📈 It's not how much it rose, but how fast.

The US stock market has been strong again recently. The S&P 500 has climbed back near its all-time high, and AI and semiconductor-related stocks are leading the entire market. In particular, $MRVL Marvell Technology surged around 30% in a single day after NVIDIA CEO Jensen Huang mentioned it could be the “next trillion-dollar company.”

Seeing this scene naturally brings two thoughts to mind.

One is, “The demand for AI infrastructure is truly strong.”

The other is, “Isn’t this pace a bit excessive?”

The key point I see is not simply “it will fall because it has risen a lot.”

The important question is whether this rally is explained by fundamental improvements, or if it's a move created by a concentration in AI/semiconductors and an over-acceleration of fund flows.

1. A Good Story and a Good Price Are Different

The directional trend for AI infrastructure itself remains strong.

Data centers are getting larger, AI models are demanding more computation, and within servers, bottlenecks are expanding beyond just GPUs to include networking, storage, memory, packaging, and power.

From this perspective, the reason $MRVL is gaining attention is also clear.

AI data centers require thousands of chips to operate like a single giant computer. To do that, chips must be connected to chips, servers to servers, and equipment within the data center must be connected quickly. The “connectivity” Jensen Huang mentioned refers to this very point.

Therefore, $MRVL is not just a semiconductor company, but a stock to watch for the connectivity bottleneck in AI clusters.

However, a good industry and a good entry price are different things.

The fact that $MRVL is an important company and the fact that a price that surged over 30% in a day is a comfortable price are two completely different stories.

2. The Real Question of This Rally Is 'Pace'

The fact that the S&P 500 has risen at a historically rapid pace over the last two months is hard to ignore.

In the past, such sharp rallies typically occurred in two phases.

First, a strong recovery market following a recession.

Second, a phase where overheating and structural vulnerabilities built up together, like just before Black Monday in 1987.

Of course, this doesn't necessarily mean this time is the same as 1987.

Rather, there is now a strong earnings story called AI, and some companies are proving those expectations with actual revenue and profits. In this respect, we should be cautious about simple dot-com bubble-style comparisons.

However, when the market becomes too concentrated in one area, other problems arise.

More money flows in because stock prices are rising,

and stock prices rise again because more money is flowing in.

During this time, the fundamentals can still be good.

The problem is when the price starts to run ahead of the fundamentals.

3. A Rally Driven by Fund Flows Can Fall Just as Quickly

When looking at the current market, the part to be cautious about is not the AI story itself, but the concentration of fund flows.

Passive funds are forced to buy more of the stocks whose market caps have grown.

CTAs and trend-following funds mechanically increase their buying exposure in a rising market.

In the options market, demand for call options can further push up stock prices.

Thematic ETFs buy up stocks labeled as AI and semiconductor all at once.

These funds become a powerful force in a bull market.

However, they are also fast when moving in the opposite direction.

If stock prices fall below a certain level, trend-following funds can turn to selling, and options positions can amplify volatility in the opposite direction. Passive and ETF funds can also be shaken together when money flows out of a specific theme, regardless of the fundamentals of individual companies.

Therefore, the necessary attitude right now is not AI pessimism.

Rather, it is to acknowledge the long-term direction of AI while examining how much of the short-term price is driven by the over-acceleration of fund flows.

4. Look at the Stocks, but Also Consider the 'Pace Risk'

The group of stocks to watch in this trend remains clear.

These stocks are linked to the structural growth of AI infrastructure.

However, this may be a period where “at what price should I look?” is more important than “what should I buy?”

Especially for a stock like $MRVL that was re-evaluated by a large margin in a single day, it's necessary to check if short-term expectations have been priced in too quickly, separate from the company's direction.

A good company and a good price are different.

Summary from an Investment Perspective

The important thing in this market is not whether AI is over.

AI is not over yet.

Rather, infrastructure demand continues to broaden.

But the market sometimes prices in a good story too quickly.

Conclusion

This AI rally is difficult to view as a simple thematic market.

Real demand exists across AI data centers, networking, memory, and power infrastructure. The reason $MRVL is getting attention is not just because of Jensen Huang's one comment. As AI clusters grow, connectivity clearly becomes a critical bottleneck.

However, the market loves a good story, and when a good story spreads too quickly, the price sometimes runs ahead of reality. Therefore, now is not the time to be negative on AI, but to check the pace of the AI rally.

It's not risky just because it has risen a lot.

If price and fund flows are running faster than the pace of fundamental improvement, that is when you need to be cautious even within a good industry.

AI is a structural growth story.

But even structural growth stocks should always be bought only at a good price.