Approximate amount | Approximate year-over-year growth | Read |
|---|---|---|---|
| US supply chain, customer-facing and technology | $7.7bn | Up 15% | Largest line, slowest growth. Consistent with an automation build maturing. |
| Store and club remodels | $3.6bn | Up 24% | In-store execution is absorbing incremental capital. |
| International | $1.8bn | Up 48% | Geographic expansion re-entering the capital story. |
| New stores and clubs | $1.1bn | Up 88% | Smallest base, fastest growth. A genuine reversal of the 2016–2020 posture. |
This reconciles two readings that look contradictory. The automation line is decelerating in growth terms even as the total accelerates, which is broadly consistent with our earlier observation that retail logistics capex has been staying roughly flat while automation’s share of it climbs. The cycle has moved on; the money has not stopped.
It also changes what kind of capex this is. Distribution-center automation is a productivity investment with a measurable unit-cost payback. Store remodels and new builds are a demand investment, justified on traffic and basket rather than on cost per case, and historically harder to defend when comparable sales soften.
The vendor landscape is adjusting to the same shift. Our earlier read on the move toward modular automation describes the direction of travel: smaller, faster-deploying systems that fit store backrooms and urban sites rather than only regional distribution centers. That architecture suits a capital cycle rotating toward stores.
One divergence is worth holding onto. Kroger, reporting its fiscal second quarter on 11 September 2026, reaffirmed full-year capital expenditure at $3.8bn–$4.0bn while cutting its full-year identical sales guidance excluding fuel to 0.2%–0.8% from 1% to 2%. The mid-tier grocer is not matching the pace, which means the capital gap between the top two and everyone else is widening rather than closing.
Implications for retailers, brands, vendors and investors
For competing retailers, the operative number is not Walmart’s total. It is the remodel and new-build line, because that is where the competitive pressure lands in a specific trade area within twelve to twenty-four months. A remodel programme growing at 24% is a local market event in a way that a distribution-center upgrade is not.
For brands and suppliers, the rotation changes which readiness work pays off next. Case-pack, labeling and palletization conformance for automated distribution centers remains table stakes, but the incremental requirement over the next two years is likelier to be in-store: reset cadence, planogram changes, fixture compatibility and the merchandising support that remodeled stores demand. Supplier teams organized entirely around distribution-center compliance are likely to be fighting the previous cycle.
For automation vendors and third-party logistics operators, the signal is mixed and should be read as such. The order book is large and contracted, which protects near-term revenue, but the growth rate of the retailer line funding it is the slowest in the portfolio. Vendors whose roadmap is weighted to large greenfield regional systems face a narrower addressable pipeline than those with modular store-adjacent products.
For investors, the mechanical consequence arrives later and lands in the income statement rather than the cash flow statement. A capital wave of this size converts into depreciation over the following several years, which is the dynamic we set out in our note on depreciation becoming a named retail margin headwind by spring 2027. If the FY2028 guide plateaus rather than falls, that headwind extends rather than peaks.
The practical discipline for anyone tracking this is to stop treating the ratio and the dollars as interchangeable. On a sales base growing 4% to 5%, a flat dollar spend produces a falling ratio, and a flat ratio produces rising dollars. Management can tell a discipline story in one unit while spending more in the other, and in February 2027 it will likely be worth checking which unit the guidance is denominated in before reacting to it.
Caveats: what could go wrong
The strongest counter-argument is that nothing was actually broken. The February 2026 peak remark was scoped to supply chain automation and store remodels, and on the first-half evidence the automation line genuinely is the slowest-growing part of the portfolio. On that reading the total rose for unrelated reasons, the narrow peak was reached as described, and the framing of this piece overstates a guidance failure.
That reading has real force, and it is the single most likely reason this prediction looks wrong in hindsight. The counter to it is that remodels, the other named category, grew 24%, which is hard to reconcile with a peak in remodel spending. Both things cannot comfortably be true at once, but the resolution depends on definitions the company has not published.
The second risk is demand. Kroger cut its identical sales guidance in September 2026, and if US grocery and general merchandise volumes soften through the holiday season, both Walmart and Costco acquire a straightforward reason to defer discretionary builds. Capital committees move faster in that direction than in any other.
