Decision: November 16, 2001, 09:00 CET
Evidence cutoff: November 15, 2001, 23:59:59 CET
Question: Commit to a large Zara logistics site at Zaragoza, stage capacity, or use more outsourced and distributed alternatives?
Recommendation
Secure and design Zaragoza as modular redundant capacity, but do not treat the full build and automation program as one unconditional authorization. Release irreversible phases only after stable-scope peak throughput, twice-weekly service, committed store and selling-area demand, project economics, supplier-worker safeguards, and a twelve-month stressed-liquidity floor clear predeclared gates. Preserve central control of demand data, merchandising, pricing, allocation, and cadence; continue to use external producers, workshops, transport, and overflow capacity where enforceable service and safeguard terms outperform ownership. This is alternative.inditex.gated-zaragoza, supported by judgment.inditex.cutoff.gated-zaragoza at moderate confidence.
That recommendation is deliberately narrower than “vertical integration.” In FY2000, external producers supplied 56% of Zara merchandise on the prospectus's retail-price-participation basis; sewing was performed by roughly 400 external workshops, and about 80% of transported units moved by an external ground carrier. The system integrated information and decisions more completely than it integrated asset ownership. source · inditex.cutoff.ipo-2001
What the cutoff record establishes
Inditex closes its fiscal year on January 31. “FY2000” therefore means February 1, 2000 through January 31, 2001—not calendar 2000. The IPO prospectus reports Spanish-GAAP consolidated figures in millions of pesetas: revenue of 435,049, EBITDA of 86,768, EBIT of 63,208, and parent-attributable net income of 43,133 in FY2000, versus 268,665, 54,184, 40,183, and 25,480 respectively in FY1998. These are as-reported values in table.inditex.cutoff.reported-results; no silent euro conversion, constant-currency adjustment, perimeter normalization, or return attribution is applied. source · inditex.cutoff.ipo-2001
The physical footprint expanded at the same time. Stores rose from 748 at January 31, 1999 to 1,080 at January 31, 2001, of which 908 were group-managed and 172 franchised. Selling area rose from 320,500 to 536,600 square metres under a definition that includes sales floor, windows, and fitting rooms but excludes stockrooms and other back-of-house space. Store count and selling area are related but distinct capacity drivers; table.inditex.cutoff.footprint preserves both. source · inditex.cutoff.ipo-2001
The latest interim filing showed continued expansion. Unaudited H1 FY2001 revenue was EUR1,331.8 million, gross profit EUR678.0 million, EBITDA EUR225.7 million, EBIT EUR137.6 million, and attributable net income EUR91.7 million. The filing reports revenue growth of 26% and comparable-store growth of 9%, defining the latter as stores open for both full comparison periods translated at constant exchange rates. It also reports 68 openings, 1,148 stores at July 31, EUR215 million invested in H1, and a EUR400 million full-year investment budget. source · inditex.cutoff.h1-2001
El País described the same EUR1,331.8 million H1 revenue as 32% growth. That conflicts with the primary filing's 26%. The financial table uses 26%; conflict.inditex.cutoff.h1-revenue-growth leaves 32% disclosed and unresolved because no correction or alternate scope was captured by the cutoff. source · inditex.cutoff.elpais-h1
The operating mechanism
The most defensible mechanism is a linked control loop:
- Store teams communicate sales and customer reactions to commercial teams daily.
- Zara leaves meaningful external purchases and most internal production responsive during the season.
- Commercial teams set market-referenced retail prices first, then select materials and suppliers consistent with target margin.
- Stores receive a product offer and place orders twice weekly; the central distribution system replenishes and introduces new models on that cadence.
- Central allocation can move inventory through the distribution center rather than requiring direct store-to-store exchange.
