ASML customer co-investment decision

Cutoff-isolated packet. This report includes only evidence public by 2012-07-09T23:59:59Z; later participant decisions and outcomes are excluded from the learner bundle. Because this named case was assembled retrospectively and later artifacts were pre-staged before freeze, the source boundary does not prove cognitive blinding by the researcher.

Executive Summary

• Proceed only toward staged, ring-fenced approval. ASML announced conditional maximum customer funding and equity terms for separate EUV and 450mm projects. Those workstreams should have separate technical, demand, economics, and termination gates. source · asml.2012.ccip-announcement
• The opportunity was real but production readiness was not proven. ASML reported six operational NXE:3100 systems, about 9,000 exposed wafers, and eleven NXE:3300B orders; contemporaneous reporting placed demonstrated source power at 30 W versus 100 W for a minimum 60 wafers per hour. source · asml.2012q1.results source · optics.2012.euv-delay
• Abstain from a target price or unconditional implementation. Announced maximum terms are conditional and prospective; final subscriptions, complete rights, workstream cash flows, capitalization, and a verified market price are missing.

Decision frame

At 2012-07-10T00:00:00Z, after the July 9 public announcement, ASML must decide whether to proceed toward approval and implementation of customer equity and R&D co-investment for EUV and 450mm. The alternatives are a combined commitment, a staged and ring-fenced program, or continuation of the existing self-funded roadmap. The objective is not simply more funding: it is faster learning without surrendering roadmap independence or hiding commercial concessions.

The announced terms are prospective, not realized economics

ASML described two funding projects—450mm and next-generation EUV—and said that, if the maximum aggregate 25% available shares were fully subscribed, customers would acquire shares valued at EUR 4.19 billion and would commit EUR 1.38 billion of R&D funding for 2013–2017. table.asml.cutoff.announced-euro-terms and table.asml.cutoff.announced-share-limit preserve that conditional basis. They do not represent completed issuance, realized funding or proceeds, or workstream spend. source · asml.2012.ccip-announcement

Intel described a EUR 3.3 billion set of agreements and an objective of shortening deployment schedules by as much as two years. That is a participant's forward-looking objective, not a measured schedule benefit. source · intel.2012.asml-agreements

The outside view is incomplete

The cutoff record contains an actual filed annual report, regulated results, an independent technical publication, and issuer and participant program announcements, but no unbiased base rate for customer-funded frontier manufacturing platforms. The 2011 Form 20-F supplies the audited backdrop; the decision must still be conditioned on this case's measurable milestones rather than a claimed generic acceleration benefit. source · asml.2011.20f

Orders validate learning demand, not production economics

ASML reported that all six NXE:3100 systems were operational and printing wafers, approximately 9,000 wafers had been exposed, and eleven NXE:3300B systems were on order. These are useful adoption and learning signals in table.asml.cutoff.euv-learning. They do not establish production availability, cost per wafer, yield, or project return. source · asml.2012q1.results

The strongest disconfirming evidence was the source-power gap. optics.org reported 30 W in the production environment versus a desired 100 W for 60 wafers per hour. That tension remains explicit in conflict.asml.cutoff.euv-readiness; orders cannot silently resolve it. source · optics.2012.euv-delay

A profitable base does not establish project returns

table.asml.cutoff.financial-base records Q1 2012 net sales of EUR 1.252 billion and net income of EUR 282 million. table.asml.cutoff.liquidity records EUR 2.953 billion of cash and cash equivalents. These are reported values, not annualized estimates. They demonstrate consolidated capacity to negotiate, not the economics of either proposed workstream. source · asml.2012q1.results

ASML guided to EUR 145 million of R&D cost for Q2, without a public allocation to EUV, 450mm, or other programs. table.asml.cutoff.program-gaps therefore leaves final co-funded R&D and verified project return as unknown—not zero. source · asml.2012q1.results

Governance is part of the economics

ASML announced proposed foundation-held, ordinarily non-voting, locked-up shares to protect strategic and operational independence. The public record still contains no final aggregate subscription, complete voting and exclusivity rights, IP terms, pricing concessions, milestone remedies, or abandonment provisions. A program that appears to lower R&D cost but transfers roadmap control or embeds discounts could destroy value. Any final agreement should preserve product-roadmap independence, equal product availability, ASML-owned shared IP, and workstream-specific pause or reallocation rights. source · asml.2012.ccip-announcement

Alternatives and scenarios

An unconditional combined commitment exposes ASML to correlated technical and demand risk. Continuing self-funding preserves independence but may forgo risk sharing and customer validation. The staged alternative offers the best bounded option on this record because it buys information while preserving the option to expand, pause, or reallocate.

The packet assigns 25% to both workstreams clearing their gates, 50% to exactly one clearing, and 25% to neither clearing. This is an explicitly uncalibrated illustration: it assumes independent 50% workstream gate-clearance chances solely to map decisions. It is not an empirical base rate, forecast, calibrated prior, or deterministic valuation output. [assumption.asml.cutoff.illustrative-independent-workstreams]

Recommendation and monitoring

Select alternative.asml.staged-customer-coinvestment with moderate confidence. Before each tranche, verify:

1. EUV source power, availability, throughput, overlay, and customer acceptance in at least two customer environments.
2. Binding customer funding and tool demand net of discounts, credits, and remedies.
3. Non-voting, non-exclusive governance that preserves roadmap and IP independence.
4. Workstream-level net funding relative to incremental R&D, capital, support, and termination cost.

Pause EUV after two consecutive missed technical reviews; pause 450mm if binding fab transition demand fails the gate; and re-underwrite either project if net customer funding no longer covers the approved incremental risk budget.

Further questions

1. What is the net customer funding after every discount, credit, warranty, remedy, exclusivity promise, and other economic concession is valued?
2. Which independent protocol will certify EUV source power, throughput, availability, overlay, and customer acceptance before each tranche is released?
3. What binding customer fab-transition commitments establish sufficient 450mm demand, and what date or volume shortfall triggers a pause?

Caveats and source boundary

No target price is supported. The full Part A source set is src.asml.2011.20f, src.asml.2012q1.results, src.optics.2012.euv-delay, src.asml.2012.ccip-announcement, and src.intel.2012.asml-agreements. The annual report is retained as immutable primary evidence, while exact structured excerpts use deterministic HTML replay. The July 9 sources establish the public decision premise and prospective terms; they do not supply complete final terms, realized program economics, or a causal schedule estimate. This retrospective named-case workflow and pre-staging of later source artifacts also mean that cutoff-isolated files should not be described as proof of researcher cognitive blinding.