Costco Wholesale Corporation (NASDAQ: COST) — A Flawless Business, Priced for Flawlessness

Report date: 2026-06-10 · Sector: Consumer Staples — Consumer Staples Distribution & Retail (Warehouse Clubs) Fiscal year: ends end-August · CIK: 0000909832 · Price at analysis: ~$969 · Market cap: ~$430B · Enterprise value: ~$419B

This is an independent fundamental-research article. With the single, explicitly-labeled exception of the “Author’s Take” block below, the body carries no buy/sell recommendation and no price target — valuation is discussed only as embedded expectations and scenarios.

⚡ Author’s Take

This block is the author’s own independent opinion and general information only — not investment advice. Everything below it is the analytical body and remains strictly position-free and price-target-free.

Verdict: HOLD / great business, full price — accumulate only on weakness, not a short. Conviction: medium-high. Constructive accumulation zone ≈ $700–800 (~34–39× FY26E EPS of ~$20.6); the stock at ~$969 (~47× trailing, ~43× forward, ~30× EV/EBITDA, ~80th percentile of its own 10-year valuation) is fair-but-full to expensive, not a bargain. Tag: “Flawless business, priced for flawlessness.”

Costco is, on the evidence, one of the highest-quality businesses in all of retail — a subscription company wearing a retailer’s clothes, where membership fees supply ~half of operating income and two-thirds of net income, renewals run at 92%+, the operating business earns ~40% return on the capital actually deployed, and growth is self-funded by a negative-working-capital model. I would happily own it forever — at the right price. The problem is entirely the price. At ~49× earnings the market is paying a fair, full price for ~10–12% EPS compounding sustained for a decade and assuming the record multiple essentially never compresses. That second assumption is the whole risk: the business can execute flawlessly and the stock can still go nowhere — a de-rate from 49× to even 30× (still a premium to every direct club peer) is roughly −40% with zero fundamental deterioration. The framing is quality-compounder-at-the-wrong-price, not value, not momentum, not falling knife.

What makes this a disciplined HOLD rather than an AVOID is that there is no catalyst forcing a de-rate and the franchise keeps widening — so shorting it is a good way to be right slowly and lose money. But I refuse to underwrite a ~2.6% all-in starting yield (earnings + dividend) on a 3.8%-operating-margin grocer whose member growth just decelerated to +4.1% (“lowest in some time”) and whose headline comps are gas- and FX-flattered. The single tell that I am reading correctly: management itself prefers special dividends to buybacks “at our current valuation” — insiders are telling you the stock is too expensive to repurchase. I would become a buyer in the high-$600s to high-$700s. Flips bullish: an international/member-growth re-acceleration (China/Asia inflection) that proves the perpetuity-growth case while the multiple holds — I’d pay up. Flips bearish (to outright avoid/trim): two-plus quarters of ex-gas/FX comp trending below ~5% and member growth stuck ≤4% with no new-market offset — at which point 49× becomes indefensible and the de-rate arithmetic takes over.

1. Executive Summary

Costco is a subscription business with a giant, near-breakeven logistics operation attached. That single sentence is the most important thing to understand about the company. It operates 928 membership warehouse clubs across 14 countries, sells a deliberately narrow ~3,800-SKU assortment at the lowest gross margins in large-cap retail (~11%, almost entirely consumed by SG&A), and earns roughly half of its operating income — and effectively the majority of pre-tax profit — from the recurring annual membership fee. Reconciling the FY2025 10-K: membership fee income (MFI) of $5,323M covered 51.3% of operating income ($10,383M) and 65.7% of net income ($8,099M). The merchandise operation exists to make the membership worth renewing; the membership, renewing at 92.2% in the US/Canada and 89.7% worldwide, is the product.

The business quality is not in serious dispute. Costco possesses the strongest configuration in Bruce Greenwald’s framework — economies of scale + a cost advantage + customer captivity — and it is provable in the numbers, not asserted in narrative: ~92% renewal (captivity), ~40% operating return on invested capital once the ~$15–20B cash hoard is excluded (cost/scale advantage), 13× inventory turns and a negative cash-conversion cycle in which suppliers finance the entire inventory position (a quantified structural edge), and a Kirkland Signature private label that is both a margin lever and a switching cost. Revenue compounded at 8.9% and EPS at 12.7% over FY21→FY25; FY25 delivered $275.2B revenue (+8.2%), $8.10B net income (+9.9%), and $18.21 diluted EPS. The warehouse-club channel is the rare good neighborhood in a structurally bad retail industry: a rational three-player oligopoly (Costco, Sam’s Club, BJ’s) with a favorable capital cycle in which high incumbent returns are not attracting the mean-reverting capital that would compete them away.

The entire debate is about price, not quality. At ~$969, COST trades at ~48.8× trailing / ~43× forward earnings, ~30× EV/EBITDA, and the ~80th percentile of its own ten-year valuation distribution — the single most expensive large-cap name in staples retail, a ~2.4× P/E premium to its closest analog BJ’s (20.6×), ~16% above Walmart, and roughly 3× the multiple of Target or Dollar General. A reverse-DCF says the market is underwriting ~10–12% EPS compounding for 10+ years and near-zero multiple compression. That is reasonable on growth but heroic on multiple persistence. Even the bull case requires the market to keep paying ~45–50×; the bear case requires no operational failure at all — merely a re-rating toward where club peers trade. The asymmetry (large de-rating downside vs. capped upside off a record base) is the defining feature of the setup.

The honest blemishes are growth-rate, not existence-of-growth: paid-member growth has decelerated to +4.1% (described on the Q3 FY26 call as “the lowest level in some time”); worldwide traffic has slowed toward ~2%, so the steady 6–7% core comp leans increasingly on bigger baskets; and the headline +11.6% Q3 net-sales growth was flattered by a Middle-East-driven gas-volume surge (~2.2pts) and FX (~1pt). Two C-suite seats (CEO and CFO) turned over in 2024. None of this breaks the franchise; all of it raises the bar for a 49× multiple. This memo takes no position. It lays out why the business is exceptional, why the price is demanding, and exactly what an owner is underwriting.

2. Business Overview

What Costco does. Costco operates membership warehouse clubs — 928 worldwide as of Q3 FY26 (up from 914 at FY25 year-end, 890 at FY24, 861 at FY23) — selling a narrow assortment of merchandise in bulk at razor-thin margins, and earning the bulk of its profit from a recurring annual membership fee. It operates in 14 countries: the United States, Canada, Mexico, Japan, the United Kingdom, Korea, Australia, Taiwan, China, Spain, France, Sweden, Iceland, and New Zealand.

The economic engine — membership fees ≈ operating income (the structural fact). This is the crux of the entire investment case. Reconciling the FY25 10-K income statement:

[FACT] Membership fees cover roughly half of operating income (51–56%) and two-thirds of net income (66–73%). [INTERPRETATION] Costco runs the merchandise operation close to breakeven on a fully-loaded basis — ~11% merchandise gross profit is almost entirely consumed by the ~9% SG&A rate — and the fee converts a near-zero-margin retailer into an ~$8B/year profit machine. Encouragingly, the “merchandise operating income ex-MFI” line has risen $3.5B→$5.1B over three years, so the merchandise/ancillary engine is now modestly profitable in its own right and contributing a rising share of incremental profit (which is why the MFI/operating-income ratio has drifted down — a healthy sign, not a deterioration). MFI grew 10% in FY25 and +10.7% in Q3 FY26 ($1.373B); stripping the Sept-2024 fee increase and FX, the underlying organic MFI growth was 7% — the real annuity-growth rate.