This can shorten the delay between demand evidence and inventory action. The public record does not quantify its isolated contribution to gross margin, markdown, stockout, or owner return. It therefore supports a causal hypothesis for operating design, not an all-results attribution. source · inditex.cutoff.ipo-2001
Nor was the system inventory-free. FY2000 consolidated inventory was PES40.772 billion. Zara reported 65 inventory days under its issuer definition, stores carried roughly one month of sales, the main warehouse averaged 6-10 million garments, and third-party garments reportedly stayed about 37 days before the season and 15 days in season. Inventory amount, physical units, dwell days, store weeks, and rotation days are different measures; claim.inditex.cutoff.inventory-not-zero and table.inditex.cutoff.balance keep them separate. source · inditex.cutoff.ipo-2001
Capacity evidence cuts both ways
The principal Zara center at Arteixo moved about 2.5 million units per week in season and supplied stores twice weekly. From order receipt, reported delivery averaged 24-36 hours within Europe and 24-48 hours elsewhere. This cadence and the concentration of a global flow in one principal site support resilience and scale investment. source · inditex.cutoff.ipo-2001
But the same prospectus reported one shift for four days per week, only 50% of distribution area occupied, and 30,000 square metres expected to be freed when Pull & Bear moved. That is strong counterevidence to treating gross floor space as an immediate hard bottleneck. The unresolved question is whether the actual constraint was peak sorting, docks, labor, transport, system resilience, future growth, or something not visible in public data. A stage gate tests the constraint before converting a strategic option into fixed cost.
The balance sheet reinforces that caution. At July 31, 2001, inventory was EUR311.7 million, short-term investments and cash EUR177.6 million, long-term financial debt EUR152.0 million, short-term financial debt EUR196.0 million, and issuer-defined net financial debt EUR163.4 million in magnitude. These are unaudited point-in-time measures, not a project funding model. table.inditex.cutoff.h1-liquidity does not net or recompute them. source · inditex.cutoff.h1-2001
Alternatives and reference class
An immediate full Zaragoza commitment offers speed, redundancy, and a platform for growth, but risks building ahead of validated need. Staging Arteixo and existing category hubs preserves cash and uses reported headroom, but leaves concentration risk and may respond too late. Outsourced regional distribution lowers owned fixed assets but can fragment data, cadence, labor accountability, and peak control. Smaller distributed hubs reduce single-site exposure but can duplicate inventory and coordination.
H&M demonstrates that an outsourced production architecture can scale: its FY2000 report describes no owned factories, roughly 900 suppliers, lead times from three or four weeks to six months, 13 distribution centers, and 682 stores. It also reports price reductions EUR-equivalent comparisons should not be invented: its own measure was SEK600 million more than the prior year because of assortment imbalance and warm weather. H&M is a rival operating model, not a clean matched comparator—its year ended November 30, its currency was SEK, and its sales included VAT. source · inditex.cutoff.hm-2000
Labor, supplier, and continuity boundary
Nearshore capacity was economically useful and externally dependent. The prospectus said 96% of external sewing workshops were in Spain and Portugal and warned about regional capacity saturation, strikes, disasters, wages, labor conduct, and reputational harm. It explicitly said the newly adopted ethical code did not assure that suppliers would always act ethically. A capacity decision must therefore budget worker access, grievance, remedy, facility mapping, alternate capacity, and continuity. A code or audit count alone is not evidence of complete compliance. source · inditex.cutoff.ipo-2001
Decision gates
- Release capacity when stable-scope peak operations sustain practical utilization above the board threshold or when verified service failures cannot be remedied through shifts, process changes, or contracted overflow.
- Preserve twice-weekly ordering and replenishment, on-time-complete delivery, feedback latency, and new-product flow; square metres are not a substitute for service.
- Track inventory euros, units, aging, store weeks, warehouse dwell, issuer-defined days, stockouts, and markdown separately.
- Require a twelve-month stressed-liquidity runway after debt service, operating commitments, the store program, inventory funding, and non-cancellable project payments.
- Block supplier or logistics migration where material facilities lack traceability, worker access, grievance, remedy, labor protections, or continuity coverage.
Abstentions and missing evidence
The packet lacks the binding Zaragoza proposal, project schedule, incentives, throughput design, module economics, board paper, and rejected alternatives. It also lacks stable-scope peak Arteixo data, product-level markdown and stockout economics, contract-level outsourcing terms, worker-outcome records, and a monthly downside treasury model. No capitalization-complete valuation evidence is present, so this case abstains from a target price, security recommendation, and any claim that one logistics decision caused all later revenue, profit, return, or market value.