Are We Approaching “Peak AI’?

AI Boom Under Increasing Scrutiny

In recent days, chatter are swirling on the topic of ‘Peak AI’.

“Tech stocks,” wrote one FT columnist last week (£) after tech shares weakened “are sending a warning.” Another FT article joined the debate (£): “Is it time to sell your AI stocks?” Even the Wall Street Journal inserted a front-page article (£) on Monday (August 25) “Tech Shares Face a Murky Outlook”. Seemingly, scepticism is growing about the sector’s once-unquestionable prospect.

Given the popularity of AI stocks among GMG readers, perhaps it would be a good idea to review the sector this week. To assess things on the ground, so to speak.

For the past two years, Artificial Intelligence (AI) spearheaded the stock market boom. Generative AI (chatgpt.com – now version 5) has come of age and there’s no looking back from this newfangled technology. Investors piled into the sector and conveniently ignore the rest of the market. Why shouldn’t they? AI is the fastest growing sector and the industry’s returns are, for now, matching the hype. Large tech platforms are investing heavily in AI to enhance their business productivity and to defend their tech moats.

The “Magnificent 7” has grown to be an asset classes of its own (see below). In late August, these seven Big Tech companies are worth almost $20 trillion in total.

At $4.3 trillion in market cap – the largest in the world – Nvidia (NVDA) alone is nearly a trillion larger than FTSE 100 and 250 combined (londonstockexchange.com) .This tells you that AI stocks have the best trading liquidity, price momentum, and earnings growth. Funds love them.

Source: yardeni.com 

What Could Trigger an AI Bust?

But remember this old saying: Trees don’t grow to the sky. At some point in the future, the sector will experience some (or severe, depending on your view) financial turbulence.

First, though, are we in an AI bubble? Many say ‘yes’. Even Sam Altman, developer of the famed OpenAI, recently admitted that investors are “overexcited by AI” and that  “when bubbles happen, smart people get overexcited about a kernel of truth”.

When one of the earliest promoters of AI is scratching his head about recent funding deals, this is a warning sign. Over-eager investors are dumping oceans of cash on newly-created AI companies. Some of which will have overoptimistic business models. A re-run of the infamous ‘Pets.com’ is a distinct possibility.

Second, if we are in an AI bubble what could trigger its collapse? In 2000, repeatedly rate hikes and over-valuation crashed the DotCom bubble. Monetary tightening typically increases the cost of borrowings which, in turn, exerts downward pressure on economic activity. Less-than-expected revenue and earnings from dot-com startups caused investors to dial down their lofty expectations. This inevitably led to a further sell-off. A vicious cycle imploded the asset bubble.

Are we, then, about to face the same thing? History rhymes, but not in the same manner.

For one, President Trump has been trying frantically to get the Federal Reserve to drop rates, to the point of attacking the venerable institution left, right and centre. Sooner or later, the policy rate will be cut. For example, after the annual Jackson Hole central bank symposium, the market widely assumes that the Fed will slash rates in September.

Lower borrowing costs may prolong the bubble longer and fuel further speculative activities. Animal Spirits are riding high these days. The riskier the securities, the better price returns. Goldman Sachs, the US investment bank, has done some studies and found many speculative sectors are enjoying stunningly high returns in the higher-tariffed world (see below).

Source: GMO (Aug 2025)

Over-Investment in AI?

Another factor that could cause a retrenchment in AI is over-investment. For example, four Big Tech firms (AMZN, META, GOOG, MSFT) are estimated to be spending $750 billion over 2025-6 to build data centres and training AI models.

More worryingly, these investments will probably rise further as competition intensifies. Unlike Nvidia, the AI boom did not usher in an era of ‘bumper profits’ for these tech platforms. Instead, it drains substantial financial resources from their balance sheets as they fight to own a slice of the new, AI-powered digital world. They are forced to purchase ever-more expensive computer chips, just to keep up.

This begs the important question: Will Big Tech generate sufficient returns from these vast spending? Nobody knows. But investors may, at some point, question the viability of these gargantuan investments. The ‘DeepSeek’ moment earlier this year was a harbinger of things to come. What if some other tech startups come up with a cheaper model? It may immediately render a large swathe of current investment obsolete. Do not underestimate the ingenuity of entrepreneurs.

When Big Tech rein in their spending, it may cause an “AI Winter’.

Source: Financial Times (paywall)

In summary, whether you should continue to invest in the AI concept depends on a few factors:

Pros

• Fast-growing tech sector, with reasonably good medium-term prospects
• Increasing productivity of large tech firms; potentially, enhance their earnings power (cementing digital moat)
• Fast-changing technology, with new potential unicorns in the pipeline

Cons

• Valuation of AI firms rich (some say ‘bubbly’)
• Vulnerable to earnings and revenue disappointments given lofty expectations
• Technically overbought – AI stocks vulnerable to retracements, profit taking, and large swings in prices

In the next section, I review some AI-related stock charts.

Technical Review of AI Stock Charts

Nvidia (NVDA) – Prices have been surging to new all-time highs nonstop since late June. Given its close proximity to the magical $200 round number level, the stock may aim for a final push to this area later this year. Short-term support is noted at around $150; prices may re-test that level as support to work off the overbought momentum.

Broadcom (AVG) – The $1.4 trillion semi-conductor firm is riding high on the AI boom. Prices more than doubled from April lows. Overbought, the stock is hitting some selling pressure near $320. Like NVDA, the stock may correct into the prior peak (at around $240-250) before resuming its long-term uptrend.

 

Google (GOOG) – Google did not gain much over the past year. Prices yo-yo between $200-150 for several quarters as investors tried to work out the implications of AI on its dominant search business. As Google AI (eg Overviews) gains traction, prices have started to move north of $200. A further rally from here is possible, especially once prices decisively clear the January peak.

Meta (META) – Has been outperforming the sector for some time. The social-media company gained more than 600 percent since late 2022; but is hitting temporary selling barrier at $800. Overbought, the stock may extend its consolidation back into $600-700 before reasserting the long-term uptrend.

Microsoft (MSFT) – The world’s second most valuable company (at $3.8 trillion) may have hit a temporary peak with the ‘blow off’ peak to $550. Prices are correcting steadily back into the round number level at $500 – a level that may cushion the pullback. If this level gives, the next support is noted at $450, around 10 percent below current prices. Even at that support, MSFT retains the long-term uptrend.

Tesla (TSLA) – The EV carmaker is joining the AI race via xAI (x.ai – Grok). After Musk stepped down from Trump’s administration, the stock has been bouncing up and down the $250-350 range. This sideways consolidation is affirming TSLA’s long-term price uptrend. A bullish breakout at $360 opens door to a quick run to $400.

Other stocks not covered here but could be interesting to monitor:

• AMD (AMD)
• Apple (AAPL)
• Alibaba (BABA)
• Palantir (PLTR)
• Arista Network (ANET)
• Intel (INTC)