Core-on-Core Reverses Up, Tariff Bracketed in LIFO, Multiple Still Full — Maintaining Hold
Initial Read: The two near-term concerns we flagged at Q2 — core-on-core margin compression and unbracketed tariff exposure — both improved this quarter. Core-on-core reversed to +36bps; tariff impact landed mostly inside a $130M LIFO charge with $40–$50M sized for Q4. The franchise re-validated; the multiple still does not leave room.
Key Takeaways
- Rating: Maintaining Hold. The bear concerns we held when initiating at Q2 FY25 each retreated this print — core-on-core margin reversed up, tariff impact has been quantified inside the LIFO mechanism, and the membership-fee deferred tailwind continues to compound. We agree the cycle thesis is constructive. But the stock has rallied back into print on the same multiple that gave us pause, and the largest piece of the deferred-fee benefit is still ahead in Q4 FY25 / Q1 FY26 — a known catalyst the market is already pricing. Hold remains the honest call until either the multiple resets or Q4 confirms the deferred-fee landing at the magnitude management has signaled.
- Clean operating beat across the board. Net income $1.90B, EPS $4.28 (+13.2% Y/Y) against ~$4.24 consensus. Revenue $63.21B. Operating income up materially Y/Y. The beat absorbed a $130M LIFO charge (vs. an $11M credit prior year, ~23bps gross-margin headwind), a $40M one-time vacation-accrual catch-up, and a $35M ($0.08/share) FX translation drag. Underlying operational momentum is clean.
- Comp framework intact; mix slightly cooler. Total comps +5.7% (+8% adjusted for FX and gas deflation). US +6.6% (+7.9% ex-gas); Canada adjusted +7.8%; other international adjusted +8.5%. E-commerce +14.8% (+15.7% adj) — decelerating from Q2’s +20.9% as bullion and gift cards lap tougher prior-year comps, exactly as Millerchip pre-flagged. Worldwide traffic +5.2%; US traffic +5.5%. The traffic story is unchanged; ticket dynamics now slightly cooler on adjusted basis vs. Q2 (worldwide adj ticket +2.7% vs. +3.2%).
- Core-on-core margin reversed to +36bps — the Q2 concern resolved this quarter. Reported gross margin +41bps (+29bps ex-gas); core-on-core +36bps. Drivers: (a) fresh productivity and lower spoilage at higher sales leverage; (b) deflation tailwinds in butter, eggs, dairy, olive oil flowing through faster than COST passed savings back to members. Millerchip’s framing — “during times of inflation we feel margin pressure as we keep prices low; the opposite is often true” — is the right philosophical anchor. We cannot extrapolate this magnitude (some drivers were quarter-specific), but it does close the door on the Q2 thesis-fissure question.
- Tariff impact is now bracketed — mostly via LIFO. $130M LIFO charge (~23bps headwind) reflects non-foods inflation re-emerging primarily from imported items; ~1/3 of US sales are imported with ~2/3 of imports in non-foods, and China-sourced items are ~8% of total US sales. Q4 FY25 will absorb another $40–$50M LIFO charge at current run-rate inflation. Management is mitigating actively: rerouted high-tariff goods to non-US markets, pulled forward summer/patio inventory, sourcing more domestically (mattresses, pillows, plastic resin, US outdoor furniture), localized KS production (Asia laundry, China water). The net effect is that tariff is now a known $140–$180M FY25 LIFO drag rather than an unbracketed risk — better than Q2’s framing.
- Membership engine compounds; renewal rate slipped 30bps on a known cohort effect. Membership fee income $1.24B (+10.4% Y/Y, +11.4% ex-FX); the Sept 2024 fee increase contributed ~4.6% of fee income (vs. ~3% at Q2). US/Canada renewal 92.7% (-30bps Q/Q); worldwide 90.2% (-30bps Q/Q). Both slips are explained by the fall-2023 Groupon promotion cohort entering the renewal calculation for the first time, plus higher digital sign-up penetration at a slightly lower renewal rate. Paid households 79.6M (+6.8%); cardholders 142.8M (+6.6%); executive members 37.6M (+9%), now 47.3% of paid base / 73.1% of worldwide sales. We treat the renewal-rate slip as cohort-driven noise rather than thesis damage but watch it.
