Rudi's View | 5:01 PM
The latest updates on Conviction Calls, Best Buys, strategy focus, and preferred stock picks for 2026.
In today's update:
- "Would the real AI bubble please stand up?”
- From the Investment Strategy Desk at Bell Potter
- Australia Remains Underweight at T.Rowe Price
- Wilsons' Conviction Calls
- Macquarie's Small-Cap Consumer Picks
- Macquarie's Portfolio Changes
- Ord Minnett's Conviction Calls
- Bell Potter's Stock Picks For 2026
- Shaw and Partners' Small Cap Champions
By Rudi Filapek-Vandyck, Editor
AI. Bubble? Fraud? Or Mega accountancy trickery?
Maybe it's simply the fourth industrial revolution and this is how society changes in real time.
The public debate continues to rage on and it's becoming an emotionally-driven affair in which voices on either side dig in and defend their established opinions (and conclusions) with all might available.
The latest contribution by the Deutsche Bank Research Institute is an honest attempt to assess the phenomenon from a neutral platform with attention for both pros and cons.
Let's go straight to the conclusion drawn: No, it's not a bubble, though it may well develop into one, and there are most definitely threats and risks as AI develops further.
This, of course, has been the view here at FNArena all along. It's great to see this debate can still be assessed in a well-researched, non-emotional, not pre-determined or overly biased manner.
Said document consists of only 23 pages, including cover and disclaimer, and mostly works through the questions and answers with well-chosen graphics we cannot reproduce here (copyright).
Hence, a summary in prose:
“Would the real AI bubble please stand up?”
The two authors of the study argue the debate around an “AI bubble” is badly framed, suggesting investors are dealing with three overlapping booms –-valuations, investment and technology-– rather than a single monolithic bubble.
While headline valuations appear stretched, the the current cycle differs materially from the late-1990s dot-com surge, with Big Tech’s cash-rich balance sheets and immediate revenue lift supporting the initial phases of the AI build-out.
Red flags do exist. US equity valuations, measured by the Shiller CAPE ratio, have pushed above 40, close to the dot-com peak of 44. Forecast AI-related capex could exceed US$4trn by 2030 –-more than ten times the inflation-adjusted cost of the Apollo program-– without any guarantee the investment earns an adequate return.
And while generative AI continues to impress, concerns are growing about scaling limits, including physical constraints around data movement between chips.
Yet, several green lights temper bubble talk. Tech valuations remain at the low end of their post-2022 trend channel, and the sector’s premium over the broader market reflects genuine earnings outperformance rather than speculative multiple expansion.
The most extreme valuations sit in private markets, not among listed megacaps, where balance sheets remain robust. Capex growth is consistent with a decade-long trend and still largely funded by free cash flow, with hyperscalers such as Google generating tens of billions in quarterly operating cash.
On the technology front, models such as Gemini 3 demonstrate continued capability gains, while costs per token have collapsed by a factor of 1,000 over recent years, widening viable commercial use-cases.
The authors caution this matters because AI-linked spending has been propping up US GDP growth, with underlying economic activity otherwise close to stalling. A reversal would therefore carry macro risk.
They outline five potential shocks: circular financing structures that obscure valuations; rising leverage even among hyperscalers; diminishing returns from model scaling; societal or political backlash; and hard-capacity constraints, notably electricity supply, where global demand is projected to quadruple by 2030.
Bottom line: the Deutsche Bank Research Institute sees pockets of excess but rejects the idea of a single AI bubble ready to burst. Instead, the current cycle looks early-stage, cash-funded and still delivering tangible technological and economic gains –- but one where investors must remain alert to structural vulnerabilities.
One of the stand-out conclusions from the report: "US would be close to recession this year if it weren’t for tech-related spending, as other spending has flatlined post-Covid".
The full story is for FNArena subscribers only. To read the full story plus enjoy a free two-week trial to our service SIGN UP HERE
If you already had your free trial, why not join as a paying subscriber? CLICK HERE