Revenue segmentation. FY25 geographic segments (10-K), revenue / operating income: United States $200,046M / $6,878M (66.2% of OI); Canada $36,923M / $1,849M (17.8% of OI); Other International $38,266M / $1,656M (15.9% of OI). [INTERPRETATION] Canada and Other International together are ~27% of revenue but earn higher segment operating margins (~5.0% Canada, ~4.3% Other Intl) than the US (~3.4%) — this is both the FX exposure bears cite and the higher-margin growth frontier. Merchandise categories: Foods & Sundries (dry grocery, candy, cooler/freezer, liquor, tobacco), Non-Foods (appliances, electronics, health & beauty, hardware, sporting goods, tires, toys, apparel, furniture, jewelry), Fresh Foods (meat, produce, deli, bakery), and Warehouse Ancillary (gasoline, pharmacy, optical, food court, hearing aids, tire installation, business centers, travel, e-commerce). Gasoline was ~10% of net sales in FY25 (747 stations); e-commerce ~7%.

Membership tiers and the Executive flywheel. At FY25 year-end: 81.0M total paid members (FY24 76.2M; FY23 71.0M), 145.2M total cardholders, of which 38.7M Executive members. By Q3 FY26: 82.9M paid (+4.1%), 148.5M cardholders, 41.2M Executive (+9.6%). [FACT] Executive members are ~48% of paid members but drive ~73.6% of worldwide net sales (FY25 10-K). Executive pays an extra $65/yr (US) and earns a 2% reward on qualified purchases up to $1,250/yr. [INTERPRETATION] The 2% reward is a self-funding loyalty loop — it nudges members to consolidate spend at Costco to “earn back” their fee, raising share-of-wallet and renewal probability, while the incremental fee more than offsets the reward for heavy spenders.

Kirkland Signature. The private label is described in the 10-K as high-quality at prices generally below national brands and at generally higher margins. On the Q3 FY26 call, management cited 15–20% savings versus national-brand equivalents at “equal or better quality.” [ASSUMPTION] KS is widely estimated at roughly one-third of sales (~$86B+) — an analyst estimate; Costco does not disclose an exact KS dollar figure or penetration. [INTERPRETATION] KS is a genuine margin-and-loyalty flywheel: a higher-margin captive brand available only at Costco, deepening switching friction while improving merchandise economics.

Recurring vs. non-recurring. MFI (~2% of revenue but ~half of operating income) is the recurring, high-visibility annuity renewing at 92.2% US/Canada. Merchandise sales are technically transactional but de facto recurring given ~90% renewal and high frequency (worldwide traffic +2.4% in Q3 FY26). Few retailers have a revenue line this predictable.

Verdict: Costco is, in economic substance, a subscription business wearing a retailer’s clothes — it sells goods near breakeven to acquire and retain a paying-member base whose annual fee, renewing at ~90%+, supplies roughly half of operating income and two-thirds of net income. The model is unusually transparent and unusually durable; the merchandise loss-leader strategy is structurally sound because the fee, not the markup, is the profit. This is one of the highest-quality business models in retail.

3. Industry Dynamics

Industry definition and structure. Costco competes in (a) the broad US/global retail and grocery industry and (b) the narrower warehouse-club sub-segment, which in the US is effectively a three-player oligopoly — Costco, Sam’s Club (Walmart), and BJ’s Wholesale. Broader competitors include Walmart, Amazon, Target, Kroger, Aldi, and regional grocers. [FACT/INTERPRETATION] General US retail is a structurally hostile, low-margin, hyper-competitive industry — Walmart’s consolidated FY2026 operating margin was ~4.2% (Walmart’s public disclosures and peer analysis). The warehouse-club niche is a structurally better neighborhood within a bad industry: a limited-SKU bulk format, membership gating, and high inventory turns produce a cost structure general retailers cannot match. The club channel’s barrier is the membership wall — it pre-selects committed, higher-income, higher-frequency customers and converts an upfront fee into the profit pool, insulating the operator from the markup war that defines grocery and mass retail.

Competitive intensity — the three-way club race. Sam’s Club US is ~13% of Walmart’s sales and a direct membership-warehouse competitor. [INTERPRETATION] Public peer analysis of Walmart independently judged Costco’s membership-renewal lock-in (~90%+) “arguably a stronger captivity moat” than Walmart’s — “part of why Costco commands an even higher multiple.” Sam’s and BJ’s compete hard on price but neither matches Costco’s per-warehouse volume (AUV, the widest in the channel) or its renewal rate. [OPEN QUESTION] Exact current Sam’s/BJ’s renewal and AUV figures are not in our corpus; BJ’s renewal is commonly cited ~90% and Sam’s somewhat lower — treat as external/approximate.

The structural shift to value. [INTERPRETATION] The macro backdrop (persistent inflation, value-seeking consumers) is a channel tailwind. Q3 FY26 commentary repeatedly emphasized members “very willing to spend but with very high expectations around quality, value and newness” — exactly the demand Costco is built to capture. The trade-down dynamic that pressures mid-market retailers feeds the clubs, evidenced by record price-sensitive gas volumes in the quarter.

Regulation and exogenous factors.

• Tariffs / IEEPA [FACT]: Costco began submitting IEEPA tariff refund claims via US Customs and Border Protection in Q3 FY26, expecting refunds “on a rolling basis,” and plans to “return to our members, in some form, the portion of tariffs that were passed on to them” — subject to “developments in the lawsuit filed against the company regarding the return process.” [INTERPRETATION] A manageable, partly-recoverable cost shock, not a thesis-breaker; Costco’s narrow SKU count and buyer expertise let it re-source faster than broadline peers.
• Gas [FACT]: ~10% of net sales; a traffic magnet (gas-buyers spend more and renew higher) but a low-margin, price-volatile category that distorts reported comps and gross margin quarter to quarter.
• FX [FACT]: ~27% of revenue is non-US; FX added ~1% to Q3 FY26 sales. A translation effect, not an operating risk.
• Labor [FACT]: A March-2025 employee agreement (added vacation days, wage increases) governs a largely non-union US workforce paid above market. [INTERPRETATION] High wages are a deliberate retention/productivity lever, not a margin problem — SG&A leverage held.
• Pharmacy/optical/hearing [FACT/INTERPRETATION]: regulated ancillary services; pharmacy posted mid-20s% comps in Q3 (GLP-1 demand, Medicare-D flex cards) with “significant market share gains.”

Marathon capital-cycle read. [INTERPRETATION] Applying Marathon’s supply-side lens: capital is not flooding into the US warehouse-club channel. New supply is rational and slow — Costco targets only ~30 net new units/yr against a 928-unit base (~3%/yr), Sam’s and BJ’s add modestly, and there are no credible new entrants because the model requires enormous scale to function (buying power, depot networks, ~$6.5B FY26 CapEx). This is the favorable Marathon setup: high incumbent returns that are not attracting mean-reverting capital because the barriers are real, and the asset-growth-anomaly risk is muted because Costco’s unit growth is demand-validated, not capacity-chasing.

Verdict: General retail is a structurally bad industry, but the warehouse-club niche is structurally good and Costco’s position within it is structurally excellent. The membership wall converts a markup war into a subscription business; the three-player oligopoly is rational with disciplined capital deployment and no credible new entrants; tariffs, gas volatility, and FX are real but cyclical-not-structural. The capital cycle is favorable: high returns are not attracting the capital that would normally compete them away. This is the rare retail sub-industry worth owning.

4. Competitive Position

Greenwald verdict up front: [INTERPRETATION] Costco possesses economies of scale + a cost advantage + customer captivity — the strongest combination in Greenwald’s framework and the only configuration that is durable when all three reinforce one another. It is not primarily a network-effect business (see the pressure-test).