- Stock reaction +4% on May 30. The market read what we read — the print reversed the Q2 concerns. The buy-side bid into the Q4 deferred-fee tailwind plus the constructive tariff mitigation framing is the path-of-least-resistance trade. We respect that read; we just do not think the multiple leaves room to play it.
Rating Action
This print maintains the Hold rating we initiated at Q2 FY25, with a meaningfully constructive shift in the underlying read of the franchise.
- Q2 FY2025 (Initiating at Hold): Best-in-class franchise quality, but core-on-core margin compression (-8bps), tariff exposure unquantified, and a multiple already pricing in continued operational beats — we wanted to see how 2H FY25 navigated.
- Q3 FY2025 (Maintaining Hold): The two near-term concerns we held both improved: core-on-core flipped from -8bps to +36bps; tariff is now bracketed inside a known LIFO mechanism (~$130M Q3, $40–$50M Q4 incremental) with active supplier and sourcing mitigation. The membership-fee deferred tailwind continued to step up (3% → 4.6% of fee income) and the largest tranche is still ahead in Q4 FY25 and Q1 FY26. Operationally the thesis-validating print we were waiting for. The reason we hold rather than upgrade: the stock has already moved into print on the same valuation that gave us pause, and the deferred-fee tailwind is a known catalyst the market is pricing today — we cannot underwrite the next 12 months delivering above the S&P 500 from an entry point that already discounts a clean Q4. Outperform requires either a multiple reset on a softer Q4 surprise or evidence that tariff absorption stays inside the LIFO mechanism without forcing core-on-core compression in the back half.
Results vs. Consensus
A clean operational beat that absorbed three discrete drags (LIFO, vacation accrual, FX translation) without breaking the EPS line. Underlying composition is the cleanest print since the September 2024 fee increase took effect.
| Metric | Q3 FY25 Actual | Y/Y | Consensus | Color |
|---|---|---|---|---|
| Net Sales | $61.96B | +8.0% | ~$62.0B | In line / slight beat |
| Total Revenue | $63.21B | +8.0% | ~$63.19B | Slight beat |
| Diluted EPS | $4.28 | +13.2% | ~$4.24 | +$0.04 above |
| Net Income | $1.90B | +13.1% | n/a | Up from $1.68B |
| Total Comp Sales | +5.7% / +8% adj | n/a | ~+5.5% | Adjusted in line with Q2 trend |
| US Comp Sales | +6.6% / +7.9% ex-gas | n/a | ~+6% | Solid |
| E-commerce Comps | +14.8% / +15.7% adj | n/a | ~+13% | Decelerating from Q2 +20.9% on bullion/gift card laps |
| Membership Fee Income | $1.24B | +10.4% / +11.4% ex-FX | ~$1.22B | Sept 2024 fee +4.6% contribution (vs. 3% Q2) |
| Reported Gross Margin | 11.25% | +41bps | n/a | Core-on-core +36bps reversed Q2 |
| SG&A Rate | 9.16% | +20bps (worse) | n/a | Wage investment + vacation accrual catch-up |
| LIFO Charge | $130M | vs. $11M credit LY | n/a | Non-foods inflation; tariff-driven |
Worth bracketing: the $40M one-time vacation-accrual catch-up rolls into both gross-margin (manufacturing/supply-chain employees) and SG&A (warehouse/operations) lines. Net interest and other was $50M (vs. $87M LY), with FX the principal driver of the Y/Y improvement. Tax rate normalized at 26.2% (vs. 26.4% LY). The clean apples-to-apples operating-income line is the cleanest single metric in the print.
Segment / Geographic Performance
United States — Strong, Productivity-Led Margin
- US comp sales +6.6% (+7.9% ex-gas deflation). Traffic +5.5%; ticket +1.1% reported (+2.3% adjusted).
- Two of the all-time highest US gallon weeks during the last month, attributable to extended gas-station hours, new station openings, and lower pump prices.
