Internet Sector Corrects 5%, Giant Valuations Fall Below Historical Averages

Overview

This report is Morgan Stanley's regular tracking update on the North American internet sector. The report points out that amidst the ongoing debate over capital flows between AI hardware and AI software, the internet sector has experienced a significant correction, but valuations of core leading companies have retreated to attractive ranges. Through detailed comparable company analysis tables (Comp Sheets), the report displays valuation levels across sub-sectors, stock performance, and the potential impact of share-based compensation (SBC) on valuations, reiterating its positive outlook on the industry.

Core views

Market Performance and Capital Flows: The internet sector fell approximately 5% over the past week, underperforming the S&P 500 (-3%) and the Nasdaq (-5%). The decline was primarily driven by large-cap stocks, with Amazon (AMZN) down 9%, Meta (META) down 6%, and Alphabet (GOOGL) down 3%. In contrast, Pinterest (PINS) rose 7% against the trend, AppLovin (APP) fell 9%, and Reddit (RDDT) dipped slightly by 1%. This divergence reflects the market's continued tug-of-war and rotation in capital allocation between AI hardware infrastructure and AI software applications. Valuation Levels and Historical Comparison: Despite short-term pressure on stock prices, this provides better valuation entry points for long-term investors. Report data shows that Amazon, Alphabet, and Meta's 2026 expected P/E ratios are 26x, 25x, and 18x respectively, trading at discounts of 19%, 2%, and 27% to their trailing twelve-month (TTM) average P/E ratios. From the NTM EV/EBITDA metric perspective, as of June 5, 2026, Amazon stood at 11.8x (a 9% discount to the 2-year average), Meta at 9.7x (a 23% discount to the 2-year average), with only Alphabet trading at a premium. This indicates that, aside from individual names, the sector's overall valuation has already priced in some pessimistic expectations. Hidden Impact of Share-Based Compensation (SBC) on Valuations: The report specifically emphasizes the impact of SBC accounting treatment on valuation comparability. If SBC is adjusted to be treated as a cash expense, the EV/EBITDA multiples for the Travel/Sharing Economy and PropTech sectors would increase by an average of approximately 44%. At the specific company level, Lyft's adjusted multiple increased by up to 103%, Pinterest by 207%, and Duolingo by 110%. This analysis reminds investors that for growth-oriented internet companies with high SBC proportions, traditional GAAP valuations may underestimate actual operating costs, requiring comprehensive judgment based on adjusted metrics.

Analysis framework

The report adopts a typical 'top-down sector view + bottom-up comparable company analysis' framework. It first establishes the baseline judgment of the sector as 'Attractive', then validates this view through multi-dimensional quantitative tables. In valuation analysis, the institution not only used standard EV/EBITDA and P/E multiples but also introduced a 'historical mean reversion' perspective, comparing current valuations with 2-year, 3-year, 5-year, and 10-year long-term averages to identify oversold or overvalued targets. Meanwhile, addressing the prevalent issue of high share-based compensation in the internet industry, the institution specifically constructed a 'SBC as Cash' sensitivity analysis. By quantifying the impact of SBC on the denominator EBITDA, it restores a more realistic picture of profitability and valuation levels, avoiding misjudgments caused by differences in accounting standards.

Methodology notes

• Valuation MethodologyEV/EBITDA valuation

  EV/EBITDA multiples and their historical mean reversion analysis

  The report uses Enterprise Value multiples (EV/EBITDA) as the core valuation anchor, comparing current values with historical averages over the past 2 years, 3 years, and even 10 years. When current multiples are significantly lower than historical averages (e.g., Meta at a 23% discount), it is typically viewed as having a margin of safety or being an oversold signal; conversely, it may imply excessive premiums. This method assumes that valuations fluctuate around a central tendency in the long run.

• Company Fundamentals and Financial FrameworkEarnings Quality Analysis

  Impact of adding back Share-Based Compensation (SBC) adjustments on valuation

  Internet technology companies often add back large amounts of share-based compensation (SBC) as non-cash expenses to calculate adjusted EBITDA. The report reveals the gap between GAAP profits and true economic profits by calculating the change in EV/EBITDA if SBC were treated as a cash expense (e.g., multiples rising by 44%-207% for some companies). This reminds investors to be wary of the actual equity dilution costs behind seemingly low book valuations when comparing companies with different SBC intensities.

Asset mapping & comparison

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

• Amazon (AMZN)

  Sector heavyweight, valuation corrected below historical average

  Strengths

  2026E P/E 26x, 19% discount to TTM average; NTM EV/EBITDA 11.8x below 2-year average

  Weaknesses

  Weekly decline of 9%, leading sector drop; significant short-term capital outflow

  Comparison

  Valuation discount larger than Alphabet but smaller than Meta

• Meta Platforms (META)

  Representative of AI software/applications, most significant valuation compression

  Strengths

  2026E P/E only 18x, substantial 27% discount to TTM average; NTM EV/EBITDA 23% discount to 2-year average

  Weaknesses

  Weekly decline of 6%; post-SBC adjustment valuation rises by 19%

  Comparison

  Deepest valuation discount among the three giants, relatively higher margin of safety

• Pinterest (PINS)

  Counter-trend riser, demonstrating relative strength

  Strengths

  Weekly gain of +7%, leading the sector

  Weaknesses

  Post-SBC adjustment EV/EBITDA surged 207%, implying high hidden costs

  Comparison

  Performance significantly stronger than heavyweights like AMZN/META

  Risks

  High SBC proportion may lead to overestimation of actual earnings

• Lyft (LYFT)

  Case study with the largest impact from SBC adjustment

  Weaknesses

  Post-SBC adjustment EV/EBITDA rose 103%; short interest ratio as high as 24.3%

  Comparison

  SBC adjustment magnitude far exceeds Uber (+20%)

  Risks

  Combination of high short positions and high hidden compensation costs

Key data

• Internet Sector Weekly Change-5%S&P 500 fell 3%, Nasdaq fell 5% during the same period
• AMZN/META/GOOGL 2026E P/E26x / 18x / 25xTrading at discounts of 19%, 27%, and 2% to TTM averages respectively
• Median Increase in EV/EBITDA after SBC Adjustment+44%For Travel/Sharing Economy and PropTech sectors
• Pinterest (PINS) Weekly Gain+7%One of the few stocks rising against the trend in the sector
• Meta NTM EV/EBITDA vs. 2-Year Average9.7x vs 12.6xDiscount magnitude reaches 23%

Impact & implications

The report believes that the current stock price correction has not changed the fundamental logic of the internet sector; instead, it has made the valuations of core assets more attractive. For investors focused on AI themes, the capital rotation between hardware and software is a normal market game process, and long-term value should not be ignored due to short-term volatility. Especially for companies with high SBC proportions but still reasonable adjusted valuations, current pricing may not fully reflect their true profitability. Maintaining an 'Attractive' rating means the institution suggests investors keep watch at current levels or accumulate on dips, rather than panic selling.