Software and semiconductor inverse correlation spikes, with European software services stocks whipsawed by AI theme rotation

Overview

This report is Morgan Stanley's weekly update on the European Software & Services industry, focusing on the pronounced inverse correlation between software services and semiconductors in AI-themed trading, the fundamental support provided to software by the 2Q earnings season, hyperscalers' AI investment returns, competition in IT services reskilling, and the impact of Legal Tech M&A on traditional legal information and software companies.

Core views

The core view is that the market is treating semiconductors and AI infrastructure as the direct beneficiaries of the AI theme, while Software & Services is still viewed by some investors as a sector that could be structurally disrupted by AI models over the long term. Therefore, when AI hardware and semiconductors rise, Software & Services may become the price segment bearing the inverse-trading pressure. However, 2Q results showed resilient growth at several software companies. If more companies can demonstrate AI revenue contribution and resistance to substitution, share-price performance among application software companies may increasingly diverge during the year.

Analysis framework

The report combines European sector price correlations, company 2Q earnings updates, an AI infrastructure ROIC framework, global hiring and developer data, and Legal Tech transaction cases to assess relative opportunities and risks for software, semiconductor, IT services, and legal information services companies under the AI theme.

Methodology notes

• Theme Correlation Analysis10-Week Rolling Correlation

  Inverse trading of software and services versus semiconductors

  European constituent data are used to observe short-term changes in the correlation between semiconductors and software services, identifying the inverse pull of AI infrastructure sentiment on software services share prices.

• Investment Return AnalysisGenAI ROIC Framework

  AI infrastructure investment returns

  The report cites unit economics and incremental ROIC analyses for three hyperscaler business models: GPU leasing, model APIs, and third-party infrastructure model APIs, emphasizing the importance of data-center capacity and model innovation.

• Labor Supply and Demand AnalysisAI Reskilling Competition

  IT services talent bottleneck

  Based on approximately five million job postings and one million developer profiles, the global IT services team quantifies AI-related talent demand, supply scarcity, and divergence in job structures.

Asset mapping & comparison

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

• European Software & Services sector

  Potential disruption target and inverse-trading asset under the AI theme

  Strengths

  Overall 2Q results were positive, with growth at SAP, ServiceNow, RELX, and Sage strengthening fundamental confidence.

  Weaknesses

  The market remains concerned that advances in AI models could weaken the value of traditional software, leaving sector prices vulnerable to semiconductor and AI hardware trends.

  Comparison

  Compared with semiconductors, software services currently appear more like the price segment bearing AI-theme pressure, while semiconductors appear more like the price-setting segment.

  Risks

  If AI revenue visibility remains insufficient or customer substitution risk rises, sector valuations and sentiment could remain under pressure.

• Semiconductors and Semicap

  Direct beneficiaries of AI infrastructure build-out

  Strengths

  Benefit from AI infrastructure capital expenditure, GPU demand, and the rising value of data-center capacity.

  Weaknesses

  Price performance is highly dependent on AI infrastructure sentiment and capital-expenditure expectations.

  Comparison

  They benefit more directly than software services from AI infrastructure expansion and have recently exerted an inverse pull on software in trading.

  Risks

  If returns on AI infrastructure investment fall below expectations, hardware-chain valuations could retreat.

• AMZN, GOOGL, MSFT, META

  Hyperscalers and AI infrastructure investors

  Strengths

  Their scale, data-center capacity, and model ecosystems make them more likely to capture GenAI ROIC and unit-economics advantages.

  Weaknesses

  They require continued substantial investment, with returns dependent on model innovation, token pricing, and efficiency gains.

  Comparison

  They benefit more than smaller participants from scale advantages and infrastructure scarcity.

  Risks

  Excessive capital expenditure, insufficient utilization, or declining prices could compress ROIC.

• Global IT Services companies

  Service providers for AI implementation and reskilling demand

  Strengths

  AI talent scarcity increases the value of credible implementation services, and early transformers may replicate the share gains seen during the cloud-transition cycle.

  Weaknesses

  The transition involves not only changes in revenue streams but also a restructuring of delivery models.