- Fresh productivity and lower spoilage on higher sales leverage drove most of the core-on-core gross-margin expansion. Croissant, butter, eggs, dairy, olive oil cited as specific margin-relief examples.
Canada — Healthy, FX Headwind Persisting
- Reported comp +2.9%; adjusted for gas deflation and FX, +7.8%. Constant-currency growth is the right read; reported is FX-distorted.
- Sweden second warehouse, Korea twentieth warehouse, and Canada 110th warehouse all opening in Q4 FY25 — the international cadence remains roughly half of the 27 net new openings planned for FY25.
Other International — All Markets Positive Comps
- Reported comp +3.2%; adjusted +8.5%. Millerchip explicitly noted that all international countries had positive comp growth in the quarter — a clean disclosure that was not the case in the prior-year setup.
- Localized production deepening: KS Ultra Clean laundry products now sourced in Asia for APAC warehouses, lowering member prices ~40% in the region.
E-Commerce — Cooling, Inflection Intact
- E-commerce comps +14.8% (+15.7% adj), decelerating from Q2’s +20.9% on bullion and gift-card laps.
- Gold/jewelry, toys, health and beauty, majors, housewares, small electrics, apparel all double-digit Y/Y.
- Costco Logistics: items delivered +31% Y/Y. Now ~20-25% of total deliveries; ~80-85% of big-and-bulky.
- Costco Next: Q3 FY25 sales matched the entirety of FY22 sales for the platform. Still ramping the vendor pipeline.
- Affirm BNPL launched. White-label partnership offering exclusive member-pricing on big-ticket. Management notes some members were already using Affirm for COST purchases outside the ecosystem — this brings them into the funnel with native pricing.
- Digital is ~8% of business on COST’s strict definition (excludes Instacart and travel); ~10%+ on broader peer comparison; ~12% if gas is excluded from total.
Membership — Slight Slip on a Known Cohort Effect
- Paid household members: 79.6M, +6.8% Y/Y.
- Cardholders: 142.8M, +6.6% Y/Y.
- Executive members: 37.6M, +9.0% Y/Y. Now 47.3% of paid base; 73.1% of worldwide sales.
- Renewal rates: US/Canada 92.7% (-30bps Q/Q); worldwide 90.2% (-30bps Q/Q). Driven by fall-2023 Groupon promo cohort entering the calculation for the first time, plus rising digital sign-up share at slightly lower renewal. Millerchip explicitly characterized digital members as “a net positive, as they grow the overall membership base and are generally younger members, but they also renew at a slightly lower rate.”
- Membership fee income: $1.24B (+10.4%, +11.4% ex-FX). Sept 2024 fee contribution ~4.6% of fee income, up from ~3% at Q2. Largest deferred-accounting impact still ahead in Q4 FY25 / Q1 FY26.
Key Topics & Management Commentary
Tariffs: Bracketed Inside LIFO, Mitigated Across Five Vectors
The biggest net-positive in the print versus Q2 framing. Vachris detailed the active mitigation playbook:
“During the third quarter, we rerouted many goods sourced from countries with large tariff exposure to our non-U.S. markets. In the U.S., we pulled forward some items that we had planned for the summer and sourced additional locally produced goods to reduce tariff impacts and ensure that we were in stock. ... We continue to move more Kirkland Signature product sourcing into the countries or regions the items are sold, and this is helping us to lower costs and mitigate some of the potential impacts of tariffs.”
— Ron Vachris, CEO
Five distinct mitigation vectors were articulated: (1) reroute high-tariff goods to non-US markets; (2) pull forward US summer/patio inventory ahead of duties; (3) increase US domestic sourcing (mattresses, pillows, plastic resin, outdoor furniture); (4) localize KS production by region (Asia laundry, China water); (5) discrete pricing decisions item-by-item (held pineapples and bananas; absorbed flowers price increases on more discretionary positioning). Millerchip’s tariff philosophy was explicit:
“At Costco, we remain committed to providing quality items at the lowest possible prices, and raising prices is always seen as a last resort.”