Cost-advantage mechanism (supply side). [FACT] Costco carries roughly 3,800 SKUs versus 30,000+ at a typical supermarket. [INTERPRETATION] The mechanism, quantified through its consequences: (1) buying power per SKU — concentrating volume into ~3,800 items gives maximal leverage with each supplier, extracting cost that funds the price gap; (2) inventory turns — limited assortment plus bulk pallets drive ~13× turnover, so Costco often sells goods before it pays the supplier, financing merchandise with negative working capital; (3) low shrink and labor — pallet merchandising and few SKUs reduce handling; (4) treasure-hunt — rotating, limited-quantity non-foods drive urgency and frequency. These compound: lower costs → lower prices → more member value → higher renewal and volume → more buying power → still-lower costs. The ~11% merchandise gross margin (vs. 25%+ at supermarkets) is a deliberate cap — Costco self-imposes a markup ceiling because the fee is the profit, which structurally prevents a competitor from undercutting it without bleeding.

Customer-captivity proof (demand side). [FACT] The captivity is measured, not asserted: US/Canada renewal 92.2% and worldwide 89.7% (Q3 FY26), stable for years; Executive members (~48% of base) drive 73.6% of sales. [INTERPRETATION] A ~90%+ renewal rate is the cleanest possible proof of captivity — members re-pay annually for the right to shop, and 9 of 10 do so every year. The fee creates a sunk-cost/consolidation psychology, and the 2% Executive reward formalizes it.

Scale economies and brand. [FACT] 928 warehouses, ~$275B revenue, $6.5B FY26 CapEx, a proprietary depot network, and in-house manufacturing (hot-dog, coffee-roasting, KS production). Fixed costs spread over the largest club volume in the world — the classic Greenwald scale-plus-captivity barrier no entrant can replicate without first achieving the scale. Kirkland adds a layered intangible: a captive brand at 15–20% below national brands that a trusting member can only re-buy at Costco.

Head-to-head and the Amazon question. [INTERPRETATION] Versus Sam’s Club and BJ’s, Costco wins on per-club volume, renewal rate, and the Executive/KS flywheel. Why Amazon hasn’t killed it: (1) Costco’s gross margin is ~11% — there is almost no markup for Amazon to undercut; you cannot win a price war against a competitor that isn’t trying to make money on the goods. (2) Bulk fresh food, the treasure-hunt impulse, gas, food court, and in-person discovery are poorly suited to e-commerce. (3) The membership fee is the moat Amazon Prime copied, not one it can erode. [FACT] Costco is itself building digital fast (Q3 FY26 digitally-enabled comps +21.5%; site/app traffic +37%; same-day delivery; AI-search the fastest-growing, highest-converting traffic source), narrowing the only axis where Amazon led.

Pressure-test — where the moat is weakest. [INTERPRETATION] Three honest weak edges: (1) The “network effect” is overstated — Costco’s lock-in is switching friction + sunk-cost + habit, not a true network effect; a member’s value does not rise because other members join. Calling it a network effect flatters the moat; it is captivity, full stop — powerful but mechanically weaker than a self-reinforcing network. (2) Geographic, not contractual, lock-in — captivity partly depends on warehouse proximity and the absence of an equally-cheap nearby alternative; where Sam’s/BJ’s are competitive and convenient, the finite fee differential is the main switching cost. (3) The renewal rate is mix-sensitive — online sign-ups renew at a lower rate than warehouse sign-ups, a structural drag on the headline as digital grows, currently offset by targeted retention. None of these breaks the moat; they bound it.

Verdict: Durable competitive advantage — scale + cost advantage + customer captivity (Greenwald’s strongest configuration), with a layered Kirkland intangible. The moat is provable in the numbers (92.2% renewal, 73.6% of sales from Executive members, ~11% self-capped gross margin, ~13× turns) and survives the Amazon threat precisely because there is no margin to attack. The one intellectual-honesty caveat: this is customer captivity, not a network effect — calling it the latter overstates it. Among retailers, this is a top-tier, genuinely durable moat that would visibly deteriorate (renewal, turns, gross-margin cap) without the advantage — passing the “tie the moat to a number” test.

5. Growth History and Forward Opportunities

Historical comparable-sales growth (10-K MD&A), reported / ex-gas-and-FX:

[FACT] FY25 net sales rose 8% to $269,912M; of the $20,287M increase, $14,788M (6%) was comp-driven and the rest new warehouses. [INTERPRETATION] The ex-gas-and-FX “core” comp of 5%→6%→8% shows genuine acceleration into FY25 — high-quality, mostly volume/frequency-led growth.

Recent quarterly trajectory (FY26). [FACT] Q1 FY26: net sales $65.98B (+8.2%); comp +6.4%; MFI $1.329B (+14%); EPS $4.50. Q3 FY26: net sales $69.15B (+11.6%); comp +9.8% (+6.6% ex-gas/FX); traffic +2.4%; ticket +7.3% (+4.2% ex-gas/FX); digitally-enabled comps +21.5%; net income $2.192B / $4.93 EPS (+15%). [INTERPRETATION] Core (ex-gas, ex-FX) comp has run a remarkably steady 6–7% for over a year — management explicitly said they “haven’t really seen any variation from that performance.” Steady, not accelerating, but high-quality.

Traffic vs. ticket — a quality flag. [FACT] Q3 FY26 worldwide traffic +2.4%, but management said recent monthly traffic is running ~2% (vs. ~5% a year ago), with basket size offsetting. [INTERPRETATION] Traffic deceleration is the quality watch-item in the comp. A 6–7% core comp increasingly leans on bigger baskets rather than more visits. Management attributes it to lapping prior-year strength, tariff-driven stock-up distortions, and physical capacity constraints at high-volume clubs. The mix shift toward ticket is not inherently bad (more units + trade-up) but raises exposure to a consumer pullback.

Unit growth. [FACT] 861 (FY23) → 890 (FY24) → 914 (FY25) → 928 (Q3 FY26). Costco now expects 26 net new openings in FY26 (cut from 28; 2 slipped to FY27) and targets 30+ net new openings/yr “in the coming years.” FY26 CapEx ~$6.5B funds new units, relocations of capacity-constrained clubs, depot expansion, KS manufacturing, and digital. [INTERPRETATION] ~3%/yr unit growth on a 928 base is disciplined and demand-validated; relocations to larger sites unlock trapped demand at maxed-out clubs.

International runway and unit economics. [FACT] Management has Canada “charted out” for 3–5 years; Asia (China, Korea, Japan) holds “the greatest potential”; Europe (France “very young,” Spain “significant growth,” UK strong). Costco launched Executive Membership in China in Q3 FY26 with adoption “ahead of expectations.” [INTERPRETATION] International is the multi-decade runway and is accretive — Canada/Other-Intl already earn higher segment margins than the US. [OPEN QUESTION] New-market clubs sign up many members who renew at a lower rate initially, so the membership-growth benefit of new markets is partly offset by lower stickiness until the base matures.

E-commerce / digital and ancillary. [FACT] Digital is the fastest-growing channel (digitally-enabled comps +20.8–21.5%; site/app traffic +37%; personalized carousels delivering ~3× conversion and ~$0.5B of e-com sales in the quarter; new Google Commerce Media/YouTube retail-media partnership). Ancillary comps +mid-20s% (pharmacy-led); gas comps +high-20s%. [INTERPRETATION] Digital and retail media are nascent, capital-light, higher-margin optionality layered on the core, and a source of SG&A leverage (faster checkout → lower payroll).

The bear flag — membership-growth deceleration. [FACT] Paid-member growth decelerated through FY26: +5.2% (Q1 FY26) → +4.1% (Q3 FY26), called “the lowest level in some time.” [FACT] Management’s defense: ex-fee-increase and FX, total membership revenue grew 7%, of which Executive +9.6% (heavier spenders); the slowdown reflects no recent major new-market openings, cycling strong prior-year sign-ups, and renewal “normalization,” with “4% to 5%” framed as a “more normal rate” absent a special event. [INTERPRETATION] This is the most important growth debate. New members are the future comp pipeline — they ramp warehouse usage over years — so 4.1% growth (down from 5.2% two quarters earlier) is a genuine yellow flag for out-year comp durability. The bull rebuttal is legitimate (Executive +9.6% and renewal stabilization mean revenue-per-member and stickiness are improving even as headcount growth slows), but the deceleration is real and the domestic-saturation ceiling is finite. New-market openings are the lever to re-accelerate, and their cadence (especially Asia) is the swing factor.

