Tom Lee’s latest interview: S&P sees 8,000 by year-end, but there could be a bear-market-level pullback this fall

To be honest, Tom Lee’s calls on crypto assets like ETH don’t differ much from what retail investors would say, and his repeated judgments also haven’t been especially precise. But you have to admit that his annual outlook on the mainline of U.S. stocks has consistently had a fairly solid logic framework and supporting rationale—it’s worth breaking down carefully.
In this interview with Prof G Markets, he laid out the logic behind the S&P 8,000 level in a systematic way, and he also pre-emptively warned about the possibility of a sharp pullback in the fall. I’ll first整理 his core观点, and then add a few of my own views at the end.
Mid-year recap: U.S. stocks have logged a fourth consecutive year of double-digit growth, but valuations are actually getting cheaper
After a strong rally in U.S. stocks from 2023 to 2025, in the first half of 2026 the S&P 500 is up nearly 9% again. It looks like it has already been rising for a long time, but the data Tom Lee provided is kind of counterintuitive.
Valuations are actually getting cheaper. At the start of the year, the market consensus for S&P 500 2027 earnings per share (EPS) was $350, but it has since been raised significantly to $400. Meanwhile, the S&P 500’s price-to-earnings ratio based on 2027 forward earnings has actually fallen—from 19.4x at the start of the year to 18.4x now. In other words, earnings are rising faster than the stock price, and valuation digestion is healthier than people expect. Using 2027 EPS of $400 multiplied by a reasonable 20x P/E, he raised his S&P 500 year-end target price from the previous 7,700 to 8,000.
A V-shaped, “bear-market-level” pullback could happen once in the fall—watch four variables
However, Tom Lee also clearly warned that before the market pushes up to 8,000, during the period from June to December (especially before the autumn midterm election), it will face four key tests, which could trigger a relatively sharp, bear-market-level pullback. Still, he believes this pullback will be V-shaped and the rebound will be fast and strong.
The first variable is policy uncertainty brought by the newly appointed Fed chair, Walsch. Walsch has set up five working groups to reshape how the Federal Reserve operates, including redefining the inflation measure, canceling routine press conferences, and removing forward guidance. This will make the market lose the policy anchor it has been used to relying on. Once that anchor is missing—i.e., the Fed can’t soothe the market in advance—volatility will naturally amplify.
The second variable is the unlock wave of IPO shares for SpaceX and other big tech giants. SpaceX’s current valuation is $1.5 trillion, but the float is only $90 billion (IPO fundraise of $18 billion). During the fall into year-end, there will be large-scale share releases, so the supply of shares the market needs to absorb will clearly increase.
The third variable is potential shortages in oil derivatives triggered by the situation around the Strait of Hormuz and Iran. Even though U.S. gasoline supply is sufficient, a shortage of industrial oil derivatives like lubricants could drag on the global supply chain.
The fourth variable is elevated margin debt. Year over year, margin debt has surged 55%, the fifth-highest growth rate over the past 70 years. This often signals that leveraged retail and short-term traders are about to run out of buying power. Historically, after signals like this appear, the following period within the next 6 months is often accompanied by a relatively large correction.
Semiconductors are breaking traditional cycles, and humanoid robots are the next super source of demand
In response to criticisms like “cyclical procurement among AI companies” and “inflated profitability quality via book investment gains,” Tom Lee’s reply is: while tech stocks’ Shiller P/E is indeed high, tech companies currently account for nearly 60% to 70% of S&P earnings growth and 40% of total earnings volume. That’s fundamentally different from the tech stocks in the 1999 internet bubble, which essentially lacked real earnings support—these are two completely different situations at the core.
On semiconductors, he offered an interesting view. Semiconductors currently account for 19% of the weight in the S&P 500. He believes semiconductors are breaking the traditional 50-year cyclic rule and entering an entirely new growth story. The core driving force is humanoid robots: the chip demand per humanoid robot is roughly 50 times that of an iPhone. Add in demand for space-grade chips, and the semiconductor long-term market opportunity (TAM) is being reshaped from the ground up. He also mentioned that AI is filling a large amount of idle time in white-collar jobs. Over the next few years, industrial robots with more precise hand flexibility will reshape residential construction (for example, creating residential details comparable to those at a Louvre-level) and the entire ecosystem of logistics giants like Amazon.
The sectors and key stocks he likes
On the sector level, he favors small-cap stocks, financials, industrials, energy, and the “Seven Giants” and software sectors (IGV). On the individual stock level, he specifically highlighted Microsoft and Meta. Even though near term the stock prices pulled back due to increased capital expenditures, these two companies have repeatedly proven their foresight in strategic transitions and their execution capability in history. He is very bullish on them being the biggest beneficiaries of AI downstream applications. Amazon was also called out separately. As a logistics giant and a company that currently has 1 million industrial robots, he believes Amazon will be one of the largest beneficiaries in the AI and robotics era.
My own take
I basically agree with Tom Lee’s logic framework, especially the rhythm of “rally first, pull back, then rally again,” which overlaps with my own view over this period.
My specific timing call is: in July and August, the broad market will first surge toward around 7,800. In the seasonal-weak window from September to October, before the midterm election, there will be a pullback. I think the depth would likely be in the range of 7,000 to 7,200. Then in November and December, i.e., the fourth quarter, it will reignite the push again. The year-end target would be 8,000 to 8,200, which is basically aligned with Tom Lee’s 8,000-point target.
On the magnitude of this pullback, my view is that it won’t exceed 10%. Because this year’s March already saw a pullback of around 10%. Historically, it’s rare for two deep pullbacks of the same magnitude to occur consecutively within the same year. So I lean toward this one being in the 5% to 10% range, not worse than the March one.
The window where I think you really need to be more careful is the third quarter (mid-August through the end of September), when uncertainty and volatility will clearly rise. Besides the variables Tom Lee mentioned—Fed policy uncertainty, the IPO unlock wave, oil-supply-chain risk, and margin debt—I want to add one more: the political uncertainty ahead of the midterm election itself. In election-year historical statistics, volatility in the market often rises systematically in the months before the election. After the election outcome is settled and the uncertainty fades, the market can then enter a corrective rebound. That’s essentially the same logic as Tom Lee’s “accelerated push in the fourth quarter,” just from the other side.
I also agree with his view on the robotics theme. I’ve always felt that robotics will be the next “super main theme” the market starts pricing ahead of time, after AI infrastructure and the space sector. The logic is similar to the way expectations for SpaceX’s IPO helped drive a re-rating of the entire space industry chain back then. And this theme will ultimately feed back into demand on the semiconductor chip end. The figure Tom Lee mentioned—“humanoid robot chip demand is 50 times that of an iPhone”—also matches my earlier analysis on the robotics industry chain. The highest certainty is in the core components and chip suppliers that no matter which OEM wins at the end, can’t be avoided. When the broad market pulls back in October, focus on this segment—it could pop in Q4.