US semiconductors and semiconductor equipment: UBS argues semicap valuations do not fully discount a potentially much larger WFE cycle

Overview

UBS's semiconductor update covers Qualcomm's edge-AI positioning, the valuation debate around semiconductor-equipment stocks, and the durability of Micron-related memory long-term agreements. Its central conclusion is that elevated WFE expectations are not necessarily fully reflected in semicap valuations and that sustained memory profitability could support continued equipment spending.

Core views

At Qualcomm's Snapdragon Summit, UBS saw a reinforced rather than transformed investment narrative. Qualcomm positioned the smartphone as an intelligence layer connecting PCs, wearables, XR devices, audio products and automobiles, supported by the Snapdragon 8 Elite Gen 6 and Snapdragon 8 Elite Extreme Gen 6 launches for upcoming HONOR, Motorola and Xiaomi flagships. The company also highlighted larger on-device AI models, software support and partners including Google, Microsoft, Xiaomi, Lenovo, HP, Motorola and Mastercard. UBS came away constructive on Qualcomm's long-term edge-AI role across Android and Windows, but did not regard the event as a major debate changer because investors already broadly recognize these themes. The report's main analytical focus is the argument that semiconductor-equipment expectations may be less fully priced than investors assume. UBS says investor discussions increasingly center on whether roughly US$300bn of WFE in calendar 2028 is already embedded in AMAT, LRCX and KLAC valuations. UBS models approximately US$280bn, but disputes that consensus expectations are already above US$300bn. Its EPS-power work indicates that, under a US$300bn annual WFE scenario, the stocks trade at approximately 15x for AMAT, 17x for LRCX and 19x for KLAC. Under a US$400bn scenario, the respective multiples decline to about 11x, 13x and 14x. UBS therefore argues that even US$300bn WFE could warrant somewhat higher earnings multiples than the stocks currently command. UBS supports this view with supply-chain checks that have strengthened its conviction that high WFE outcomes are achievable. Some major semiconductor-production-equipment suppliers have told their supply chains to prepare for US$500bn of WFE over the next three to five years. Its C2029 upside scenarios, excluding potential average-selling-price uplift, show EPS power rising from US$18.06 to US$29.14 for LRCX, US$31.50 to US$56.96 for AMAT, and US$9.65 to US$16.40 for KLAC as assumed WFE rises from US$300bn to US$500bn. UBS flags Terafab as an important uncertainty: spending approaching TSMC levels could help equipment demand near term, while a more efficient wafer-flow model could be unfavorable for equipment suppliers over the longer term. On memory, UBS addresses investor concern that LTAs supporting memory-company gross margins of roughly 70% to 75% may fail if customers view about 60% as normalized and seek spot-market-like pricing as supply normalizes. UBS argues that DDR pricing would have to fall approximately 75% to 80% from peak levels before customers would consider breaking LTAs. Even then, customers could forfeit more than US$10bn in prepayments, which UBS views as an effective break fee. UBS does not model price per GB returning to roughly US$5 until November 2029 or February 2030, late in the deal term. The report also contrasts Micron's approximately 6.7x next-twelve-month P/E, which UBS says implies limited market confidence in current memory earnings durability, with LRCX, AMAT and KLAC at roughly 27x to 32x, which it interprets as reflecting expectations for continued memory investment through 2028 and beyond. UBS argues that this valuation disconnect helps explain why semicap stocks can still work: memory suppliers could remain highly profitable and sustain elevated capex and WFE investment even if gross margins decline by about 20 percentage points year over year.

Analysis framework

UBS combines event interpretation, supply-chain checks, valuation comparisons and scenario-based EPS-power analysis. It tests equipment-company earnings sensitivity at US$300bn, US$400bn and US$500bn WFE assumptions, then compares those earnings outcomes with current trading multiples. For memory, it evaluates LTA durability through price-decline thresholds, customer prepayments and the timing of its price-per-GB assumptions.

Methodology notes

• Valuation methodsP/E and PEG Valuation

  P/E valuation and EPS-power scenarios

  UBS compares current trading multiples with estimated earnings power for AMAT, LRCX and KLAC at different WFE levels to assess how much equipment-demand upside may be reflected in valuations.

• Industry AnalysisSupply-demand framework

  WFE demand scenarios and equipment-supply-chain checks

  The report uses assumed WFE spending levels and supplier supply-chain preparations to assess the potential scale and durability of semiconductor-equipment demand.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Qualcomm (QCOM)

  UBS is constructive on Qualcomm's long-term positioning in edge AI across Android and Windows ecosystems.

