Quick Summary
Covering the latest research from top Wall Street investment banks

Institutional Underweighting of Mega-Cap Tech Stocks Narrows; Preference Shifts to AI Hardware Bottleneck Areas

Institution
Morgan Stanley
Date
20260519
Company
NVIDIA, Apple, Microsoft, Amazon, Google-A, Meta Platforms, IBM, Intel, Intuit, Oracle, Broadcom, Cisco, Dell Technologies, Salesforce, KLA Corporation, Lam Research, Micron Technology, Western Digital, Now, Seagate Technology, SanDisk, Qualcomm, Palo Alto Networks, Applied Materials, AMD, Adobe
Ticker
NVDA, AAPL, MSFT, AMZN, GOOGL, META, IBMN, INTC, INTU, ORCLN, AVGO, CSCO, DELLN, CRMN, KLAC, LRCX, MU, WDC, NOWN, STX, SNDK, QCOM, PANW, AMAT, AMD, ADBE
Industry
Artificial Intelligence, Information Technology Services, Software - Infrastructure, Computer Hardware, IT Hardware
Rating
MixedMedium confidenceThe report adopts a neutral stance on changes in institutional tech stock holdings, but individual stock ratings diverge markedly: Meta, Intuit, and NVIDIA are viewed favorably as AI beneficiaries, while Adobe and Dell face downward pressure from cost or competitive challenges.
CoverageUnited States

AI summary card

Institutional Underweighting of Mega-Cap Tech Stocks Narrows; Preference Shifts to AI Hardware Bottleneck Areas

Q1 2026 data shows a modest improvement in institutional underweighting of mega-cap tech stocks (e.g., NVDA, AAPL), though still significantly below their S&P 500 weights; capital continues flowing into AI 'picks-and-shovels' stocks—particularly memory/storage—while software stocks face neglect.

—|Target Price —
Institutional HoldingsAI InfrastructureMega-Cap Tech StocksSemiconductor Memory/StorageSNDK OverweightNVDA Underweight
  • Institutional underweighting of mega-cap tech stocks (e.g., NVDA/AAPL/MSFT) narrowed from −137 bps to −125 bps.
  • NVDA remains the most severely underweighted large-cap tech stock (−2.39%), followed by AAPL and MSFT.
  • SanDisk (SNDK) is the most heavily overweighted tech stock (+2.16%), far exceeding all other names.
  • Institutional positioning reveals a pronounced structural preference: overweighting memory/storage (AI bottleneck areas), underweighting traditional software.
  • Quantitative analysis indicates that extreme underweighting by active managers often precedes technical price rebounds.

Report interpretation

Overview

Based on Q1 2026 13F filings, Morgan Stanley analyzes institutional ownership of major U.S. technology stocks. Key findings include: although the relative underweighting of mega-cap tech stocks (e.g., NVIDIA, Apple, Microsoft) versus their S&P 500 weights narrowed slightly during the quarter, they remain meaningfully underweighted overall. Meanwhile, institutional capital displays strong structural preferences—significantly overweighting memory and storage semiconductor stocks tied to AI infrastructure bottlenecks (e.g., SanDisk, Seagate), while sharply underweighting traditional software stocks. The report posits that such extreme positioning may signal potential technical corrections ahead.

Core views

Underweighting of mega-cap tech stocks improved marginally but remains at elevated absolute levels. As of end-Q1 2026, the average active institutional ownership of mega-cap tech stocks (including NVDA, AAPL, MSFT, AMZN, GOOGL, META, TSLA) was 125 basis points (bps) below their S&P 500 weights—narrowing from −137 bps at end-Q4 2025. This implies institutions modestly increased exposure to these giants over the quarter—or rising share prices boosted their passive benchmark weights, thereby narrowing the gap. Nevertheless, the −125 bps underweight signals continued caution among active fund managers toward these market leaders. At the individual stock level, NVIDIA (NVDA) remains the most severely underweighted large-cap tech stock, with institutional ownership 2.39% below its S&P 500 weight—a gap near historical highs and down only 18 bps sequentially. Apple (AAPL, −2.32%) and Microsoft (MSFT, −1.86%) follow closely, alongside Amazon (AMZN, −1.24%). In contrast, SanDisk (SNDK) is the most heavily overweighted stock (+2.16%), more than double the second-most overweighted name, Seagate Technology (STX, +0.84%). Salesforce (CRM) also ranks highly among overweighted names (+0.63%). Institutional positioning exhibits a clear 'AI hardware vs. software' bifurcation. At Q1 2026’s close, institutions held significantly higher weights in large memory and storage-related stocks (e.g., SNDK, STX, Lam Research, KLA, Western Digital) relative to their S&P 500 weights—reflecting strong preference for AI 'picks-and-shovels' and supply-chain bottleneck exposures. Conversely, traditional software names (e.g., IBM, Oracle, Palo Alto Networks, ServiceNow, Adobe) were markedly underweighted. This divergence suggests investors currently favor hardware segments of AI infrastructure where outcomes appear more certain, while remaining cautious about AI monetization prospects or competitive dynamics in the software layer. Extreme positioning carries predictive value. Morgan Stanley’s quantitative analysis finds that, after controlling for market cap and earnings surprise factors, significant deviations between active institutional ownership and S&P 500 weights—whether extreme under- or overweighting—are statistically correlated with future stock returns. Specifically, when active ownership falls far below benchmark weight, stocks tend to experience technical upside rebounds; conversely, extreme overweighting often precedes downside correction. Thus, NVDA’s current severe underweight may imply upside potential driven by position rebalancing, whereas SNDK’s extreme overweight warrants caution regarding pullback risk.

