Mutual funds still hate battered software stocks: By the numbers

Software stocks remain deeply out of favor: mutual funds have the lowest exposure to the group since at least 2012, and hedge funds now hold software at its smallest weight since 2019 while semis are at record highs. The S&P Software & Services Index is down 12% year to date, with Salesforce, Adobe, and ServiceNow each falling 25% to 30% as investors price in AI disruption and slower enterprise spending. The article flags Salesforce earnings on Wednesday as a near-term test of whether AI fears and growth slowdown are justified.

Analysis

The positioning signal matters more than the price action: when both mutual funds and hedge funds are de-grossing software at the same time, marginal buyers disappear and multiple compression can continue well after fundamentals stabilize. That creates a reflexive setup where any weak guidance from the large-cap application layer forces ETF and benchmark sellers to liquidate the rest of the cohort, including higher-quality names that are not directly exposed to AI substitution risk.

The real second-order winner is still the picks-and-shovels AI supply chain, but the crowded nature of semis means leadership may narrow further toward companies with the cleanest capacity/throughput leverage and away from broad beta. LRCX and AMAT look better supported than the software group because capex cycles are being driven by foundry/HBM and advanced-node demand, not discretionary enterprise spend; ASML remains the highest-quality but is more exposed to any pause in EUV ordering if customers push out fabs.

Within software, the market is likely over-penalizing all recurring-revenue models as if AI adoption is binary. The next few quarters will probably separate workflow-software incumbents with embedded distribution and data advantages from point-solution vendors that can be more easily unbundled; that argues for dispersion trades rather than blanket sector exposure. The catalyst window is short: guidance season over the next 1-2 earnings cycles will decide whether this is a valuation reset or the start of a deeper 12-18 month derating.

The contrarian setup is that current sentiment may already reflect a recession-like outcome for software while enterprise IT budgets are merely being deferred, not destroyed. If Salesforce can frame AI as net retention accretive instead of cannibalistic, it could force a sharp short-covering rally across CRM/ADBE/NOW given how crowded the negative narrative is.

Key Decisions for Investors

• Run a pair trade: long LRCX/AMAT basket, short CRM or NOW into earnings over the next 2-6 weeks. Theses: semicap demand has clearer capex visibility, while software still faces valuation compression if guide-outs continue; target 10-15% relative upside if guidance confirms divergence.
• Buy short-dated downside protection on CRM ahead of earnings via put spreads. Risk/reward favors hedging because implied move is likely lower than the potential de-rating if management emphasizes AI-driven seat pressure or slower deal cycles.
• Add selective exposure to ASML only on weakness, not as a broad semis beta trade. Use it as a high-quality core position with a 3-6 month horizon, but size smaller than LRCX/AMAT because order timing risk can create near-term drawdowns.
• Fade indiscriminate software weakness with a relative-value basket long a profitable, platform-dominant name vs short a lower-moat application name. The trade should be implemented only after earnings confirm no broad demand collapse, with a 1-2 quarter catalyst window.
• Avoid catching the whole software sector outright until post-earnings revisions stabilize. The better expression is dispersion: long highest-quality names after capitulation, short names where AI can plausibly reduce spend per workflow.