Have you ever watched a market that looked completely broken one day and then realized, years later, that the real opportunity was sitting right there in the middle of the chaos? That feeling hit me hard while following the latest tech selloff. Anthropic’s highly anticipated public debut was supposed to inject fresh energy into the sector. Instead, the opposite happened. Shares of the AI company climbed fast at the open only to give back a large chunk of those gains by the close, and the broader technology complex followed it lower. Suddenly everyone was talking about whether the AI trade had finally run out of steam.
Yet one well-known market voice is not buying the panic narrative. Tom Lee of Fundstrat has been making the case that what we are seeing right now could turn out to be one of those classic moments of maximum discomfort that later reward the investors who stayed calm. His comparison to an earlier chapter in technology investing is particularly interesting, and it is worth unpacking carefully.
Why the Recent Tech Selloff Feels Familiar to Some Veterans
The session that followed Anthropic’s listing was rough. The S&P 500 dropped more than two percent, marking its sixth straight day of losses. The Nasdaq fell even harder. The Dow shed over a thousand points. All seven of the so-called Magnificent Seven names finished lower. Nvidia, fresh off reporting eye-popping quarterly numbers, still declined several percentage points. Smaller names tied to AI infrastructure, such as certain specialized cloud providers, got hit harder still.
In the middle of that noise, Lee sat down for a conversation and offered a perspective that cut against the prevailing mood. He pointed back to 1997 and Cisco Systems. At the time, the internet infrastructure story was still young and full of questions. Cisco shares dropped sharply over a three-month stretch, sliding from roughly nine dollars to five. Plenty of people assumed the party was over. Those who held on, or who bought into the weakness, watched the stock more than double over the following year.
I think that we are at that moment where it feels like max pain, but for the patient investor, this is going to have a big payoff.
That single sentence captures the entire thesis. The pain is real. The question is whether the pain is the end of the story or simply a chapter that patient capital will later look back on with a certain amount of satisfaction.
What Actually Happened During Anthropic’s First Day
Anthropic came to market with a large offering. The company priced shares at eighty-three dollars and sold a substantial number of them. At the open, the stock jumped more than thirty percent, trading as high as the mid-one-hundred-teens before retreating. By the closing bell it settled near ninety-seven dollars. Investors who received an allocation at the IPO price still walked away with a solid gain. Anyone who bought at the opening print, however, finished the day underwater by a meaningful amount.
The size of the deal itself was notable. After expanding the offering following the roadshow, the company raised more than eighty-one billion dollars in gross proceeds. That figure put it in rare company among recent large listings. Underwriters still hold an option to purchase additional shares, which could push the total higher if exercised.
The market’s reaction was telling. Early enthusiasm quickly gave way to selling pressure that spilled across the entire technology complex. It was not simply a case of one stock disappointing. The whole sector seemed to lose momentum at the same moment.
The Cisco Parallel and Why It Still Matters
Lee’s reference to Cisco in 1997 is more than just a historical anecdote. In the mid-to-late nineties the internet was clearly transforming business, yet the path of investment spending was far from smooth. Questions about overbuilding, about returns on capital, and about which companies would ultimately dominate created bouts of sharp volatility. Cisco, as one of the purest plays on network infrastructure, absorbed a good deal of that uncertainty in its share price.
What followed, of course, was a multi-year period in which the underlying demand proved more durable than the temporary doubts. The same pattern has appeared in other technology cycles. The initial excitement creates high valuations and high expectations. Then a period of digestion arrives. Companies that were once unstoppable start to look vulnerable. Headlines turn skeptical. And then, gradually, the actual growth in usage and spending reasserts itself.
Whether the current AI cycle will follow a similar trajectory is the open question. Lee’s view is that the long-term demand for computing power remains intact even if the near-term financing environment has become more complicated.
Where the Money for AI Infrastructure Will Come From
One of the more interesting parts of the recent discussion has centered on financing. Building data centers, securing power contracts, and locking in specialized chips requires enormous amounts of capital. Some of the pure-play infrastructure providers have seen their borrowing costs rise. That pressure showed up in their share prices during the latest selloff.
