Decision boundary
This learner packet stops at December 23, 1998, 11:59:59 p.m. Pacific time. The decision is deliberately hypothetical: on the morning of December 24, should Costco's board commit the next expansion phase to its membership-first operating system, execute management's announced plan without added gates, harvest more fee or merchandise margin, or slow owned growth while testing selective alternatives? Nothing in the packet asserts that an actual board meeting or vote occurred on that date.
Management's observed action at the cutoff is narrower. Costco said it planned a net 18-20 new warehouse clubs in fiscal 1999 plus seven relocations, after six net openings in the first quarter. Its fiscal 1998 annual filing had described approximately $525-$575 million of U.S./Canadian capital and $75-$125 million of international capital, explicitly subject to change. Those disclosures establish a baseline plan, not the hypothetical board's recommendation or the plan's later result. [evidence.costco.cutoff.expansion-1998; evidence.costco.cutoff.q1-expansion-footprint]
What business is the board scaling?
The membership fee is visible, recurring revenue, but it is not the whole business and is not synonymous with profit. Costco separately reported merchandise net sales, membership fees and other revenue, merchandise costs, SG&A, preopening expense, operating income, and net income. It defined gross margin as net sales minus merchandise costs. The underwriting question is therefore whether a system of mechanisms reinforces member value and warehouse economics—not whether one can compare fee revenue with operating income and call the difference causal. [evidence.costco.cutoff.reported-results; evidence.costco.cutoff.expense-waterfall; evidence.costco.cutoff.gross-margin-1998; evidence.costco.cutoff.operating-net-income]
The disclosed system has distinct parts:
- Member value and loyalty. Costco said its format was designed to reinforce loyalty and create continuing fee revenue. At August 30, 1998, it reported approximately 3.7 million Business memberships and 8.7 million Gold Star memberships. These are membership categories—not unique people, households, or cardholders. [evidence.costco.cutoff.membership-policy-1998; evidence.costco.cutoff.membership-counts-1998]
- Low price and limited selection. Management described very low prices and a limited branded/private-label assortment. It said only products offering meaningful member savings were carried and reported roughly 3,600-4,400 active SKUs per warehouse, far fewer than conventional discounters and supermarkets. This selection discipline is separate from the fee mechanism. [evidence.costco.cutoff.operating-model; evidence.costco.cutoff.sku-value-policy]
- Treasure hunt and assortment judgment. The frozen filing supports limited, fast-selling assortment and significant savings, but this packet does not promote a numerical “treasure-hunt” cadence or economics because the source set does not provide a governed measure. The board should monitor rotation and member response rather than turn a familiar label into unsupported causality. [evidence.costco.cutoff.sku-value-policy]
- Scale, vendor terms, distribution, and handling. Costco attributed its low-margin viability to volume purchasing, efficient distribution, reduced handling, rapid turns, and no-frills facilities. It further claimed that mature warehouses often sold much inventory before vendors were paid and still used early-payment discounts. These are company mechanism claims, not warehouse-level causal proof. [evidence.costco.cutoff.operating-model; evidence.costco.cutoff.vendor-working-capital]
- Operating-cost discipline. Costco generally limited advertising to openings and occasional direct mail. This is one cost mechanism; it should not be conflated with merchandise margin, fee economics, employee cost, or vendor financing. [evidence.costco.cutoff.low-cost-operations]