Verdict: High-quality growth, decelerating at the margin but still well above the retail field. Core (ex-gas/FX) comp has held a steady, volume-and-ticket-led 6–7% for over a year; unit growth is disciplined (~3%/yr, targeting 30+/yr) and accretive internationally; digital/ancillary/retail-media add capital-light optionality. The two honest blemishes — paid-member growth slowed to 4.1% and traffic decelerated to ~2% — are bounded, not broken, and partly offset by Executive-tier strength. The growth is real and high-quality; the open question is its rate of deceleration, not its existence.

6. Financial Quality

Multi-year build (FY2021–FY2025, $M):

[FACT] Revenue compounded at 8.9% FY21→FY25; net income at 12.8% and diluted EPS at 12.7% — earnings grew ~1.4× faster than revenue, evidence of modest but real operating leverage (op margin +35bp). [FACT] Q3 FY26 (12 weeks ended ~May 10, 2026, 10-Q filed 2026-06-03): net sales $69,154M (+11.6%), MFI $1,373M (+10.7%), operating income $2,815M (+11.3%), net income $2,192M, EPS $4.93 (+15%), tax rate 25.4%. Nine-month FY26 revenue $207,431M (+9.7%), operating income $7,884M (+12.0%).

The structural crux — MFI as the real profit engine. [INTERPRETATION] Roughly half of operating income — and, given near-breakeven merchandise after SG&A, effectively the majority of pre-tax profit — comes from membership fees, not from selling goods. MFI is ~1.9% of revenue but ~51–58% of operating income. The MFI/operating-income ratio drifted down from 57.8% (FY21) to 51.3% (FY25) — positive, showing the merchandise/ancillary engine contributing a rising share of incremental profit. The merchandise operation’s job is to drive the renewal rate (92.2%), not to earn a retail margin.

Gross margin, SG&A, operating margin. [FACT] Merchandise gross margin is structurally ~10.5–11.1% by design, the lowest in large-cap retail; reported margin is heavily distorted by gasoline price swings and LIFO (a LIFO charge ran through FY25; Q3 FY26 carried a smaller $44M LIFO charge vs. $130M prior-year). SG&A crept up off the FY22 low as Costco invests in wages, technology, and new-market build-out. [INTERPRETATION] The operating-leverage story is real but thin and slow — +35bp over four years on a 3.4% base. There is no margin-expansion bonanza: Costco deliberately reinvests scale economies into lower member prices and higher wages — the Greenwald scale-plus-captivity advantage being spent on the moat rather than dropped to the P&L. The risk is that the leverage actually shows up in renewal and member growth (compounding MFI at ~100% incremental margin), so if member growth stalls (+4.1% and slowing), the engine that carries the leverage slows with it.

FCF, ROIC, ROE. [FACT] FCF: $5.4B / $3.5B / $6.7B / $6.6B / $7.8B (FY21–25); FCF conversion ~95–100% in normal years — cash earnings are high quality (net income does not materially diverge from CFO). [FACT/INTERPRETATION] Properly excluding the ~$15B cash + ST-investment hoard from invested capital, operating ROIC runs 36–40% every year (FY25 ~40%); including the cash drags reported return-on-capital to ~19–24%, which is why unadjusted screens understate the operating business. ROE is 25–31% (FY25 27.8%), understated by the cash drag and achieved with near-zero financial leverage. This is an elite-return business on every Greenwald test.

The negative-working-capital machine. [FACT] Costco sells inventory before it pays suppliers: Days Inventory ~28, Days Payable ~30, near-zero receivables → cash-conversion cycle ≈ −3 days (FY25), roughly −4 to 0 across FY21–25. Accounts payable ($19,783M FY25) exceeds inventory ($18,116M FY25) by $1.67B — suppliers finance the entire inventory position and then some. Inventory turns ~13× (vs. Walmart ~9×, Target ~6×). [INTERPRETATION] Negative working capital means growth is self-funding — every new warehouse and dollar of comp growth releases cash. This is a genuine, quantified structural advantage, not a narrative.

Balance sheet, leases, SBC. [FACT] Net cash $9.5B at FY25 year-end, rebuilt to a ~$20B gross cash position (cash $18,946M + ST investments) by Q3 FY26 (post-special-dividend rebuild well advanced). LT debt $5.7B, minimal current maturities. Operating-lease liabilities are modest (~$2.67B total) — Costco owns ~80%+ of its real estate, [OPEN QUESTION] a hidden asset plausibly tens of billions above book (a Greenwald asset-value cushion). SBC: $665M (FY21) → $860M (FY25), ~0.3% of revenue and ~10.6% of net income — low for its size, overwhelmingly RSUs. Buybacks only modestly exceed SBC, so dilution is roughly neutralized but not aggressively retired.

One-time items to normalize. [FACT] The $15/share special dividend (~$6.7B) paid Jan 2024 drove FY24 dividends-paid to $9,041M (non-recurring; FY25 normalized to $2,183M). Gasoline price swings and LIFO move reported revenue/margin each period — always read management’s “ex-gas, ex-FX” comp.

Verdict: Do economics improve with scale? Yes — but the improvement is deliberately suppressed and re-routed into the moat. The economics are elite on the metrics that matter for this model (40% operating ROIC, −3-day cash cycle, 13× turns, ~95% FCF conversion, half of profit from a high-renewal subscription) and they do improve with scale, but operating margin barely moves because Costco returns scale economies to members (lower prices) and employees (higher wages). Disconfirming evidence weighed: a skeptic can correctly say “net margin is 2.9% and op margin barely expands — where’s the leverage?” The answer is that the leverage shows up in renewal and member growth that compound MFI at near-100% incremental margin — which is precisely why the +4.1% member-growth deceleration matters. High-quality economics, correctly understood as a subscription business, but priced for the leverage to keep compounding.

7. Capital Allocation

The framework. [FACT] Costco runs a deliberate, three-pronged return policy: (1) a steadily-growing regular dividend, (2) periodic large special dividends funded by accumulated cash, and (3) a token buyback that roughly offsets SBC dilution. CapEx into new warehouses is the primary use of cash; M&A is essentially absent. The regular dividend has compounded at a low-double-digit CAGR (raised to $1.30/qtr in April 2025 from $1.16); the regular payout ratio is low (~28% of FY25 EPS), which is why the cash pile rebuilds quickly.

Special-dividend history:

[FACT] The 2024 special ($15.00) is confirmed directly from the 8-K corpus; the pattern — a ~$5–$15 special every 2–4 years, escalating with the cash balance — is well established. [INTERPRETATION] The cadence is the tell on Costco’s capital philosophy: management refuses to lever the balance sheet permanently or to chase buybacks at a high multiple, so it lets cash accumulate to ~$15–20B and disgorges a lump. This is tax-aware and discipline-revealing — but it is not the most per-share-accretive policy available, because it returns cash via (taxable) dividend rather than via buybacks that would shrink the share count.

The buyback question — direct. [FACT] Buybacks: $0.439B / $0.676B / $0.700B / $0.903B (FY22–25). Against ~$860M of FY25 SBC and a ~$430B market cap (~49× P/E), the buyback retires well under 0.25% of shares/year — it barely exceeds dilution. [INTERPRETATION, applying a returns-focused lens] At ~49× earnings, the minimal buyback is the correct decision, not a flaw — repurchasing a 2.9%-net-margin retailer at a ~2% earnings yield would be value-destructive relative to holding cash earning ~4–5% or reinvesting at 40% ROIC. The legitimate critique is a meta-one: management’s refusal to repurchase is itself a valuation signal — insiders are implicitly telling you the stock is too expensive to buy. Indeed, on the Q3 FY26 call management stated that “at our current valuation, special dividend is typically the most effective way to return excess cash” and noted the stock is “materially higher” than at the last special. This corroborates the skeptical valuation read.