The third risk is the free cash flow promise functioning as designed. Management reaffirmed double-digit full-year free cash flow growth while first-half free cash flow fell $1.414bn, which implies a disciplined second half. If that discipline materializes, the FY2028 guide has an easier path back below 3.5% and the step-down scenario becomes the base case.
The fourth risk is distortion from the tariff refund cycle, which has been moving unusual amounts of cash through retail balance sheets. A capex year part-funded by one-off refunds is not a structural capex year, and we have argued separately that the refund windfall is likelier to end in buybacks than in price cuts. Capital allocated from a windfall is the first thing cut when the windfall stops.
The fifth risk is definitional and the most boring, which is usually a sign it deserves attention. If Walmart stops guiding capex as a percentage of net sales and moves to a dollar range or a multi-year framing, the primary claim becomes unscoreable rather than wrong. That is a genuine possibility and is carried at roughly 10% in the scenario table above.
Finally, the Costco leg should not be overweighted. A company that has just guided to 33 openings has a visible pipeline, and the $7.5bn figure is a plan rather than a commitment. Openings slip routinely for permitting and construction reasons that say nothing about strategic intent.
FAQ
What exactly is being predicted, in one sentence?
That Walmart’s fiscal 2028 capital expenditure guidance, expected at its Q4 FY2027 release in the second half of February 2027, likely comes in at 3.5% of net sales or higher rather than below it. In dollar terms on a growing sales base, that points to roughly $30bn or more at the upper end of the range.
Is this just saying capex goes up every year?
No, and the distinction matters. Walmart’s own February 2026 guidance of approximately 3.5% implied roughly flat dollar spend against the $26.642bn it spent in fiscal 2026, so a step-down was the stated plan. The prediction is that the stated plan does not survive, not that spending rises because spending usually rises.
Could the February 2026 statement simply have been misread?
That is the most serious objection and it is addressed directly in the caveats. The remark was scoped to supply chain automation and store remodels, and the automation line is in fact the slowest-growing part of the first-half spend. The weakness in that defence is that remodels, the other named category, grew about 24%.
Why does the rotation inside the capex envelope matter more than the total?
Because the two kinds of spend behave differently and affect different counterparties. Distribution-center automation is a cost-per-unit investment with a defensible payback, while remodels and new stores are demand investments that compete locally and are more exposed to soft comparable sales. A supplier or competitor reading only the total will misjudge where the pressure arrives.
What would make this prediction clearly wrong?
A fiscal 2028 guide below 3.5% of net sales, framed as the automation build completing and the remodel programme normalizing. A fiscal 2027 outturn at or below 3.5% would also undercut the premise, since it would suggest the August revision was conservative framing rather than a real increase.
How does Costco fit if its capex ratio is much lower than Walmart’s?
Costco is included for the structure of its guidance language, not for the level of its spending. The relevant comparison is the growth rate, roughly 17% for fiscal 2027, paired with a verbal promise that growth slows afterwards. That is the same construction Walmart used, which is what makes it a second independent observation rather than a repeat of the first.
Why use an automation vendor’s backlog as a signal rather than retailer disclosure?
Because it is independent of the retailers’ own framing and harder to adjust rhetorically. Systems in deployment rising from 46 to 77 year over year is a physical measure of committed spend in flight, and contracted multi-year sequencing constrains how quickly a single-year step-down can be executed. It is a constraint on the retailer’s freedom of action, not a forecast of its intent.
Does Kroger’s flat capex contradict the thesis?
It complicates the sector-wide version of the thesis, which is why the thesis is written about Walmart and Costco specifically. Kroger reaffirming $3.8bn–$4.0bn while cutting its sales guidance is evidence that capital discipline is available to retailers who want it. The reasonable conclusion is divergence between the top two and the rest, not a uniform capex wave.
When can a reader check the outcome?
There are three useful dates. Walmart’s Q3 FY2027 results, expected in the second half of November 2026, show whether the capex-light second half required by the free cash flow promise is materializing; the Q4 FY2027 release in the second half of February 2027 carries the scoreable FY2028 guide; and Costco’s fiscal 2027 outturn arrives at its Q4 call expected in late September 2027.