— Gary Millerchip, CFO
Read-through: tariff is no longer an unbracketed downside. The LIFO mechanism captures the inventory-cost inflation; the active mitigation reduces the absorption COST is forced to take on gross margin. The principal residual risk is whether non-foods inflation accelerates enough that the Q4 LIFO charge runs above the $40–$50M sized number, which Millerchip explicitly flagged as sensitive to tariff trajectory.
LIFO Mechanics — Why $130M This Quarter Is Three Quarters of FY25
Millerchip’s detailed LIFO walk in response to Bernstein’s Jiang Ma was the most analytically valuable exchange of the call. The key facts:
- LIFO is calculated on US inventory only (~$12–$13B); international uses retail method.
- Annual estimate based on net landed cost at start vs. end of fiscal year.
- FY25 estimate at current inflation: ~$145M total LIFO charge; $130M (9/13 weeks) recognized in Q3, ~$40–$50M (4/13) sized for Q4 at constant inflation.
- Implies overall blended inflation of ~1–1.5% on US inventory — below post-COVID peak but a directional change from H1 FY25 deflation.
- FY26 LIFO depends on FY26 inflation trajectory; no carryover from FY25 cumulative.
The cleanest disclosure-quality data point of the call — LIFO mechanism is now sized for the year and the residual variability is bounded.
Core-on-Core Margin Re-expansion — What Reverses, What Doesn’t
Core-on-core margin reversed from -8bps at Q2 to +36bps at Q3. Drivers:
- Fresh productivity + spoilage: structurally repeatable on continued sales leverage; we expect this to persist.
- Commodity deflation tailwinds: dairy, butter, eggs, olive oil. COST passed some back to members but absorbed margin relief faster than price-down cadence — this is the “feel margin pressure during inflation, relief during deflation” mechanic Millerchip articulated. Quarter-specific magnitude.
- Non-food margin: up slightly worldwide, down slightly US. Tariff drag emerging.
Asked by Oppenheimer’s Rupesh Parikh whether the drivers persist, Millerchip was deliberately conservative:
“Some of those were fairly unique to the quarter in the sense that the adjustment that we saw around some of the deflation in some of the ingredients ... we will continue to look for ways to invest in the member to drive top-line growth in our sales.”
— Gary Millerchip, CFO
Our model: core-on-core +36bps is the high mark for FY25; expect Q4 to come in flat-to-modestly-positive as tariff drag in non-foods offsets continued fresh productivity.
Affirm BNPL Launch — A Quiet Strategic Move
Buy-Now-Pay-Later via Affirm is now live across appliances, furniture, consumer electronics, with exclusive member rates. Strategic rationale:
“There were some Costco members that were already using Affirm as a solution for part of purchasing certain products at Costco while not coming through our ecosystem and not getting the full value from Costco with the exclusive pricing that we can offer.”
— Gary Millerchip, CFO
This is a quiet but thesis-relevant capital-light addition. Big-ticket discretionary on a finance-enabled basis brings a category COST has historically underpenetrated (vs. specialty retailers) into the membership ecosystem with native pricing. Early traction was characterized as “pleased with the initial sales results.” We treat this as a multi-quarter ramp that could compound the e-commerce big-and-bulky gains COST is already seeing through Costco Logistics.
Membership Fee Increase — ~4.6% Contribution, Building
The Sept 2024 US/Canada fee increase contributed ~4.6% of fee income in Q3, vs. ~3% in Q2. Trajectory remains on the path Millerchip laid out at Q2: largest deferred-accounting tailwind in Q4 FY25 and Q1 FY26. This is the biggest known forward catalyst on the COST EPS bridge into FY26 and is well-understood by the market.
Renewal Rate Slip — Cohort, Not Damage
The 30bps Q/Q slip at both US/Canada (93.0% → 92.7%) and worldwide (90.5% → 90.2%) is explained by the fall-2023 Groupon promotion cohort entering the renewal calculation for the first time, plus rising digital sign-up share at lower-than-average renewal. Millerchip’s framing — that digital members are net-positive (younger, growing the base) but renew at slightly lower rates — is the right narrative. The strategic imperative: convert digital sign-ups into in-warehouse engagement to migrate them up the loyalty curve.
