Is AI already on the verge of a bubble? History is reminding us of one thing.

Recently, there is a piece of data worth noting:

The weight of AI-related stocks in the S&P 500 has approached 40%.

Many people see AI and think this is the future, the trend, and an opportunity that cannot be missed.

There’s no mistake in that; AI is indeed a long-term direction.

But the problem is, even in the best tracks, if you buy at the hottest, highest-valued, most crowded positions driven by emotion, it can still be very uncomfortable.

What we really need to be cautious about is not whether AI has a future, but whether the market has already front-loaded too many expectations.

1. The 40% level is not ordinary in history

There is an important phenomenon in the market called excessive concentration of funds.

Simply put, all the money in the market is rushing into a few popular stocks or sectors.

Now, with AI-related stocks approaching 40%, it means that a large part of the gains in the US stock market are already dominated by the AI sector.

The market looks very strong, but in fact, it is increasingly relying on a few core assets.

It’s like a table that should be supported by four legs, but now more and more weight is being placed on one or two legs.

In the short term, the table can still hold.

But once these legs loosen, the overall market volatility will significantly amplify.

2. Similar situations have occurred several times in history

In 1972, the U.S. “Beautiful 50” bubble.

At that time, market funds were疯狂ly chasing a group of companies considered the best, most stable, and most growth-oriented.

Everyone thought these companies would never fall, and buying at any price was fine.

As a result, the U.S. stock market experienced a long adjustment period, with the index retracing nearly 50% in stages.

In 1990, the Japanese stock market bubble.

At that time, Japanese assets were globally sought after, and the Japanese stock market once had a very high proportion in global indices.

Market sentiment was extremely optimistic, with everyone believing Japan’s economy would continue to lead the world.

But after the bubble burst, the Nikkei index plummeted significantly within two years, entering a long adjustment period afterward.

In 2000, the internet bubble.

Everyone believed the internet would change the world.

In fact, the internet did change the world.

But this doesn’t mean that those who bought internet stocks at the high in 2000 all ended up making comfortable profits.

Many tech stocks were overvalued at that time, with the Nasdaq experiencing a maximum retracement of nearly 78%.

This is the cruelest part of the market:

The right direction doesn’t mean the price is right.

Having a future industry doesn’t mean there are no bubbles at present.

3. The current problem with AI isn’t lack of prospects, but overcrowding

The current AI sector has several similarities to past bubble phases:

First, market funds are highly concentrated.

Second, valuations are already high.

Third, investor expectations are very high.

Fourth, volatility is beginning to significantly increase.

Fifth, many people are buying not because they understand the company’s value, but because they fear missing out.

This is the most dangerous part.

When a sector shifts from “having prospects” to “everyone must buy,” risks are already accumulating.

Because in the end, the market is not trading reality but imagination.

As soon as earnings slightly miss expectations, or expectations for rate cuts change, or funds start flowing out of overvalued sectors, adjustments can come very quickly.

4. So what should we do now?

My approach is very simple, no need for complexity.

First, continue dollar-cost averaging on long-term favorites.

If you truly believe in the long-term development of AI, don’t try to go all-in at once.

Buy a little every week or month, using time to average down the cost.

The benefit of dollar-cost averaging is:

No need to guess the top, nor to bet on the bottom.

When prices go up, you haven’t missed out.

When prices go down, you still have the chance to lower your average cost.

This is much more comfortable than going all-in at a high point.

Second, always keep cash on hand.

Not chasing now doesn’t mean being bearish.

It means waiting for a more comfortable position.

Many people’s biggest problem isn’t not understanding the opportunity, but when the opportunity truly arrives, they have no money left.

Every major market decline will淘汰 a batch of fully invested people and reward those who still hold cash.

Cash isn’t wasting opportunities; cash itself is an opportunity.

When market sentiment is hottest, cash is defensive.

When the market is panic-selling, cash is offensive.

5. The most core message

AI is a long-term direction, but that doesn’t mean blindly chasing highs now.

History has told us many times:

When a popular sector’s weight in the index approaches 40%, the market is often already in a very crowded position.

This doesn’t necessarily mean an imminent crash.

But it definitely means the risk-reward ratio is no longer as comfortable as before.

So at this stage, the most important thing isn’t being aggressive, but pacing.

If you’re optimistic about the long term, keep dollar-cost averaging.

If you’re worried about high levels, hold cash.

Don’t let market excitement push you into a no-retreat situation.

History doesn’t repeat simply, but it often rhymes.

Maintain dollar-cost averaging, hold cash, and wait for the wind.