At a time when pessimism is already at its peak, the company argues in a newly released report that the growth rate of investments in artificial intelligence (AI) infrastructure has passed its zenith and that securing financing has become increasingly difficult. Every piece of research has its value, of course.

However, when the researcher is Citrini, one might want to exercise a bit more caution.

Let me recall the report they published in February regarding the software (SaaS) sector.

By February 22, when they released a report outlining a bleak future for software companies, the sector had already taken a severe hit.

As of February 18, shares of Salesforce and ServiceNow had already fallen by approximately 29% since the beginning of the year, while Adobe had dropped by 25%.

In other words, by the time they "discovered" the fear, investors had already paid a significant portion of the price for that fear.

Now, Citrini is doing something similar.

Citrini contends that growth in AI infrastructure investments has peaked.

I disagree with this, and I believe the evidence they present is insufficient to support such a sweeping conclusion.

Financing challenges and project delays are certainly significant.

But how do we leap from these points to the conclusion that demand for AI is falling and that the overall appetite for investment is weakening?

Furthermore, the fact that investments exceed revenues is not, in itself, proof of an impending slowdown.

Capital is spent on infrastructure first; the revenue generated from that capacity comes later.

The real points for discussion—beyond the sheer size of the gap between investment and revenue—are whether the capacity being built will find customers and whether the necessary financing can be fully secured.

My own expectation is that as the adoption of artificial intelligence spreads, there will be an even greater need for computing power and infrastructure investment.

This expectation could, of course, turn out to be wrong.

But to argue the opposite, one would need stronger evidence than just a few troubled projects. Citrini’s researchers conclude that we are nearing the end of the investment cycle, basing this on the increasing difficulty of building infrastructure.

I take issue with this specific connection.

The fact that something has become harder to build does not imply that the need for it has diminished.

On the other hand, the rise in artificial intelligence adoption certainly does not justify the current valuations of every single infrastructure stock.

Which company are we talking about? What is its debt burden, what are its customer contracts, and what is its price?

The devil is in these details.

Granted—

Citrini’s questions regarding financing and the return on investment may well be valid.

And the fact that their previous reports were well-timed does not automatically mean this new one is incorrect.

But the answer I am looking for is this: To what extent is the risk you describe already priced into the stock?

Because it is easy to tell an investor—after a market decline—just how dangerous the world has become.

What is difficult is identifying the price point at which fear transforms into an opportunity.