Asked by Baird’s Peter Benedict how long this persists, Millerchip explicitly said “for the foreseeable future” due to the ongoing digital-promotion cadence and the renewal-lag from large Asian warehouse openings (4-5x average member counts at lower first-year renewal). We watch this but do not regard it as thesis-relevant absent a sub-92% US/Canada print.
Non-foods Decelerating — By Design
Non-foods were high-single-digits this quarter, decelerating from prior double-digit prints, exactly as Millerchip pre-flagged on tougher bullion and gift-card comparison. Underlying strength remained broad: gold/jewelry, majors, toys, housewares, home furnishings all double-digit. Vachris’s tariff-mitigation pull-forward on summer/patio kept availability strong in categories that would otherwise have been most exposed.
High-Volume Warehouse Throughput — A Strategic Priority
UBS’s Michael Lasser pressed Vachris on warehouse-volume saturation. Vachris confirmed the focus:
“About 80% of those warehouses we’re opening [in Q4] are going to cannibalize some high-volume locations for us. That’s going to take some relief off. ... Our recent expansion of gasoline hours was a great indicator that the throughput for our members improved nicely.”
— Ron Vachris, CEO
Strategic cannibalization plus front-end technology pilots (scan-and-go pilots in test, digital wallet integration) plus extended gas hours all targeting throughput. The framing — “turn over parking spaces quicker” — is a useful mental model for how COST grows the same warehouse without breaking the experience. Millerchip noted COST has roughly 40 warehouses with $400M+ annual sales; this group is the priority.
Outlook & Forward Commentary
Costco does not provide formal forward guidance. The actionable forward markers from the call:
| Forward Item | FY25 Color |
|---|---|
| FY25 capex | Just over $5B (vs. $1.13B in Q3) |
| FY25 new openings (revised) | 27 gross / 24 net new buildings; total to 914 worldwide |
| Q4 FY25 LIFO incremental charge | $40–$50M at current inflation rate |
| FY25 wage headwind (Q4) | Mid-single-digit bps Y/Y from March agreement; lap July 2024 increase 10 weeks into Q4 |
| Membership fee deferred tailwind | Largest impact Q4 FY25 and Q1 FY26 from Sept 2024 increase |
| Tariff exposure | ~1/3 of US sales imported; ~2/3 of imports in non-foods; China = ~8% of total US |
| Q4 FY25 openings | ~10 warehouses including 2nd Sweden, 20th Korea, 110th Canada |
The Q4 FY25 setup: deferred-fee tailwind landing at its largest single-quarter impact, partially offset by the incremental $40–$50M LIFO charge, the tail of wage investment (lapping July 2024 ten weeks in), and continuing FX translation drag. Net: we expect Q4 to comp at ~+5–6% reported / ~+7–8% adjusted with EPS in the high single-digit Y/Y range — in line with what consensus already discounts.
Analyst Q&A — Notable Exchanges
- Simeon Gutman (Morgan Stanley) opened on price-investment posture and competitive-gap dynamics. Vachris’s response — that buyers move first on commodity declines and competitive landscape “improved slightly at the latter part of the quarter” — was a constructive incremental read on COST’s relative price positioning.
- Christopher Horvers (JPMorgan) probed the price-gap improvement and tariff pull-forward magnitude. Vachris confirmed pull-forward was “slight ... very tough to quantify a certain percent” — a credible narrow framing rather than a thesis-relevant demand pull.
- Michael Lasser (UBS) drew the bullion/gift-card lap framing and the warehouse-throughput priority discussion. Vachris’s 80%-cannibalization framing on Q4 openings was the cleanest forward-read on warehouse-volume management.
- Scott Ciccarelli (Truist) noted EBIT margin expansion across eight or nine consecutive quarters and asked if the streak persists. Millerchip’s response was deliberately quarter-agnostic, framing margin around long-term reinvestment-into-value rather than streak management.
- Jiang Ma (Bernstein) drew the LIFO-mechanics walk that was the cleanest disclosure-quality data point of the call.