Lee suggested that the next phase of demand may come less from specialized startups and more from large, established companies that are modernizing their own technology stacks. He floated the idea that certain major financial institutions could eventually evolve into meaningful providers of specialized cloud capacity for their own operations and, potentially, for clients. The broader point is that the customer base for AI infrastructure is likely to broaden over time rather than remain concentrated among a handful of hyperscalers and AI-native firms.
At the same time, regulators and central bankers are watching the balance sheets closely. Recent commentary from a Federal Reserve official highlighted that debt levels among the largest AI-related companies are expected to climb substantially. Off-balance-sheet commitments tied to leases and energy contracts could reach even larger figures. Those numbers are large enough to warrant attention, yet they also underscore just how much capital is already being committed to the build-out.
How Crypto Markets Reacted to the Same News Flow
Digital assets did not escape the pressure. Bitcoin had shown some resilience earlier in the day as traders positioned ahead of the Anthropic listing. Once the equity market turned lower, that strength faded and the leading cryptocurrency slipped back below key levels. Other major tokens followed a similar path.
Some market participants had argued that investors were selling crypto holdings in order to free up capital for the large technology IPO. Whether that specific flow was large enough to move prices is difficult to prove, but the correlation between risk assets was clearly visible. When technology equities sold off hard, crypto tended to move in the same direction.
The interplay between traditional tech stocks and digital assets has become more pronounced over the past couple of years. Both are sensitive to liquidity conditions, to interest-rate expectations, and to overall risk appetite. A genuine recovery in one often provides a supportive backdrop for the other.
Interest Rates and the Cost of Capital Still Matter
Another element that has weighed on the technology sector is the level of longer-term interest rates. The ten-year Treasury yield has been hovering near levels that many growth-stock investors find uncomfortable. Higher discount rates reduce the present value of distant cash flows, and most pure AI stories still have a large portion of their value tied to those future cash flows.
Minutes from a recent policy meeting showed that officials are aware of the rising financing costs facing AI infrastructure companies as well as the inflationary pressures that can accompany large-scale data-center construction. That awareness does not mean the central bank will pivot quickly, but it does mean the issue is on the radar.
In my view, the path of rates will continue to influence how quickly sentiment can improve. A sustained move lower in yields would almost certainly help risk assets. A further rise would keep the pressure on valuation multiples.
What Patient Investors Might Be Considering Right Now
Lee’s core message is relatively straightforward. Markets often deliver the greatest rewards after periods that feel the most uncomfortable. The current environment checks many of those boxes: consecutive down days, high-profile names under pressure, questions about financing, and a general sense that the easy money has already been made.
None of that guarantees a quick rebound. Corrections can last longer than expected, and individual stocks can remain under pressure even if the broader thesis eventually proves correct. Position sizing and time horizon become especially important in these moments.
Some investors will choose to wait for clearer signs of stabilization. Others will begin adding exposure gradually, focusing on companies with the strongest balance sheets and the most durable competitive positions. Still others will look for opportunities outside the largest names, in the specialized infrastructure providers that have sold off more sharply.
- Focus on businesses with visible cash-flow generation rather than pure narrative stocks
- Pay attention to capital expenditure plans and funding sources
- Watch the ten-year yield as a key sentiment indicator
- Consider how corporate demand for AI tools could broaden beyond the hyperscalers
- Maintain enough dry powder to take advantage of further weakness if it arrives
The Broader Context of the AI Investment Cycle
It is easy to forget how early we still are in the commercial deployment of advanced AI systems. Enterprise adoption is growing, but it remains far from universal. Many companies are still in the experimentation phase. Others have begun to see measurable productivity gains and are expanding their usage. That progression tends to be uneven rather than linear.
The capital spending required to support that adoption is enormous. Chipmakers, cloud providers, networking companies, and power infrastructure firms all stand to benefit if the demand materializes as projected. At the same time, the risk of overbuilding exists, just as it did during previous technology waves. The difference this time is that the potential productivity impact appears broader and deeper than many earlier cycles.