CapEx and new-unit economics. [FACT] CapEx rose $3.59B→$5.50B (FY21–25, +53%), funding ~30 net new units/yr. [INTERPRETATION] Given the negative-working-capital model and ~40% operating ROIC of the mature fleet, new-unit economics are highly attractive; this is the single best use of cash and the reason the buyback can stay small. CapEx is funded entirely from operating cash flow (FY25 OCF $13.3B vs. CapEx $5.5B).

M&A — near-total abstinence. [FACT] The only material acquisition in recent memory is Innovel Solutions (last-mile logistics, ~$1.0B, 2020), rebranded Costco Logistics. [INTERPRETATION] This abstinence is a positive signal under the Marathon lens — no overpriced deals to manufacture growth, no goodwill-impairment risk, no integration drag, no empire-building.

Executive comp & alignment (DEF 14A, filed 2025-12-04). [FACT] Comp is unusually simple: base, cash bonus, and performance RSUs. The FY2025 performance-RSU criteria were a 3% increase in net sales OR a 2% increase in pre-tax income (FX-adjusted) versus FY2024; the Committee determined both were exceeded. [INTERPRETATION, direct] Those hurdles are trivially low — thresholds Costco clears almost automatically; the “performance” gate is effectively a formality, and the metrics reward size, not return on capital or per-share value. A mild governance weakness, though modest in dollar terms (SBC ~0.3% of revenue) and free of egregious option mega-grants. CEO Vachris and CFO Millerchip elected three-year vesting on FY25 grants — a small alignment positive.

Insider ownership. [FACT, correcting external data] Per the 2025 proxy, all directors and executive officers as a group (23 persons) own 460,270 shares — “less than 1%” of shares outstanding (the “16.3%” figure circulating in third-party data feeds is incorrect — it conflates with an aggregate/institutional number). Largest holders are index funds (Vanguard ~9.8%, BlackRock ~7.8%). [INTERPRETATION] Sub-1% insider ownership means alignment runs through annual RSU grants and culture, not a large founder stake — neutral-to-mildly-negative, typical of a mega-cap, and it removes the “owner-operator skin-in-the-game” argument some bulls make.

Verdict: Yes — capital is allocated intelligently, with one soft spot. Reinvesting at ~40% operating ROIC into new warehouses, funding it from a self-financing negative-working-capital model, paying a covered growing dividend, returning excess via tax-aware specials, avoiding overpriced M&A, and declining to buy back stock at 49× — a coherent, disciplined, value-rational program. Disconfirming evidence weighed: soft comp hurdles (a governance nit), sub-1% insider ownership (weakens the alignment story), and the choice of specials over more-accretive buybacks. None is thesis-breaking. The honest framing: management allocates capital well and is implicitly signaling, through its refusal to repurchase, that it considers its own stock expensive.

8. Changes and Headwinds — Last Two Years

Leadership transitions (executed cleanly). [FACT] Two simultaneous top-of-house transitions in 2024: Ron Vachris (a 40-year Costco veteran, ex-COO, who started as a forklift driver) succeeded Craig Jelinek as CEO (Jan 2024); Gary Millerchip (ex-Kroger CFO) succeeded Richard Galanti as CFO (March 2024) after Galanti’s ~38-year tenure. [INTERPRETATION] Both were promote/recruit-from-strength successions; cultural continuity (low markups, high wages, treasure-hunt SKUs) appears intact through two earnings years. Galanti’s departure removed the Street’s single most-trusted communicator — a minor information-quality loss — but Millerchip has executed the IR cadence smoothly.

September 2024 membership fee increase. [FACT] First increase since 2017: US/Canada Gold Star $60→$65, Executive $120→$130 (higher 2% reward cap), effective Sept 1, 2024. Per the Q3 FY26 call, the increase accounted for “a little more than 1/4” of MFI growth; excluding it and FX, MFI grew 7%. [INTERPRETATION] Recognized ratably over the membership year, it is a multi-year tailwind to MFI (and, at ~100% incremental margin, to operating income) running through FY25–FY27. Renewal did not deteriorate post-increase (92.2%) — textbook Greenwald pricing power validated by data.

Tariffs / IEEPA + member-refund lawsuit. [FACT] Costco began submitting IEEPA tariff refund claims via US CBP, expecting refunds “over the next few months,” and referenced “a lawsuit filed against the company regarding the return process.” [INTERPRETATION] Tariffs are a margin/price headwind being actively managed via sourcing shifts; the IEEPA refund is a modest potential tailwind; the returns lawsuit is [OPEN QUESTION] not yet quantified in filings — appears immaterial but flag for monitoring.

Membership-growth deceleration. [FACT] Total paid members +4.1% at Q3 FY26 (82.9M), down from higher-single-digit prior-year growth. [INTERPRETATION] The most important watch item in the thesis — because MFI compounding at near-100% incremental margin is the profit engine, a sustained slide toward low-single-digit member growth would cap the MFI growth the 49× multiple underwrites. So far offset by the fee increase, Executive mix shift, and international additions; renewal at 92.2% is reassuring (the base is sticky); the question is top-of-funnel new-member velocity.

Other developments. [FACT] China/international expansion (strong early China adoption; same-day delivery in Spain/France); gasoline and FX continue to swing reported sales/margin (Q3’s +11.6% net sales was gas- and FX-inflated). [FACT, financial media, 2026-05-29] Q3 FY26 “beat across the board” but shares fell on a soft market reaction; sell-side maintained ratings (BofA Buy → $1,200 target; Truist Hold → $1,011). [INTERPRETATION] A beat-and-fade on a 49× stock signals a high embedded-expectations bar — valuation, not fundamentals, is the swing factor.

Verdict: Net strengthen, with one watch item. The fee increase (validated by an undamaged 92%+ renewal) and the smooth dual C-suite succession strengthen the franchise; tariffs and the returns lawsuit are manageable second-order headwinds; the IEEPA refund is a small positive. The one genuine weakening vector is member-growth deceleration to 4.1%, because it pressures the exact engine that justifies the premium. Disconfirming evidence weighed: headline growth is gas/FX-flattered, the fee increase is a one-time step not a recurring lever, and member adds are slowing — all fair; the rebuttal is that renewal at 92.2% and Executive-mix richening keep MFI growing high-single-digits even with slower member-unit adds. The operating thesis is intact and modestly stronger; the valuation thesis is where the risk sits.

9. Risk Analysis

Risks scored by likelihood × impact with the evidence basis. The headline: operational and competitive risks are remarkably well-contained — the dominant risk is valuation / multiple compression, a market risk, not a business risk.

Ranked: (1) Valuation/multiple compression — the only risk simultaneously likely and severe, requiring zero operational failure; everything else is secondary. (2) Growth deceleration — the fundamental risk most likely to trigger (1). (3) Gas normalization + FX — near-certain optics headwinds. (4) Tariff/lawsuit, succession, labor — manageable, matter mainly as potential catalysts for the de-rate. (5) Amazon/e-comm, recession-ticket, intl execution — slow-burn watch-items. (6) Cyber (tail) and margin self-cap (structural feature).

Verdict: The cleanest and most lopsided risk profile imaginable for a retailer. Operationally, Costco is about as de-risked as retail gets — fortress balance sheet, recurring membership cash, top-tier renewal, share-gaining value model. The risk that actually matters is the price paid for that safety. A buyer at 49× is underwriting multiple persistence; the asymmetry (large de-rating downside vs. capped expansion upside off a record base) means market risk dwarfs business risk — an unusual and important inversion of the typical retail risk stack.

