Mentioned in This Article
1. Lee Says the AI Selloff Has Found Its Floor
Tom Lee, co-founder of Fundstrat and chairman of Bitmine, said on November 2 that the recent selloff in artificial intelligence stocks has reached its bottom, arguing that demand for AI capacity is still limited by tight supply. In a CNBC interview, he added that crypto's muted reaction to the multibillion-dollar Binance hack suggests many investors have already sold.
Lee made the remarks on CNBC's Squawk Box. His blunt summary was that the market is at the bottom. The comments arrive after a bruising stretch for large technology shares, which matters well beyond individual AI companies because those same stocks sit inside many retirement-investing funds and broad index products.
Lee also acknowledged the limits of such a call. Picking a bottom is inherently risky, he indicated, because nobody can know in advance how much further prices might fall. His argument therefore rests less on a precise price level than on his reading of how supply, demand and investor positioning currently line up.
2. Why Does Lee Think This Is Not an Overcapacity Market?
Lee's central claim is that the AI industry is struggling to expand quickly enough, which is a different problem from having built more capacity than customers need. In his view the AI market remains very tight, and it is not a market with too much supply chasing too little demand.
That distinction is the basis of his optimism. An overcapacity market typically ends in falling prices, idle equipment and write-downs, and recoveries can take years. A supply-constrained market looks different: customers are waiting, orders are queued, and the main limit on growth is how fast suppliers can build.
According to Lee, constraints on AI spending, including political opposition, have slowed the pace of growth. He does not read that slowdown as proof the boom has peaked. Growth that is delayed by bottlenecks is, in his framing, not the same thing as growth that has ended, and he sees the current pause as the former rather than the latter.
The argument is a forecast and a matter of judgment rather than an established fact. Lee is presenting a market view, and the evidence he and others cite points in more than one direction, as the next sections show.
3. Strong Demand Can Still Produce Weak Earnings
Shortages cut two ways for companies that build AI infrastructure, such as data center operators and chipmakers. They can leave customers waiting for equipment, which signals healthy demand, and they can also prevent suppliers from converting those orders into revenue on schedule.
Intel illustrates the tension. On the Friday before Lee's interview, its shares fell nearly 30 percent to $65.76 after the company issued a weak forecast for the fourth quarter. Its data center and AI revenue had risen 78 percent from a year earlier to $7.3 billion, a figure that would normally be read as a strong result.
Manufacturing constraints, however, limited Intel's ability to meet demand. The company projected quarterly revenue of $15 billion to $16 billion, a midpoint about $1.6 billion below analyst expectations. A weaker PC business added to the pressure on the outlook.
The episode shows how a company can report booming AI demand and still see its stock punished. For Lee's thesis it cuts both ways: it supports the idea that supply, not demand, is the bottleneck, yet it also shows that investors are willing to sell even when the demand picture is strong, which keeps the question of a market floor open.
4. What Does the Binance Hack Say About Crypto Sellers?
Lee's second argument turns to crypto, where he points to the market's relatively restrained response to the security breach at Binance, which the exchange initially estimated at about $6.3 billion in impact. His reasoning is that a market able to absorb such bad news without a much larger decline may have few sellers left.
In his words, if crypto tends to lead on risk, then people have already been washed out. The logic is common among market strategists: capitulation happens when holders who intended to sell finally do so, and what follows is often a stabilization because the pool of motivated sellers has shrunk.
Lee tied the two markets together, suggesting that investors in AI shares and in digital assets are exposed to a similar risk appetite. If crypto has already absorbed a shock and held up, he argues, that is a constructive sign for risk assets more broadly.
Binance said on the Friday in question that about $3 billion of the affected assets had been frozen. It also pledged to cover losses using its emergency reserve and company funds, steps that may have helped limit the damage to confidence among users and traders.
5. The Evidence That Cuts the Other Way
The data does not point only in one direction. Bitcoin (BTC) fell about 4 percent to roughly $76,400 on Friday, and more than $1.3 billion in leveraged positions were forcibly closed during the move. Forced closures of that kind often accelerate a decline because they trigger further selling.
U.S. spot bitcoin exchange-traded funds also recorded $643.2 million in outflows that day. That was the largest single-day outflow for the funds in October, and it indicates that some investors were still withdrawing money rather than buying the dip.
There is a counterweight. A 4 percent drop is considered fairly ordinary for crypto traders and can occur even during bull markets, which tempers the significance of the liquidations and the fund outflows. Whether the day marked exhaustion or the start of further weakness is exactly what Lee's call is trying to settle.
6. Lee's Position and What to Watch
Lee is a prominent voice in both worlds. He leads strategy research at Fundstrat and chairs Bitmine, a company that accumulates ether, so his views carry weight with investors but also come from someone with a stake in the asset class. Decrypt, the outlet that reported his comments, disclosed that Lee is one of several investors in its parent company, Dastan.
For readers, the practical takeaways are straightforward. Lee believes the AI selloff reflects supply limits and not collapsing demand, and he reads crypto's resilience after a large hack as a sign that sellers are largely spent. The counterevidence includes Intel's sharp drop, large liquidations and the heaviest fund outflows of the month.
No scheduled catalyst accompanied the interview, so the test will come from market behavior itself: whether AI-linked shares stabilize as suppliers work through their constraints, and whether bitcoin and the funds that hold it stop bleeding capital. If prices keep falling, Lee's bottom call will be revised by the market rather than by commentary.
7. How Market Bottom Calls Are Usually Judged
A market bottom is only identifiable in hindsight, which is why strategists tend to frame such calls around conditions rather than dates. Common signals include exhausted selling, extreme pessimism in sentiment gauges, and the failure of bad news to push prices lower. Lee's reference to crypto's reaction to the Binance hack falls into the last category: bad news arrived and the market did not fall apart.
Analysts also distinguish between a price bottom and a valuation bottom. A price bottom is the lowest point reached before a sustained recovery. A valuation bottom is a level at which an asset looks inexpensive relative to expected earnings or demand. Lee's supply-constraint argument is aimed at the second idea, since he contends that the underlying business demand for AI has not weakened.
Investors who disagree typically point to positioning. If large funds are still reducing exposure, as the bitcoin fund outflows suggest, then prices can keep sliding even when the fundamental story is intact. This is why the two sets of evidence in the interview, supportive demand data and ongoing redemptions, are not necessarily contradictory. They describe different layers of the same market.
The link between technology shares and digital assets is itself a recurring theme. Both are widely treated as risk assets, meaning investors tend to buy them when optimism is high and sell them together when it fades. That correlation is why a comment about AI stocks from a crypto-focused strategist is relevant to bitcoin traders, and why Lee framed the two as sharing the same ship.
8. Why the Supply Question Matters to Investors
For investors the difference between a supply constraint and a demand shortfall shapes how long a downturn might last. A supply constraint is usually temporary, because companies respond by building more factories, securing more components and adding more data center space. A demand shortfall can persist for as long as customers stay away.
Intel's results show how this plays out in practice. The company reported rapid growth in its data center and AI business, but its own production limits kept it from selling more. Investors then focused on the lower-than-expected revenue forecast and marked the shares down sharply, even though the underlying demand figure was strong.
That reaction suggests that markets are currently rewarding delivery rather than promises. Lee's view is that, once capacity catches up, the revenue should follow. Skeptics would answer that the timing is uncertain and that share prices can stay depressed while the bottlenecks are cleared.
The same lens applies to crypto. A market that has already priced in a shock, such as a large exchange breach, may have less downside left, but only if new selling does not appear. The fund outflows and liquidations recorded on Friday are a reminder that such selling had not fully stopped when Lee spoke.