Analysis framework

This report employs a 'relative weight comparison methodology' to assess institutional positioning trends. Its core logic compares the average weight of a given stock across the top 100 actively managed portfolios against its weight in the S&P 500 index, calculating the spread (difference). 1. **Data Source**: SEC-mandated 13F filings covering the 28 largest U.S. tech companies by market cap. 2. **Metric Definitions**: - **Underweight**: Institutional weight < S&P 500 weight → negative spread. - **Overweight**: Institutional weight > S&P 500 weight → positive spread. 3. **Analytical Dimensions**: - **Aggregate View**: Track changes in average under-/overweighting across mega-cap tech stocks to gauge macro-level capital flows. - **Structural View**: Drill down to individual stocks to identify which subsectors (e.g., memory vs. software) attract or repel capital. - **Historical Context**: Compare current spreads against historical distributions (e.g., NVDA’s underweight near historical peaks) to assess extremity. 4. **Inference Logic**: Historical regression analysis demonstrates that extreme relative positioning deviations often serve as leading indicators of subsequent price reversals (mean-reversion logic), offering actionable trading signals.

Methodology notes

  • Event-Driven Strategy & Behavioral FinanceFund Flow / Positioning Analysis

    Active Ownership Spread

    Measures fund crowding by comparing active fund portfolio weights to a market benchmark (e.g., S&P 500). Extreme underweighting typically signals exhaustion of selling pressure and accumulation of latent buying power—potentially triggering technical rebounds; extreme overweighting reflects concentrated long positions vulnerable to profit-taking on adverse news. This embodies a quantified form of behavioral 'contrarian investing'.

Asset mapping & comparison

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

  • NVIDIA (NVDA.O)
    Most severely underweighted stock; core AI compute enabler
    Strengths
    Blackwell architecture maintains performance leadership; data center demand continues to outstrip supply
    Weaknesses
    Valuation already prices in high growth; multiple expansion potential is limited
    Comparison
    Underweighting far exceeds that of other mega-caps, implying high rebound elasticity
    Risks
    Supply rapidly catching up to demand causing growth deceleration, or emergence of strong competitors
  • SanDisk (SNDK.O)
    Most severely overweighted stock; beneficiary of memory cycle recovery
    Strengths
    Strong AI-driven data center storage demand; tight industry supply-demand balance
    Weaknesses
    Excessive positioning concentration; fragile trade structure
    Comparison
    Overweighting is more than double that of second-place STX
    Risks
    Reversal in memory pricing, or AI demand falling short of expectations
  • Apple (AAPL.O)
    Severely underweighted; awaiting acceleration in iPhone replacement cycle
    Strengths
    Stable services business; AI features expected to drive hardware upgrades
    Weaknesses
    Near-term uncertainty from rising memory costs
    Comparison
    Second-most underweighted after NVDA
    Risks
    Weak consumer spending; iPhone 17 sales missing expectations
  • Adobe (ADBE.O)
    Representative underweighted software stock; facing AI disruption concerns
    Strengths
    Deep moat in complex marketing solutions
    Weaknesses
    Low visibility into generative AI’s potential erosion of Creative Cloud business
    Comparison
    Part of the broader software segment avoided by institutions
    Risks
    Intensifying AI competition; macroeconomic headwinds affecting SMB spending

Key data

  • Average Underweighting of Mega-Cap Tech Stocks-125 bpsNarrowed by 12 bps from −137 bps in Q4 2025
  • Average Overweighting of Other Large-Cap Tech Stocks+32 bpsFlat quarter-on-quarter
  • NVDA Institutional Underweighting-2.39%Most severe underweight among large-cap tech stocks, near historical peak
  • SNDK Institutional Overweighting+2.16%Most severe overweight among large-cap tech stocks, more than double the second-highest

Impact & implications

For mega-cap tech stocks—especially NVDA, AAPL, and MSFT—their current extreme underweight status may constrain further downside and provide a technical foundation for a rebound. If fundamentals remain robust, short-covering or position normalization could drive upward price momentum. For the memory and storage sector (SNDK, STX, MU, WDC, etc.), broad institutional overweighting reflects strong consensus on AI-driven storage demand. However, high positioning concentration means even minor negative catalysts could trigger sharp profit-taking. Investors should closely monitor storage pricing sustainability and the pace of AI-related capital expenditures. For the software sector, institutional underweighting reflects concerns about AI monetization speed and competitive intensity. Should upcoming earnings reports demonstrate that AI features translate meaningfully into revenue growth and margin expansion, this sector may offer substantial expectation-revision upside.

Risks

  • Macroeconomic recession reducing enterprise IT spending and advertising budgets
  • Slowing growth or signs of bubble bursting in AI infrastructure investment
  • Geopolitical tensions disrupting global supply chains—especially semiconductors
  • Regulatory shifts (e.g., antitrust, data privacy) impacting tech giants’ business models

What to watch

  • Trend in mega-cap tech underweighting in next quarter’s 13F filings
  • Capital expenditure guidance and AI server order visibility from major cloud providers (hyperscalers)
  • Pricing trajectory and inventory levels for memory chips (DRAM/NAND)
  • Monetization progress and customer adoption rates for AI features at software firms
Zhejiang ICP No. 2022035445-5
Disclaimer: Market data, charts, indicators, research views, and other information provided on this website are intended solely for information display, research communication, and educational reference. They should not be regarded as personalized investment advice, securities recommendations, trading instructions, solicitations, or guarantees of return. While we strive to improve the reliability of our data and content, such information may still be subject to delays, errors, incompleteness, or untimely updates due to source differences, methodological limitations, system processing, or market volatility. Users should exercise independent judgment based on their own circumstances and bear all risks and responsibilities arising from the use of this website.

Settings

Sign in to view recent logins