10. Valuation Discussion

Embedded-expectations and scenario analysis only. No price target. No recommendation.

Where the multiples sit. At ~$969 (2026-06-10): market cap ~$429.5B; enterprise value ~$419B (a fortress net-cash balance sheet pulls EV below market cap, unusual for a retailer); trailing P/E 48.8×; forward P/E (FY26E ~$20.6) ~43×; EV/EBITDA ~30×; P/S 1.46×; dividend yield ~0.6%.

Peer multiple table (public market data, 2026-06-10):

[FACT/INTERPRETATION] The single most important comparison: COST 48.8× vs. BJ’s 20.6× — BJ’s is the closest structural analog (a membership warehouse club with renewal-annuity economics), yet COST commands a ~2.4× P/E and ~2.3× EV/EBITDA premium. Against mass/discount (TGT 16.7×, DG 15.5×), COST is a ~3× earnings multiple; even against Walmart — itself near a record own-history valuation — COST is ~16% more expensive on P/E and ~30% on EV/EBITDA. COST is the single most expensive large-cap name in staples retail.

Costco vs. its own history. [FACT] A 10-year own-history valuation index (vs. COST’s own ~10-year distribution; 0 = cheapest, 100 = most expensive) reads: P/E percentile 80.2, P/B 77.1, P/S 84.2, composite 80.5. COST sits in the ~80th percentile of its own decade — this expensive or more only ~20% of the time in ten years, and near the richest it has ever been on price-to-sales. [INTERPRETATION] Even relative to its own historically premium average, COST is expensive today; a reversion merely to its own 10-year median implies meaningful compression with zero change in the business.

Embedded-expectations / reverse-DCF. A ~49× trailing P/E is a ~2.0% earnings yield; with the ~0.6% dividend, the equity offers a ~2.6% “all-in” starting yield before growth. For that to clear a reasonable ~8–9% equity hurdle (beta-0.87, bond-like), the market must underwrite ~10–12% nominal EPS compounding for a long time with no terminal multiple compression. Decomposing the achievable EPS growth from disclosed drivers: unit growth ~3.0–3.2%/yr (30+ units on a 928 base, decelerating); ex-gas/FX comp ~6–7% (steady); blended consolidated sales growth normalizes to ~8–9% once gas/FX/fiscal noise washes out; MFI growing ~7% ex-fee-hike with periodic step-function increases; thin operating-margin accretion of ~10–30bp/yr (much of it MFI-mix). Stack it: ~8–9% sales + thin leverage → ~10–12% EPS growth is a reasonable base, consistent with the historical ~13–15% CAGR decelerating as the law of large numbers bites. The verdict: at 49×, the market is paying a full, fair price for ~10–12% EPS compounding for 10+ years and assuming the ~49× multiple barely compresses. The embedded expectation is reasonable on growth; it is heroic on multiple persistence.

To make the multiple-persistence assumption explicit: if EPS compounds ~11%/yr for five years (~$18.21 → ~$30.7) and the exit P/E holds at 49×, price roughly mirrors EPS (~11%/yr). But if the exit multiple drifts to 35× (still a large peer premium), the 5-yr price CAGR collapses to ~3–4%; at 30× it is roughly flat-to-slightly-negative over five years despite ~70% cumulative EPS growth. This is the entire investment risk in one sentence: the business can do everything right and the stock can still go nowhere.

Scenario analysis (5-year) — illustrative multiple ranges only, no point target:

[INTERPRETATION] The dispersion is almost entirely the exit-multiple column. Even the bull case requires the market to keep paying ~45–50× — a bet on multiple persistence, not operations. The bear case requires no operational failure at all — merely a reversion toward where every direct club peer trades. This asymmetry is the defining feature of the COST setup.

Where the premium is justified vs. stretched. Justified: the ~$5.5B run-rate membership annuity (92.2% renewal, near-100% incremental margin, genuinely bond-like); durability and low beta (0.87); the Kirkland and Executive-upgrade optionality; the high-ROIC reinvestment runway. Stretched: it is, at the end of the day, a ~3.8%-operating-margin grocer with ~3% unit growth wearing a 49× multiple; member growth has decelerated to +4.1%; reported EPS is gas-/FX-flattered; and core-on-core margin is declining by management’s own choice, capping the operating-leverage half of the EPS algorithm. Special-dividend optionality (a ~$10–15 special roughly every 2–3 years) is real and supports a floor but is a ~1–1.5% return event — color, not thesis.

Verdict: Great business, demanding price. The premium is partly earned by the membership annuity, durability, and reinvestment runway; it is stretched by an absolute 49× / ~30× EV-EBITDA multiple at the ~80th percentile of COST’s own decade on a 3.8%-margin business with decelerating member growth. The dominant valuation question is not “will COST execute?” (it will) but “will the market keep paying a record multiple for that execution?”

11. Variant Perception

Consensus view. The Street holds a “quality-compounder, own-it-forever, pay-(almost)-any-price” view: a generational business with a self-reinforcing membership moat, a 92%+ renewal annuity, decades of (especially international) unit runway, and management discipline — and therefore a multiple that is “always expensive and always right.” WS target ~$1,082 (~12% above spot) embeds continued premium-multiple persistence. [FACT] Short interest is ~1.7% of float (short ratio ~4.1) — not a crowded short; almost no one is positioned against it, which is itself a mild contrarian warning: consensus is one-sided and the multiple has no skeptical float pushing back.

Strongest bull case. The bull does not need to win on valuation — it needs the business to be so good the multiple is justified and persists: the ~$5.5B MFI annuity compounding ~7% ex-fee-hike with 92.2% renewal and Executive members +9.6%, the closest thing in retail to a subscription; decades of unit/international runway (China Executive launch, Japan, Europe, with new markets historically driving outsized member growth); Kirkland as a structural cost/pricing flywheel deepened by vertical integration; pricing power without price hikes (“first to come down, last to go up,” gaining share in stress); a fortress balance sheet (~$45/share cash) with special-dividend optionality. Bull in one line: a bond-like, share-gaining, low-beta compounder with a widening moat deserves to trade like a perpetuity — and 49× is what a perpetuity of 10–12% growth costs.

Strongest bear case. The bear case is almost entirely valuation — and that is exactly why it is dangerous. No fundamental deterioration is required for a poor 3–5 year outcome. At 49×, compression toward even 30× (still a premium to every direct peer) is ~−40% with the business performing flawlessly; at WMT’s 42× it is a mid-single-digit annual drag; at BJ’s 21× it is a halving. The market is paying a record, ~80th-own-percentile price for a 3.8%-margin grocer whose member growth has decelerated to +4.1% (“lowest in some time”), whose reported comps are gas-/FX-flattered (ex-gas/FX +6.6% vs. headline +11.6%), and whose core-on-core margin is contracting (−9bps). Layer in the dual key-person transition, tariff/IEEPA overhang plus the member-refund lawsuit, FX on ~27% international, and the eventual gas-normalization comp headwind, and you have a stock priced for perfection with several plausible disappointment vectors — any one of which could be the narrative catalyst for the de-rate the arithmetic already invites. Bear in one line: the best business in retail at the wrong price is still a poor investment; the only question is whether the multiple re-rates on a stumble or on its own gravity.

The 3–5 assumptions that matter most. (1) Will the market sustain a ~45–50× multiple, or de-rate toward a high-quality-retailer multiple (30–38×) as member/unit growth matures? (the highest-impact swing factor). (2) Can ex-gas/FX comps + MFI hold ~7–9% blended for a decade, or does the law of large numbers + US saturation pull normalized growth toward ~5–6%? (3) Does international (esp. China/Asia) inflect to a new growth engine, or has the highest-return domestic phase peaked? (4) Does the new CEO/CFO duo preserve the price-leadership culture and capital discipline? (5) Will membership economics survive a real consumer recession (renewal + ticket)?