I have found that the most useful way to think about these periods is to separate the short-term price action from the medium-term fundamental trajectory. Prices can and do disconnect from fundamentals for extended stretches. Eventually they tend to reconnect. The investors who keep that distinction clear are often the ones who navigate the volatility more successfully.
Risks That Should Not Be Ignored
A balanced view requires acknowledging the genuine uncertainties. Financing conditions could tighten further. Power constraints in key data-center regions could slow the build-out. Regulatory scrutiny of AI applications could intensify. Competition among model developers could compress margins faster than expected. Any of those developments would make the current correction deeper or longer-lasting.
There is also the simple reality that valuation multiples for many of the leading names remain elevated by historical standards. Even if the growth story plays out, the path to higher share prices may involve a period of multiple compression alongside earnings growth. That combination can produce frustrating sideways or lower price action even while the underlying businesses continue to expand.
Perhaps the most interesting aspect is how concentrated the market has become. A relatively small group of companies has driven a large share of the index gains over the past couple of years. When those names come under pressure at the same time, the impact on broader indexes is outsized. That concentration risk works in both directions.
Looking Ahead: What Could Change the Narrative
Several developments would likely help sentiment recover. Clearer evidence that enterprise spending on AI tools is translating into measurable returns would reduce some of the skepticism. Any sign that financing conditions for infrastructure projects are stabilizing would also help. A sustained decline in longer-term yields would provide a more constructive backdrop for growth stocks in general.
On the company-specific side, continued strong results from the leading chipmakers and cloud providers would reinforce the idea that demand remains robust. If Anthropic and other newly public AI-related companies can demonstrate solid post-IPO performance over subsequent quarters, that would further calm nerves.
None of these catalysts is guaranteed to arrive on a particular timetable. Markets have a way of testing patience precisely when conviction feels hardest to maintain. That is usually the point at which the eventual payoff begins to take shape, even if it is not yet visible in the daily charts.
A Personal Take on Navigating This Environment
In my experience, the most dangerous emotion during these periods is the urge to do something dramatic simply because the market is moving. Selling everything at the bottom of a six-day losing streak rarely improves long-term results. Equally, loading up on leverage just because a respected strategist has turned constructive can create its own set of problems if the weakness continues for a while longer.
A more measured approach tends to work better. Reassess position sizes. Make sure the companies you own still meet your original investment criteria. Consider whether the recent price declines have created better entry points in names you already liked. And leave room for the possibility that the correction has further to run.
The comparison to earlier technology cycles is useful precisely because it reminds us that periods of doubt are a normal part of transformative investment themes. The internet did not stop growing after 1997. Mobile computing did not stop after the first wave of skepticism around smartphone valuations. Cloud computing survived multiple periods when investors questioned the pace of migration. Each time, the underlying trend eventually reasserted itself.
Whether the current AI wave follows a similar pattern remains to be seen. The early evidence on adoption and capital spending is encouraging. The financing questions and valuation concerns are real. Somewhere between those two realities sits the opportunity that patient investors are trying to evaluate right now.
The recent session that followed Anthropic’s public debut delivered a clear message: the market is in a more cautious mood toward technology and AI-related names. Consecutive declines, high-profile underperformance, and rising questions about capital intensity have created an atmosphere of discomfort. Tom Lee’s view is that this discomfort may ultimately prove to be a buying opportunity for those with a longer horizon. History offers some support for that perspective, even if it cannot guarantee the same outcome this time.
For investors willing to look past the daily noise, the next several quarters will provide important information about the durability of demand, the availability of financing, and the ability of leading companies to convert massive capital investments into sustainable returns. Those answers will matter far more than any single trading session, no matter how dramatic it felt at the time.
The road ahead is unlikely to be smooth. Corrections rarely end cleanly, and narrative shifts take time. Yet the fundamental drivers that attracted capital to the sector in the first place have not disappeared overnight. That tension between near-term price action and longer-term potential is exactly where thoughtful investors spend their time, and it is the space that Lee’s comments invite us to consider more carefully.