- Greg Melich (Evercore ISI) drew the digital penetration breakdown (~8% strict / ~10%+ broad / ~12% ex-gas) and the Costco Logistics share metrics (~20-25% of total deliveries / ~80-85% of big-and-bulky).
- Chuck Grom (Gordon Haskett) asked about front-end technology pilots and warehouse-hour extension. Vachris flagged scan-and-go testing and digital-wallet checkout integration as the principal levers; warehouse hours not yet expanded but on the watch list.
- John Heinbockel (Guggenheim) asked about depot/supply-chain efficiency and tariff inventory planning. Millerchip framed depots as cross-dock-throughput-optimized; Vachris emphasized 6-8 month buying commitments and treasure-hunt flexibility on cross-region rebalancing.
- Rupesh Parikh (Oppenheimer) drew Millerchip’s explicit framing that core-on-core drivers were “fairly unique to the quarter” — an important caveat against extrapolating the +36bps print.
- Kelly Bania (BMO) asked about pricing posture and Affirm strategy. Pineapples/bananas (held), flowers (passed through) were the discrete examples on tariff absorption discipline. Affirm framed as bringing existing off-platform Affirm-COST users into the native pricing ecosystem.
- Kate McShane (Goldman Sachs) asked about international sentiment toward COST as a US brand. Millerchip acknowledged some member feedback on US-relations tension but flagged that all international countries had positive comp growth in the quarter — a clean operational counter-narrative.
- Peter Benedict (Baird) drew the renewal-rate cohort-explanation framing and the “foreseeable future” characterization of when digital sign-up dilution stabilizes — not a near-term reversion.
What They’re NOT Saying
- No sizing of the Affirm BNPL contribution. “Pleased with initial sales results” is the entire disclosure. Watch Q4 for any commentary on attach rate or AOV uplift.
- No commitment on continued core-on-core expansion. Millerchip explicitly framed the Q3 drivers as “fairly unique.” The market may be over-extrapolating.
- No formal Q4 guidance on the deferred-fee impact. The directional framing (largest impact Q4 FY25 / Q1 FY26) is consistent with prior quarters but unsized.
- No update on US membership fee subsequent increase. The Sept 2024 increase is still rolling through; international increases (Japan, Korea announced) are out of phase. No US color.
- No special-dividend signaling. The January 2024 $6.7B special is rolled off the cash-balance comp; capital return cadence remains an unaddressed framework question.
- No detailed scan-and-go rollout timeline. Pilots in test “extremely successful” with no rollout cadence. Material if it scales given throughput implications.
- No quantified renewal-rate floor. “Foreseeable future” for the digital cohort drag is the only forward color. We watch for a sub-92% US/Canada print as a downgrade trigger.
Market Reaction
COST closed Friday May 30, 2025 up approximately 4.1%, a clean post-print rally. Volume was elevated. Reaction reflected the buy-side reading the same two signals we read: (1) core-on-core margin reversed cleanly, removing the Q2 thesis-fissure question; (2) tariff impact landed mostly inside a known LIFO mechanism rather than as an open-ended cost-of-goods absorption.
The rally is consistent with a stock that responds favorably to thesis-validating prints. We respect the read; our reservation is that the multiple now discounts further validation in Q4, and the deferred-fee tailwind is the single biggest known catalyst still ahead. Buy-the-confirmation is a path-of-least-resistance trade we choose not to chase.
Street Perspective
The bull case being made on the Street post-print converges on three planks: (1) core-on-core margin reversal demonstrates the long-run beat-and-raise template is intact, with fresh productivity and spoilage improvements as repeatable structural drivers; (2) tariff exposure is now sized inside the LIFO mechanism, with active mitigation across five vectors compressing the residual EPS risk into a known bracket; (3) the membership-fee deferred tailwind is the largest known forward catalyst, landing at its peak quarterly impact in Q4 FY25.
The bear case being articulated centers on: (1) renewal rate has now slipped 30bps Q/Q at both US/Canada and worldwide; the digital-cohort explanation is plausible but the absolute number is the lowest US/Canada print in several quarters; (2) e-commerce comps decelerated meaningfully (+20.9% → +14.8%) and the bullion/gift-card lap continues into Q4; (3) Q3 core-on-core drivers were explicitly characterized as “fairly unique,” suggesting the +36bps print is the high mark not the new run-rate; (4) at ~50x forward earnings on consensus, the multiple already prices in the deferred-fee tailwind.