  Comparison

  Compared with traditional P/E, metrics based on FCF after stock-based compensation and M&A may better reflect underlying value.

  Risks

  Rising talent acquisition valuations, M&A becoming a recurring reinvestment cost, and slower-than-expected transformation.

• RELX and Wolters Kluwer legal businesses

  Established content and software platforms in the Legal Tech competition

  Strengths

  RELX's legal-business revenue growth accelerated to approximately 10% underlying year over year, showing continued potential to participate in upside.

  Weaknesses

  Competition in the legal workflow tools market is intense, with Harvey and Legora making progress among corporate legal departments and large law-firm clients.

  Comparison

  RELX is more content-centric, while Wolters Kluwer covers both legal software and content; the legal research market has so far changed relatively little.

  Risks

  Intensifying competition in Legal Workflow Tools could affect growth and pricing; Thomson Reuters' results will require monitoring.

Key data

• 10-week rolling correlation between European semiconductors and European software servicesApproximately -0.7xIndicates a strong negative correlation, significantly below the roughly 10-year average of +0.55x.
• Intraw​​eek rebound in Software & ServicesSome covered companies rose more than 15% through Wednesday's closeThey subsequently retraced as AI infrastructure and semiconductor stocks rallied.
• GenAI incremental ROICApproximately 25% or moreDerived from analysis of hyperscaler AI infrastructure-related business models.
• Demand for AI build-out and scaled deployment rolesUp 107% year over year in May 2026Based on a demand proxy from Morgan Stanley's global IT services team.
• Demand for junior or automatable rolesDown 5% year over yearReflects a K-shaped divergence in the technology talent market.
• Revenue forecast for approximately 20 IT services companies covered by Morgan StanleyApproximately $300 billion in 2028Corresponds to approximately 4% CAGR from 2026 to 2028, with a scenario range of approximately -3% to +7% CAGR.
• Sage nine-month organic total revenue growthApproximately 10.4%Implies approximately 11% in 3Q; the share price rose approximately 9% that day.
• Capgemini FY26 growth guidance8.5%-9%Raised from the previous 6.5%-8.5%, but the share price fell approximately 5% that day.
• Nemetschek share-price reactionDown approximately 14% that dayPressure resulted from lower media-business guidance and the impact of consolidating HCSS, despite solid 2Q growth.
• Legora cash from its April 2026 Series D financingMore than $600 millionTo fund Legal Tech M&A expansion, including the acquisition of Wexler.

Impact & implications

For investors, the Software & Services sector's relative performance may remain dominated in the short term by semiconductor and AI infrastructure sentiment. Over the medium term, attention should focus on whether software companies can demonstrate that customers are reluctant to replace third-party packaged software, that AI revenue can spread across more companies, and that competitive and pricing pressures remain manageable. IT services companies need to adapt to the AI transition through reskilling, M&A, and new delivery models, while the cost of M&A and the quality of free cash flow after stock-based compensation will become key valuation considerations.

Risks

• If AI infrastructure and semiconductor stocks remain strong, the Software & Services sector could continue to face inverse-trading pressure.
• If software companies cannot demonstrate clear AI revenue or resistance to substitution, investor concerns about AI disruption could resurface.
• IT services companies may face talent scarcity, excessive M&A valuations, and declining free-cash-flow quality during the AI transition.
• Expansion by new Legal Tech entrants in the legal workflow tools market could weaken the competitive position of traditional legal information and software companies.
• External factors such as conflicts in the Middle East could continue to affect the earnings-recovery trajectory of companies related to air travel, including Amadeus.

What to watch

• Whether the 10-week rolling correlation between European software and semiconductors recovers from approximately -0.7x toward a normal range.
• Whether subsequent 2Q and 3Q results show software companies disclosing AI revenue contribution and evidence of accelerating growth more broadly.
• AI infrastructure capital expenditure, data-center capacity utilization, and validation of hyperscalers' GenAI ROIC.
• Whether IT services companies can transform their revenue mix through reskilling and new delivery models rather than relying excessively on high-priced acquisitions.
• Thomson Reuters' results next week and the signals they provide about competition in legal research and legal workflow.