  Strengths

  Smartphone-centered distributed AI strategy, new Snapdragon platforms, on-device AI support and broad ecosystem partnerships.

  Weaknesses

  The Snapdragon Summit largely reinforced themes already appreciated by investors rather than changing the debate.

• Applied Materials (AMAT)

  Semiconductor-equipment company with earnings leverage to higher WFE spending.

  Strengths

  UBS sees valuation support under US$300bn to US$500bn WFE scenarios.

  Weaknesses

  Recent performance has been weighed down by the perception that high WFE expectations are already priced in.

  Comparison

  UBS estimates roughly 15x P/E at US$300bn WFE and roughly 11x at US$400bn, versus LRCX and KLAC scenario valuations.

  Risks

  Longer-term equipment demand could be affected if Terafab materially improves fab wafer-flow efficiency.

• Lam Research (LRCX)

  Semiconductor-equipment company with earnings leverage to higher WFE spending.

  Strengths

  UBS's C2029 EPS power increases from US$18.06 at US$300bn WFE to US$29.14 at US$500bn.

  Weaknesses

  Recent performance has been weighed down by concerns over elevated WFE expectations.

  Comparison

  UBS estimates roughly 17x P/E at US$300bn WFE and roughly 13x at US$400bn.

  Risks

  Longer-term equipment demand could be affected by more efficient fab manufacturing models.

• KLA (KLAC)

  Semiconductor-equipment company with earnings leverage to higher WFE spending.

  Strengths

  UBS's C2029 EPS power increases from US$9.65 at US$300bn WFE to US$16.40 at US$500bn.

  Weaknesses

  Recent performance has been weighed down by concerns over elevated WFE expectations.

  Comparison

  UBS estimates roughly 19x P/E at US$300bn WFE and roughly 14x at US$400bn.

  Risks

  Longer-term equipment demand could be affected by more efficient fab manufacturing models.

• Micron (MU)

  Memory supplier whose LTA durability and profitability are linked to continued memory capex and WFE investment.

  Strengths

  UBS believes customer prepayments and the required scale of DDR price declines make LTA breaks unlikely during most of the deal term.

  Weaknesses

  Investors remain skeptical about the durability of memory LTAs and current elevated gross margins.

  Comparison

  MU trades at approximately 6.7x NTM P/E, versus roughly 27x-32x for LRCX, AMAT and KLAC.

  Risks

  A normalization of supply and movement toward spot-market pricing could pressure gross margins and challenge the LTA thesis.

Key data

• UBS C2028 WFE estimate~US$280bnUBS estimate versus the commonly discussed ~US$300bn investor expectation
• Semicap multiples at US$300bn WFE~15x AMAT, ~17x LRCX, ~19x KLACUBS assessment of trading multiples under a US$300bn-per-year WFE scenario
• Semicap multiples at US$400bn WFE~11x AMAT, ~13x LRCX, ~14x KLACUBS assessment of trading multiples under a US$400bn WFE scenario
• C2029 EPS power at US$300bn to US$500bn WFELRCX US$18.06-US$29.14; AMAT US$31.50-US$56.96; KLAC US$9.65-US$16.40Upside cases excluding potential ASP uplift
• Memory-company gross margins supported by LTAs~70%-75%Consistent with UBS's interpretation of Micron's indication of margins well above the prior peak
• DDR price decline needed before LTA breaks are considered~75%-80% from peak levelsUBS estimate; customers could also forfeit potentially more than US$10bn in prepayments
• Micron valuation~6.7x NTM P/EUBS interprets this as limited confidence in the durability of current memory earnings

Impact & implications

UBS argues that WFE upside remains a meaningful potential support for AMAT, LRCX and KLAC because their valuations do not appear to fully reflect higher spending scenarios. It also sees durable memory LTAs and sustained memory profitability as important support for elevated semiconductor-equipment investment, while Terafab's manufacturing approach remains a longer-term uncertainty.

Risks

• A macroeconomic downturn or disruption to international trade could weaken semiconductor demand.
• Technological disruption or business-model innovation could alter industry unit sales, ASPs and revenue trajectories.
• Terafab's attempt to redesign wafer flow could reduce longer-term demand for semiconductor-production equipment.
• Different assumptions on WFE, pricing, margins or memory supply conditions could materially change the analysis.