What falsifies each side. Bull falsified by a two-plus-year trend of ex-gas/FX comp below ~5%, and/or renewal sliding below ~91% US/Canada, and/or member growth stuck ≤4% with no international re-acceleration. Bear falsified by ex-gas/FX comps re-accelerating toward ~8%+, an international unit/membership inflection, a clean new fee increase, and the market holding the multiple through a recession.

Verdict: Consensus and bear are not arguing about the business — they agree COST is exceptional. They are arguing about price and multiple durability. The bull owns the business, the bear owns the multiple, and the next five years will be decided by which mean-reverts first — the growth rate (toward the multiple) or the multiple (toward the peers).

12. Fact vs. Interpretation Table

13. Open Questions

1. Mark-to-market value of owned real estate (~80%+ of warehouses owned) — a material, undisclosed asset-value cushion potentially tens of billions above book.
2. Quantified exposure/reserve for the member returns-process lawsuit and the dollar magnitude/timing of recoverable IEEPA tariff refunds.
3. Whether the full 248-Form-4 corpus confirms zero open-market (code-P) insider purchases across the 5-year window (the recent sample strongly implies it).
4. The cadence of new-market (especially Asia) warehouse openings — the lever that determines whether member growth re-accelerates from +4.1% or grinds toward low-single-digits.
5. Exact Kirkland Signature penetration and KS-specific gross margin — not disclosed; the ~1/3-of-sales figure is an analyst estimate.
6. Current Sam’s Club and BJ’s renewal rates and per-club AUV — needed to size Costco’s intra-channel lead precisely.
7. Whether the new CEO/CFO duo preserves the price-leadership-and-discipline culture through a genuine downturn — unproven, as their tenure has coincided with a benign environment.

14. What Must Be True (Bull and Bear, with Falsification Tests)

Bull case — what must be true:

1. The membership annuity keeps compounding high-single-digits — renewal stays ≥92% US/Canada, Executive-tier mix keeps richening, and periodic fee increases land without renewal damage.
2. International (especially Asia) inflects into a genuine new growth engine, re-accelerating member growth above ~5% and extending the unit-growth runway a decade-plus.
3. The market keeps paying a ~45–50× multiple — i.e., COST is durably treated as a quasi-perpetuity, not a cyclical retailer.

Falsification test (bull): Two-plus years of ex-gas/FX comp trending below ~5%, and/or renewal sliding below ~91% US/Canada, and/or member growth stuck ≤4% with no international re-acceleration. Any sustained trend (not a single print) here breaks the perpetuity-growth premise.

Bear case — what must be true:

1. The multiple mean-reverts toward club/quality-retail norms (30–38×) — whether triggered by a growth stumble or simply by its own gravity off a record base.
2. Normalized growth decelerates as US saturation, slowing traffic (~2%), and the law of large numbers pull EPS growth toward ~7–8%, removing the justification for a perpetuity multiple.
3. A narrative catalyst (gas-normalization comp headwind, a soft member-growth print, a succession or tariff/lawsuit wobble) crystallizes the de-rate.

Falsification test (bear): Ex-gas/FX comps re-accelerate toward ~8%+, an international unit/membership inflection materializes, a new fee increase lands cleanly, and the market holds the multiple through a recession — i.e., the franchise proves it compounds fast enough and is durable enough that 49× converts from “expensive” to “fairly priced for the quality.”

15. Source Appendix

Full source detail in the standalone Source Appendix (Appendix B of the combined report). Primary sources, prioritized:

1. Costco FY2025 Form 10-K (period ended 2025-08-31, filed 2025-10-08) — income statement, segment note, membership tables, merchandise categories, comparable-sales table, gas/e-commerce penetration, Kirkland description.
2. Form 10-Q, Q3 FY2026 (12 weeks ended ~2026-05-10, filed 2026-06-03) — quarterly financials.
3. Q3 FY2026 earnings call transcript (2026-05-28) — membership counts/renewal, MFI, comps, traffic/ticket, tariffs/IEEPA, gas, China Executive launch, capital allocation, valuation commentary.
4. Q1 FY2026 earnings call (2025-12-11) and prior-quarter calls — trajectory.
5. DEF 14A proxy (filed 2025-12-04) — executive comp metrics, insider ownership, beneficial-ownership table.
6. 8-K corpus (FY24–FY26) — $15 special dividend (2023-12-14), CEO/CFO transitions, fee increase, earnings releases.
7. Form 4 corpus (~248 filings, 5-yr) — insider transaction read (zero open-market purchases in recent sample).
8. SEC EDGAR XBRL company facts (CIK 0000909832) — multi-year revenue, net income, operating income, equity, cash, debt reconciliation.
9. Public market data (2026-06-10) — COST and peer multiples (WMT, AMZN, BJ, TGT, DG, KR).
10. Third-party financial data & media — fundamentals snapshot, 10-yr own-history valuation index, financial-media coverage (2026-05-29). Third-party signal; reconciled to filings.
11. Walmart public disclosures — warehouse-club (Sam’s Club) framing and Costco renewal-moat cross-read.
12. Investment-research frameworks — Greenwald & Kahn, Competition Demystified; Marathon/Chancellor, Capital Returns.

Independent research article. The body carries no investment recommendation and no price target; “Author’s Take” is a clearly-labeled exception and reflects the author’s own opinion. General information only — not investment advice.

APPENDIX A — Standard Diligence Questionnaire

Costco Wholesale (NASDAQ: COST) — Standard Diligence Questionnaire Appendix

Supplemental to the research article. Answers are labeled Fact / Interpretation / Assumption where it matters. As-of 2026-06-10.

General

What thoughtful questions have other investors asked about this company? The recurring institutional debate is not about business quality (universally acknowledged as elite) but about price and multiple durability: (1) Is a 49× P/E / ~30× EV/EBITDA defensible for a 3.8%-operating-margin retailer with ~3% unit growth? (2) How much of the profit is really the membership annuity, and how fast can MFI compound? (3) Is member-growth deceleration to +4.1% a saturation signal or a temporary new-market-opening gap? (4) Can the post-Jelinek/Galanti management preserve the culture? (5) How much is current EPS flattered by gas and FX? [Interpretation] The most sophisticated bears concede the franchise and short only the multiple — a “right business, wrong price” stance.

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? [Interpretation] Neither extreme, but the reported run-rate is modestly flattered — Q3 FY26 net sales +11.6% included ~2.2pts of gas-price inflation and ~1pt of FX; underlying ex-gas/FX comp was +6.6%. Gas volumes hit all-time records on a Middle-East oil-supply shock, a non-recurring tailwind. Normalized earnings power is high-single-digit comp + ~3% units → ~10–12% EPS growth.

Driven by the external environment or internal actions? [Interpretation] Predominantly internal (membership model, Kirkland, price leadership, unit growth, Executive-tier conversion), with external swing factors (gas prices, FX, tariffs) that distort the optics quarter to quarter but not the underlying trajectory.

How stable are revenues? [Fact] Exceptionally stable — MFI (~$5.5B run-rate) renews at 92.2% US/Canada; core ex-gas/FX comp has held a remarkably steady 6–7% for over a year. Few retailers have a more predictable revenue base.

Outlook for products/services? [Interpretation] Durable secular demand (value/staples-led, gains share in consumer stress), with capital-light digital/retail-media and pharmacy/ancillary as incremental growth layers.

How big will this market be — growing, shrinking, domestic or international? [Interpretation] The US warehouse-club channel is mature but still growing; the multi-decade growth frontier is international (China, Japan, Korea, Europe), which already earns higher segment margins than the US. Total addressable runway supports 30+ net new clubs/year “in the coming years.”