- Employee execution. Costco reported about 63,000 employees, roughly 90% hourly, with about half of hourly employees part-time and approximately 11,000 represented by the Teamsters. The filing establishes the workforce perimeter, not whether compensation and benefits caused better productivity or renewal. [evidence.costco.cutoff.employees-1998]
- Geographic and format scale. At November 22, 1998, Costco reported 284 operated warehouses across the United States, Canada, the United Kingdom, Korea, and Taiwan, plus 15 Mexico warehouses through a 50%-owned equity-method venture. Consolidated operations and the Mexico equity-method perimeter must remain separate. [evidence.costco.cutoff.q1-warehouse-perimeter; evidence.costco.cutoff.q1-equity-method-perimeter]
- Emerging channel option. Costco already identified Internet commerce as increasingly popular, alongside warehouse clubs and category specialists. The packet has no Costco e-commerce revenue, capital, fulfillment-cost, cannibalization, or member-retention data; digital is an option to test, not an evidence-backed growth forecast. [evidence.costco.cutoff.competition-1998]
Reference class and history
The 1993 Price/Costco merger was expected to create a company with about $16 billion in annual sales, 196 stores, more than 28,000 employees, and greater buying clout. That was a contemporaneous expectation, not proof that scale later caused margin or member outcomes. The 1994 filing also says pooling-of-interests accounting restated the companies as if combined from inception and conformed policies and interim periods. Pre-merger series should not be read as untouched historical company results. [evidence.costco.cutoff.merger-expectations; evidence.costco.cutoff.merger-accounting]
The combination also produced a useful warning: Costco reported a decline in renewals from overlapping memberships and reciprocal privileges. Membership totals and renewal can be affected by structure and definitions rather than customer-value deterioration alone. [evidence.costco.cutoff.merger-membership-overlap]
Shareholders approved changing the issuer name from Price/Costco, Inc. to Costco Companies, Inc. on January 29, 1997, and the filing identified the COST ticker. That legal-name and ticker transition is a presentation event, not another merger, acquisition, or operating gain. [evidence.costco.cutoff.name-change-1997]
Cornell's contemporary industry study identified Sam's Club, Price Costco, and BJ's as the three major club operators. It described steep late-1980s/early-1990s growth followed by slower—but not stagnant—growth. Costco's filing identified approximately 800 clubs across the United States and Canada, category specialists with substantial share, and emerging Internet retail. The class offered scale, but not an uncontested runway. [evidence.costco.cutoff.reference-class; evidence.costco.cutoff.competition-1998]
Financial reconstruction
All amounts below are company-reported, in $000, and not analyst-derived. Fiscal 1996 contained 52 weeks; the cited filing contrasts it with a 53-week fiscal 1995 and also identifies opening, closure, and existing-location effects. [evidence.costco.cutoff.fiscal-comparability-1996]
| Reported measure ($000) | FY1996 | FY1997 | FY1998 |
|---|---|---|---|
| Net sales | 19,213,866 | 21,484,118 | 23,830,380 |
| Membership fees and other | 352,590 | 390,286 | 439,497 |
| Gross margin, defined as net sales less merchandise costs | 1,868,551 | 2,169,633 | 2,450,689 |
| Operating income | 490,723 | 580,712 | 787,276 |
| Net income | 248,793 | 312,197 | 459,842 |
Source lineage: [evidence.costco.cutoff.reported-results; evidence.costco.cutoff.margin-causes-1996; evidence.costco.cutoff.gross-margin-1998; evidence.costco.cutoff.operating-net-income; evidence.costco.cutoff.net-income].
