Tech stocks' valuations return to pre-AI boom levels! Apollo's Chief Economist: Forward P/E ratio compressed from 40 times to 20 times

Author: Claude, Deep Tide TechFlow

Deep Tide Guide: Apollo Global Management Chief Economist Torsten Slok’s latest chart shows that the forward P/E ratio of the S&P 500 Information Technology sector has compressed from about 40 times at the peak of the AI boom to around 20 times, returning to pre-AI prosperity levels. The sector rotation triggered by Middle East conflicts, doubts about AI capital expenditure returns, and slowing profit growth create a complex pressure, and tech giants are experiencing their most painful valuation re-pricing since 2022.

The valuation bubble of the S&P 500 Information Technology sector is being rapidly squeezed.

According to Apollo Global Management’s April 11 Daily Spark report, partner and chief economist Torsten Slok revealed the current situation of tech stocks with a chart: the forward P/E ratio of the S&P 500 Information Technology sector has shrunk from the high of about 40 times during the AI boom to approximately 20 times, back to the valuation levels before the AI boom.

This chart covers the ten largest constituent stocks in the index: Nvidia, Apple, Microsoft, Broadcom, Oracle, Micron, Palantir, AMD, Cisco, and Applied Materials. In other words, the core winners of the AI era have collectively given up the valuation premiums accumulated over the past two years.

Triple pressures intertwine, leading to valuation re-pricing for tech giants

The valuation compression is driven by multiple factors, not a single cause.

The Middle East conflict is the most direct catalyst. Since the outbreak of war in Iran, the energy sector surged over 34% in the first quarter, and ExxonMobil’s year-to-date gains approach 42%. Large capital flows have shifted from tech stocks to energy and defensive sectors, making tech stocks the largest outflow. The S&P 500 formed a “death cross” at the end of March (50-day moving average crossing below the 200-day), and by early April, the index struggled around 6,582 points, just below the 6,300 point “correction zone” threshold.

Doubts about the returns on AI capital expenditures are the second pressure. According to FactSet data, the S&P 500’s expected earnings growth for Q1 is 12.6%, with a forward P/E ratio of about 20.4 times. Tech giants have invested enormous capital expenditures over the past two years (Amazon plans to spend $200 billion by 2026, Microsoft, Meta, and others have also planned hundreds of billions), but AI-related revenues still lag far behind the scale of investments. A report from CEIBS cites estimates that about $400 billion in AI capital expenditure in 2025 would need to generate $1.6 trillion in annual revenue to break even, but actual revenues are only around $15-20 billion.

Slowing profit growth constitutes the third pressure. According to Bloomberg Intelligence, the “Big Seven” are expected to see an 18% profit growth rate in 2026, the lowest since 2022, narrowing the gap with the other 493 companies in the S&P 500, which are expected to grow at 13%. David Lefkowitz, head of US equities at UBS Global Wealth Management, stated in January that the trend of profit growth diffusion is happening, and technology is no longer the sole focus.

Nvidia at 21x, Microsoft down 23%: increasing divergence among giants

Valuation compression is even more pronounced at the individual stock level.

According to Zacks analysis, Nvidia’s forward P/E has fallen to about 21.4 times, well below its median of 45.3 times over the past decade, despite an expected annualized earnings growth rate of 39.1% over the next three to five years. Microsoft has declined about 23% since the beginning of the year, with its market cap dropping below $3 trillion from over $4 trillion last October. Apple remains relatively stable among the “Big Seven,” partly because its AI capital expenditure is much lower than peers, and it repurchased $24.7 billion worth of stock in a single quarter, with disciplined capital management earning a premium when the market punishes large spenders.

Insider activity may better explain the situation. According to SEC Form 4 data cited by Motley Fool, insiders of Nvidia, Apple, Alphabet, Microsoft, and Amazon sold a total of about $16.1 billion worth of stock over the past two years as of April 2. Although most of these sales are related to tax-related compensation, the lack of buy signals amid such large net sales still unsettles the market.

Debate over the AI bubble heats up, but there are fundamental differences from the 2000 internet bubble

Does the return of tech stock valuations to pre-AI levels mean the AI bubble has burst?

There are clear disagreements among institutions. BlackRock’s report on the tech sector states that the forward P/E of the S&P 500 Information Technology index was about 30 times in October 2025, still high historically but significantly below the peak of around 60 times during the dot-com bubble for the Nasdaq 100. BlackRock emphasizes that current valuations reflect real revenues, validated business models, and accelerating AI adoption, which are fundamentally different from 2000.

Goldman Sachs previously noted that while the implied long-term dividend growth embedded in current stock prices is unrealistically high, it remains below the extreme levels seen during the internet bubble and the “Beautiful 50” period of the 1960s.

However, warning signs are also evident. According to Globe and Mail, the S&P 500’s cyclically adjusted P/E (CAPE) ratio entered its second-highest valuation zone in 155 years at the start of the year. Historically, the two times the CAPE ratio exceeded 40 (during the internet bubble and January 2022), the S&P 500 subsequently declined by 49% and 25%, respectively.

Zacks analysts adopt a more pragmatic view: as stock prices fall, earnings expectations are being revised upward, leading to passive valuation multiple compression. The risk-reward ratio for some stocks is improving. Nvidia is considered the best match between growth and valuation at current levels, while Microsoft is seen as having potential for a “catch-up rebound.”

For investors, the key question is not whether AI is valuable, but whether the massive capital expenditures can be converted into profit returns aligned with valuations within a reasonable timeframe. If 2026 marks the cyclical peak of large-scale client capital spending, then even if technology continues to evolve, the investment return cycle around AI infrastructure may far exceed market patience.