BNP Paribas just raised its price targets for AMD, Nvidia and Intel.
Goldman Sachs expects AI spending to push S&P 500 earnings even higher. And on Friday, Nvidia pushed the Nasdaq-100 to a new record high.
Everyone is talking about an AI bubble. And yet, analysts keep raising their price targets.
THE UPGRADES
Let’s start with Lisa Su’s AMD. BNP says the company is no longer simply selling chips. It is becoming a complete platform for AI infrastructure. That is why it raised its price target to $960, representing a 60% increase.
Its Helios accelerators have become a “credible second source” after Nvidia. With commitments for 14 gigawatts and agreements with OpenAI, Meta and Anthropic. BNP believes AMD can capture at least 8% of the GPU market, which will exceed $1 trillion by 2030. The rating, however, remained Neutral.
Now let’s move to Nvidia. Its price target was raised to $345 from $285. And it remains BNP’s top semiconductor pick.
Why? Because of its broad range of hardware, its CUDA software and its networking products. And because it remains strong as AI moves into inference.
What does that mean?
In very simple terms, it is the process of running AI models after they have been trained. BNP also expects gross margins above 70%, despite the competition.
And then there is Intel. The price target is $125, up from $75, with a Neutral rating.
It should benefit from demand for server processors, as AI agents require increasingly more computing power. The big test, however, is manufacturing chips for third parties.
And until those customers materialize, BNP is staying on the sidelines.
TOWARD $6 TRILLION
And this is where things get even more interesting.
Yesterday, Nvidia rose 2.12% and hit a new all-time high. That ended seven weeks of stagnation surrounding Trump’s meeting with the heads of what he calls “superintelligence.”
And because Nvidia accounts for 13% of the Nasdaq, it pulled the Nasdaq-100 to a new record as well. Its market value reached $5.77 trillion.
“So where can it go from here?” you may be wondering.
The answer lies in options. Their prices indicate how likely the market considers a particular move in the stock.
And right now, they imply roughly a 50% probability that Nvidia will reach a $6 trillion valuation by the end of the month. Around 67% by December 18. And a one-in-16 chance of reaching $7 trillion by November 20.
Be careful, though. An equally large decline is just as possible.
S&P 500 EARNINGS
Now let’s move on to the other piece of the puzzle.
Goldman Sachs expects another strong earnings season. Estimates point to a 27% increase in S&P 500 earnings per share for the third quarter. And Goldman believes most companies will beat those estimates.
But look at where those earnings are coming from.
Technology and energy are responsible for almost 80% of the increase. AI infrastructure companies alone account for more than half. And Micron together with Nvidia account for more than one-third.
“And are all these billions being spent on AI actually generating revenue?” you may be wondering.
That is what will be determined now.
Spending by the major cloud providers is expected to increase by 116%. And cloud revenue at Amazon, Alphabet, Microsoft and Oracle is expected to accelerate to 55%, up from 48%.
Of course, it is not all rosy.
For the typical company, estimates point to just a 9% increase. And the typical stock is 17% below its record high, while the S&P 500 is only 2% below its record.
Goldman, however, sees earnings per share reaching $375 in 2026, up 36% from last year. And it expects the S&P 500 to reach 8,000 points by the end of the year and 8,700 within 12 months.
BUBBLE OR NOT?
And this brings us to the big question.
Is all of this a bubble?
Hou Wey Fook, Chief Investment Officer at DBS, says clearly: no.
Nvidia is currently valued at 17 times the earnings expected over the next 12 months. And its earnings are expected to increase by 70% over the next year.
“And what does 17 times earnings actually mean, Christos?” you may be wondering.
It means you are paying $17 for every $1 of earnings.
And to understand how low that is, consider this: before the dot-com collapse, Cisco was valued at roughly 100 times its earnings.
He put it very simply:
“When the symbol of AI is valued at around 15 times earnings, how can it be a bubble?”
That said, even he does not put all his eggs in the same basket.
He recommends a “barbell” strategy. On one end, technology stocks for growth. On the other, investment-grade bonds for stable income.
And in the middle, gold and hedge funds to spread the risk.