Business Quality & Competitive Moat

Is the industry getting more or less competitive? [Interpretation] The warehouse-club niche is a rational three-player oligopoly (Costco/Sam’s/BJ’s) with disciplined capital deployment and no credible new entrants — stable, not intensifying. Broader retail is hyper-competitive, but the membership wall insulates Costco from the markup war.

How profitable is the business (ROIC, ROE)? [Fact/Interpretation] Operating ROIC ~40% once the ~$15–20B cash hoard is excluded from invested capital (~19–24% unadjusted); ROE ~28% (FY25 27.8%) with near-zero financial leverage. Elite returns.

How profitable is the industry — competitors, barriers to entry? [Interpretation] Club-channel returns are high and protected by scale (buying power, depots, ~$6.5B CapEx), the membership wall, and Kirkland. Barriers are real — the model requires enormous scale to function — which is why the capital cycle is favorable (high returns not attracting mean-reverting capital).

Can the business be easily understood? [Fact] Yes — among the most transparent models in retail: sell goods near breakeven, earn the profit on the renewing membership fee.

Can it be undermined by foreign low-cost labor? [Interpretation] No — it is a domestic-footprint, real-estate-and-logistics business; the relevant cost lever is buying power, not labor arbitrage, and Costco deliberately pays above-market wages for retention/productivity.

Do brands matter? [Fact/Interpretation] Yes — Kirkland Signature (15–20% below national brands at equal/better quality, exclusive to Costco) is both a margin lever and a switching cost; the Costco brand itself signals trust/value.

What is the nature of competition? [Interpretation] Price and value-per-trip, where Costco’s self-capped ~11% gross margin makes it nearly impossible to undercut profitably (including by Amazon).

Customers’ switching costs? [Fact/Interpretation] The annual fee creates sunk-cost/consolidation psychology; the 2% Executive reward formalizes it; renewal at 92%+ is the measured proof. Caveat: it is customer captivity, not a true network effect — durable but not self-reinforcing in the strongest sense, and partly geographic (warehouse proximity).

Financial Condition & Balance Sheet

Assets not fully recognized on the balance sheet? [Fact/Open Question] Yes — Costco owns ~80%+ of its real estate, carried at depreciated cost; the mark-to-market value of owned land/buildings is plausibly tens of billions above book (an undisclosed Greenwald asset-value cushion).

Off-balance-sheet liabilities? [Fact] Modest — operating-lease liabilities ~$2.67B total (small for a 928-club operator given the high owned-real-estate share). No material off-balance-sheet financing.

How conservative is the accounting? [Interpretation] Conservative — LIFO inventory (charges flow through gross margin), clean ~95–100% FCF conversion (net income tracks operating cash flow), low SBC (~0.3% of revenue), no aggressive revenue recognition (MFI recognized ratably over the membership year).

How CapEx-hungry is the business? [Fact] Moderately — FY26 CapEx ~$6.5B (~2.4% of revenue), funded entirely from operating cash flow (FY25 OCF $13.3B). New-unit CapEx earns ~40% operating ROIC, and the negative working capital means comp growth releases cash rather than consuming it.

Capital Allocation & Management

How much FCF does the business generate, and how is it used? [Fact] FCF ~$7.8B (FY25). Uses, in priority: (1) reinvest in warehouses/depots/digital at ~40% ROIC; (2) grow the regular dividend (~28% payout); (3) token buyback (offsets SBC); (4) periodic special dividends for excess cash. Philosophy: disciplined, value-rational, no permanent leverage.

Significant acquisitions recently? [Fact] Essentially none — the only material deal in recent memory is Innovel Solutions (~$1.0B, 2020, → Costco Logistics). M&A abstinence is a positive signal (no overpriced empire-building).

Buying back shares? [Fact/Interpretation] Only token amounts ($0.903B FY25, <0.25% of shares/yr) — barely above SBC dilution. At 49×, this is the correct decision, and management’s stated preference for specials “at our current valuation” is a self-aware valuation signal that the stock is too expensive to repurchase.

Issuing large amounts of new shares to insiders? [Fact] No — SBC ~0.3% of revenue, overwhelmingly RSUs, no large option grants.

Compensation policy of directors/management? [Fact/Interpretation] Simple (base + cash bonus + performance RSUs) and modest in dollars, but the performance hurdles (3% sales OR 2% pre-tax income, FX-adjusted) are trivially low — a mild governance weakness that rewards size, not return on capital or per-share value. CEO/CFO elected three-year vesting (small positive).

Motivations of management? [Interpretation] Culture-driven (the “Costco way” — low markups, high wages, member value first) rather than equity-stake-driven; insider ownership is <1%, so alignment runs through annual grants and culture, not a large founder stake.

Valuation & Market Data

Is the stock an ADR, MLP, or K-1 issuer? [Fact] No — common stock of a US C-corporation (NASDAQ: COST). Standard 1099 dividend reporting.

Dividend policy? [Fact] Growing regular dividend (~$1.30/qtr; forward ~$5.88/yr; yield ~0.6%; payout ~28%) plus periodic large special dividends ($7 in 2012, $5 in 2015, $7 in 2017, $10 in 2020, $15 in 2024).

How profitable is the business? [Fact] Net margin ~2.9%, operating margin ~3.8% — thin on the surface, but the relevant return metric is ~40% operating ROIC and ~28% ROE.

Is net income diverging from cash from operations? [Fact] No — FCF conversion ~95–100%; OCF (FY25 $13.3B) comfortably exceeds net income ($8.1B) and funds CapEx with room to spare. A clean quality-of-earnings signal.

Risks & Downside

What factors would cause the stock to decline? [Interpretation] Primarily multiple compression (a de-rate from 49× toward 30–38× is ~−25% to −40% with no fundamental change), most likely triggered by: a sustained member-growth/comp deceleration, the gas-normalization comp headwind, an FX reversal, a tariff/lawsuit wobble, or a succession stumble. The business risk is low; the price risk is high.

Risk of a catastrophic loss? [Interpretation] Very low — fortress net-cash balance sheet (~$20B gross cash), recurring membership cash, no refinancing/liquidity risk, defensive demand. A catastrophic business outcome is hard to construct.

Chance of a total loss? [Interpretation] Negligible — this is not a solvency or obsolescence story. The realistic downside is a multi-year period of poor equity returns driven by multiple compression, not impairment of the enterprise.

Recent News & Events

Has the business environment changed recently? [Fact] Q3 FY26 (reported 2026-05-28) “beat across the board” but shares fell on a soft market reaction — a high embedded-expectations bar. Gas volumes hit records on a Middle-East oil shock; member growth decelerated to +4.1%; the Sept-2024 fee increase continues to lap through MFI; IEEPA tariff refund claims are being filed amid a member-refund-process lawsuit.

Significant acquisitions? [Fact] None recent.

Change in accounting policies? [Fact] None material.

Recent changes — new markets, facilities, management? [Fact] CEO (Vachris) and CFO (Millerchip) both transitioned in 2024; Executive Membership launched in China in Q3 FY26; continued international expansion (same-day delivery in Spain/France); warehouse count reached 928 (targeting 30+ net new/yr).

APPENDIX B — Source Appendix

Public primary sources used in this article, prioritized primary-first. As-of 2026-06-10. All quantitative figures reconcile to SEC filings.

1. SEC Filings (primary — US filer, CIK 0000909832; available at SEC EDGAR)

2. Earnings Call / Event Transcripts (primary — company investor relations)

3. Market & Reference Data

4. Analytical Frameworks

Primary sources (SEC filings, transcripts) take precedence over secondary/third-party data throughout. Every material number reconciles to a Costco SEC filing; third-party aggregator data was used for orientation, peer comps, and sentiment signals only, and is flagged as such.