No investor can have failed to notice the huge rises in the Nvidia (NASDAQ: NVDA) stock price in 2024 and earlier. It’s up 200% in the past 12 months, and a massive 2,300% over five years.
But it has all the marks of a tech bubble stock, just waiting to go pop, right? Maybe not.
Slowing AI spend?
Last year was full of tales of all kinds of companies we’d never heard of pumping cash into AI.
But, increasingly towards the end of the year, analysts were getting worried. How will businesses turn all this investment into profit? Won’t they surely have to cut back on that spending in 2025? Are we in for a big slowdown?
Those concerns are surely valid. Remember those long-ago days when everyone was piling into this new-fangled internet thing with few having an idea how to make money from it?
But we know what happened to the clued-up companies that got it right. They turned many of their shareholders into millionaires. Even billionaires.
Microsoft knows
If anyone knows how to develop a profitable AI business, I’d say surely Microsoft does. Is it likely to scale back its AI outlay in 2025?
Well, we’re only a few days into the New Year, and already Microsoft has announced plans to boost AI spend in 2025, with $80bn earmarked for AI data centres. A fair chunk of that should, presumably, end up as revenue for Nvidia.
It does look like the timing of the announcement could have a political aspect, with Microsoft president Brad Smith urging the incoming US administration to go easy on AI regulation.
But the next few years could see competition from China for building the world’s AI infrastructure and creating global standards.
Can’t afford to keep spending on AI at such high rates? Maybe the big players can’t afford not to.
No bubble valuation here
Bubble stocks are often marked by valuations that defy rationality. Price-to-earnings (P/E) ratios up in the thousands, and that kind of thing.
But Nvidia doesn’t have one. A forward 2025 P/E of 51 does look a bit rich. But it’s nothing like some of the stratospheric valuations we’ve seen.
And if earnings grow the way the analysts forecast, the Nvidia P/E could drop to 27 by 2027.
We’ll have to wait and see if Nvidia can beat earnings expectations yet again, in its next quarterly update in February. If it should, say, only manage to equal expectations, I think a lot of bulls could sell and drop the price.
Competition
I reckon a key risk over the next few years has to come from competition. Nvidia might be making the most in-demand AI chips today. But I’d never write off Intel, Advanced Micro Devices (aka AMD), or even ARM Holdings.
And as Apple‘s in-house CPU technology goes from strength to strength, maybe we’ll even see some inroads from that direction before long.
But the one big threat going into 2025, that the big players will scale back their AI spend? I think it could be overblown. I’ll consider buying on any fear-led price falls.