In fiscal 1998, management attributed membership-fee growth partly to 18 new warehouses and a $5 annual fee increase. It separately attributed gross-margin improvement to ancillary mix, depot use, and international operations. Earlier, management had attributed fiscal 1996 margin improvement to buying power after the merger, shrink, depots, and ancillary penetration. These are issuer attributions and multiple co-causes; they do not justify saying “membership fees caused the profit.” [evidence.costco.cutoff.fees-accounting-margin; evidence.costco.cutoff.gross-margin-1998; evidence.costco.cutoff.margin-causes-1996]
At August 30, 1998, Costco reported $361.974 million of cash and equivalents, $75.549 million of short-term investments, $1.910751 billion of net merchandise inventory, and $1.605533 billion of accounts payable. At November 22, inventory was $2.430667 billion, payables $2.059702 billion, and deferred membership income $216.803 million. Those are snapshots. Seasonality, openings, receipts, payment timing, and accounting basis can change them; without average balances and a governed deterministic calculation, this packet abstains from publishing inventory turns, payable days, or cash conversion. [evidence.costco.cutoff.balance-1998; evidence.costco.cutoff.payables-1998; evidence.costco.cutoff.q1-balance; evidence.costco.cutoff.q1-payables-deferred]
The accounting transition
Costco announced that beginning in fiscal 1999 it would recognize membership-fee income ratably over the one-year membership life rather than on the prior cash basis. In the first quarter, the transition produced a $118.023 million non-cash after-tax cumulative charge. Reported net loss was $13.789 million; net earnings before that charge were $104.234 million. Membership fees and other revenue were $103.840 million, down 4% under the new basis, while net sales were $5.894238 billion, up 11%. A reader who treats the loss, the pre-charge income, or the fee decline as interchangeable would misread the quarter. [evidence.costco.cutoff.membership-accounting-change; evidence.costco.cutoff.q1-performance; evidence.costco.cutoff.q1-sales; evidence.costco.cutoff.q1-fees-margin]
Alternatives
1. Execute the announced plan as baseline. Continue the net 18-20 openings and seven relocations without imposing new board-level staged releases. This best matches management's disclosed plan, but the public packet lacks site-level returns, committed-versus-optional capital, cannibalization, and downside liquidity.
2. Gate membership-first scale. Continue committed openings and relocations, but separate follow-on releases. Require stable-definition member cohorts; matched price and assortment evidence; warehouse-level contribution and capital; deterministic inventory and vendor metrics; employee capacity and outcomes; and stressed liquidity. This preserves the operating architecture while paying for information.
3. Harvest fee and merchandise margin. Raise fees faster, accept higher merchandise margins, broaden assortment, or compress operating and employee cost. The path could lift near-term reported results, but it risks weakening the member savings and operating loop that management itself described. The prior fee increase proves that pricing was used, not that aggressive harvesting is optimal. [evidence.costco.cutoff.fees-accounting-margin; evidence.costco.cutoff.operating-model]
4. Slow physical growth and test options. Defer most new owned warehouses while testing relocations, joint ventures, ancillary services, or e-commerce. This reduces committed capital but may surrender scale, vendor leverage, or member convenience; the packet has no comparable return series for these alternatives.
Recommendation
Select gated membership-first scale, with moderate confidence. Protect low price, limited assortment, vendor terms, distribution, employee execution, membership value, and separate fee economics as one system. Fund actions already costly to reverse only after establishing their commitments; stage the rest.
The board should not approve a target price or declare a numerical return from this packet. Capitalization, contractual commitments, site economics, renewal cohorts, and deterministic model receipts are incomplete. Instead, every release should require:
- same-definition sign-up, renewal, visit, spending, and Executive-adoption cohorts;
- matched member price gaps, availability, SKU productivity, rotation, returns, and complaints;
- stable-week, stable-perimeter warehouse sales, gross margin, SG&A, preopening cost, contribution, capital, and cash;
- governed inventory-turn, aging, shrink, payable, discount, stockout, and vendor-concentration measures;
- staffing, wages, benefits, training, safety, turnover, promotion, grievance, and union capacity; and
- a downside liquidity test covering vendors, payroll, debt, leases, construction, and already-started sites.
Reverse or slow the plan if controlled cohorts deteriorate, price savings narrow, inventory aging or vendor terms weaken, worker/service outcomes fail, site economics miss stressed hurdles, or e-commerce/category specialists structurally impair the value proposition.
What remains unknown
The public record does not reveal the actual board package or resolution, warehouse-level returns, stable renewal cohorts, price gaps, average inventory and vendor aging, employee-outcome series, full contractual commitments, or e-commerce test economics. It supports a disciplined decision architecture; it does not support precise causal attribution, inventory-turn claims, fee-to-profit equations, valuation, or certainty.