Our read sits on the bull side of (1) and (2) but agrees with the bear framing on (3) and (4). On (1) renewal, we treat the slip as cohort-driven noise unless Q4 prints below 92% US/Canada.
Model Implications
- Q4 FY25 revenue: +6–7% reported, ~+8% adjusted on continued traffic strength offset by FX drag and a modest non-foods deceleration.
- Q4 FY25 gross margin: Modest Y/Y expansion as deferred-fee tailwind contributes incremental margin offset by $40–$50M LIFO charge. Core-on-core back to flat-to-modestly-positive (vs. +36bps Q3 high mark).
- Q4 FY25 SG&A: Net wage headwind moderating as July 2024 lap arrives 10 weeks in; productivity continues to offset.
- FY25 EPS: Range moves up to $17.40–$17.90 on Q3 beat plus stronger Q4 deferred-fee landing.
- FY26 EPS: Q1 FY26 captures the second-largest deferred-fee impact; we model FY26 EPS in the $20.50–$21.50 range, with tariff trajectory as the principal swing.
- Capex: $5.0–$5.2B FY25 base case held.
- Tariff scenario: base case is FY25 LIFO charge ~$145M; bear-case sensitivity is $200M+ if non-foods inflation accelerates beyond current trajectory in Q4.
Thesis Scorecard
| Thesis Point | Status | Notes |
|---|---|---|
| Bull #1: Membership annuity compounds; Sept 2024 fee deferred tailwind builds | Confirmed + | Fee contribution stepped up 3% → 4.6% of fee income; largest impact Q4 FY25 / Q1 FY26 still ahead |
| Bull #2: International unit economics underwrite multi-year warehouse runway | Confirmed | All international countries positive comp growth; Sweden 2nd, Korea 20th, Canada 110th in Q4 |
| Bull #3: E-commerce / Logistics / Next compound past inflection | Confirmed | E-com +14.8% (decelerating on tough laps); Costco Logistics items +31%; Next = full FY22 sales in Q3 alone |
| Bull #4: Kirkland Signature flywheel + localized production | Confirmed + | +50bps KS sales penetration Y/Y; Asia laundry localization -40% prices; 40+ new KS items in Q3 |
| Bull #5 (NEW): Affirm BNPL on big-ticket extends member ecosystem | New — Early | Captures off-platform Affirm-COST users; early traction "pleased"; ramp watch in Q4 |
| Bear #1 (Q2): Core-on-core margin compounding pause | Resolved | Reversed -8bps → +36bps; primary Q2 fissure closed |
| Bear #2 (Q2): Tariff exposure unbracketed | Bracketed — Active | $130M Q3 LIFO + $40-50M Q4 sized; mitigation across five vectors; residual is non-foods inflation acceleration |
| Bear #3: Wage cadence stepped up through 2027 | Active | +20bps SG&A reported; productivity offsetting partially; July 2024 lap helps Q4 |
| Bear #4: Multiple discounts continued execution | Active | Stock +4% on print; ~50x forward leaves limited margin of safety |
| Bear #5 (NEW): Renewal rate slip on digital cohort | Active — Latent | -30bps Q/Q both US/Canada and worldwide; cohort-driven; watch for <92% US/Canada |
Action: Maintaining Hold. Operationally the print we wanted to see — the Q2 concerns reversed cleanly. But the multiple has rallied on the validation, and the largest known forward catalyst (Q4 FY25 / Q1 FY26 deferred-fee landing) is already priced. Path to Outperform requires either (a) Q4 FY25 prints with deferred-fee impact landing at the upper end of management’s framing, validating FY26 EPS upside that the market under-models, or (b) a tariff-driven multiple reset that creates margin of safety. Path to Underperform requires sustained renewal-rate slip below 92% US/Canada or a tariff scenario where Q4 LIFO runs materially above the $40–$50M sized number. Hold remains the honest call.