{"id":2284,"date":"2026-09-15T09:39:50","date_gmt":"2026-09-15T13:39:50","guid":{"rendered":"https:\/\/advisor.wellington-altus.ca\/klugeadvisorygroup\/?p=2284"},"modified":"2026-09-14T16:55:54","modified_gmt":"2026-09-14T20:55:54","slug":"is-ai-creating-a-stock-market-bubble-what-the-data-says-in-2026","status":"publish","type":"post","link":"https:\/\/advisor.wellington-altus.ca\/klugeadvisorygroup\/is-ai-creating-a-stock-market-bubble-what-the-data-says-in-2026\/","title":{"rendered":"Is AI Creating a Stock Market Bubble? What the Data Says in 2026"},"content":{"rendered":"

Clients have been asking this question since late 2022, when the original ChatGPT debuted and stocks connected to artificial intelligence (AI) appreciated sharply. The articles, op-eds, X posts, YouTube videos, etc., suggesting we are in a bubble have kept coming.<\/p>\n

The fear of a bubble is as legitimate as it is understandable. Past bubble collapses have impacted generations of investors, including some of history\u2019s most brilliant minds; nobody wants to be \u201cholding the bag\u201d when things unravel.<\/p>\n

We know from history that fear is not the investors\u2019 friend so here is what the data is saying right now.<\/p>\n

Has every major technology revolution created a bubble?<\/p>\n

Yes!<\/p>\n

When technology genuinely improves productivity at scale, it generates overinvestment. Past examples include canals, railways, electrification, and fibre optics (internet). Capital floods in faster than the underlying economics can justify as investors and businesses reach for the technology\u2019s promise. As a result, investment prices detach from reality. The bubble inflates and eventually corrects.<\/p>\n

So when people ask whether AI will eventually generate a bubble, they are asking the right question. History says it happens and I am not going to argue against history.<\/p>\n

However, for investment purposes, the relevant question is not whether it will happen; it is whether it is happening now or in the foreseeable future.<\/p>\n

 <\/p>\n

What does a real stock market bubble look like?<\/h3>\n

A bubble is when share prices detach from earnings reality. Prices go up. Earnings do not follow or follow too slowly. The gap widens until it is unsustainable and \u201csomething breaks\u201d.<\/p>\n

The poster child of the dot-com bubble was Cisco Systems. At its peak, Cisco traded at approximately 140 times forward 12-month earnings. The company was one of the leaders of the networking\/internet revolution. The price was disconnected from any reasonable projection of what it could earn in any reasonable time frame.<\/p>\n

By contrast, the \u201cposter child\u201d of the current AI cycle is Nvidia. Based on FactSet data, Nvidia trades at approximately 17 times forward 12-month earnings. Current estimates see NVIDIA\u2019s earnings growth at approximately 65% in the next 12 months. A company whose earnings are growing that fast (or even half that rate) with that valuation is the opposite of a bubble.<\/p>\n

 <\/p>\n

Is the broader technology sector overvalued?<\/h3>\n

No!<\/p>\n

The data says the opposite.<\/p>\n

When you look at the whole U.S. technology sector, forward earnings per share are rising significantly faster than share prices. That means that the stocks are getting cheaper relative to what the companies are earning. The forward 12-month earnings ratio of the U.S. tech sector is currently at 21.9 times vs. the five year average of 23 times and the 10-year average of 25.5 times.<\/p>\n

This is also the opposite of a bubble.<\/p>\n

 <\/p>\n

Is AI revenue real or is it still just a promise?<\/h3>\n

One of the best arguments against AI twelve months ago was very simple: where is the revenue?<\/p>\n

The bottom line is that it is arriving. In addition to a nearly 700-billion-dollar backlog of orders, Microsoft is reporting substantial AI-related earnings growth. Anthropic (one of the leading AI labs) is projected to grow revenue seven times this year and has reportedly achieved an operating profit in Q2 2026, per Bloomberg reports. Revenue growth of that magnitude has not been seen before at this stage of a technology cycle.<\/p>\n

The money is showing up.<\/p>\n

 <\/p>\n

What would need to be true for the bear case to be right?<\/h3>\n

For the negative case on AI to hold, you need to believe that demand for compute is now saturated.<\/p>\n

This would mean that:<\/p>\n