Bubbles Create a Shadow – Time to Move out from the Shade

Roger Montgomery | Montgomery Investments 

Let’s state the obvious right at the outset: No one knows for certain if the equity market will crash. The reality of investing is that we can rarely identify a bubble until after it has burst, which means we can never be 100% sure when we are inside one.

However, with that necessary disclaimer in place, I am reasonably confident that investors should prepare for a shift in 2026. 

While the last few years have been kind to bulls – something we’ve predicted since 2022, thanks to a combination of positive economic growth, disinflation and ample liquidity, which has been a positive blend since 1970 - I currently expect greater volatility and lower returns ahead. Stock markets tend to "cast their shadow before them," meaning investors will begin to reduce their exposure before any problems arise.

To navigate this, we must understand the anatomy of a bubble, acknowledge today's AI boom, and identify where to find shelter.

1. What is a Bubble?

We often look for a precise definition, but bubbles are best understood by the behaviours they encourage and those that create them. Roughly three-quarters of General-Purpose-Technology booms—like the internet, electricity, or aviation—share a similar pattern: Hype that brings down the cost of capital, colossal over-investment, financial carnage through creative destruction, and then, eventually, lower prices, which encourages wide adoption, followed by productivity gains built on the original infrastructure.

Statisticians might attribute the latter stage of a bubble’s lifecycle to a ‘regression to the mean’, but a clearer, if admittedly anecdotal, indicator of the bubble’s peak was often found on the newsstand. The "Time Cover Curse" suggests that gracing the cover of TIME Magazine isn’t a coronation but a harbinger of doom. By the time an investment theme makes it to the cover, it’s a saturated theme that’s already pulled in the very last of those who were ever going to buy in.

History is peppered with examples. Jeff Bezos was TIME’s Person of the Year in 1999, right at the peak of the dot-com bubble; Amazon's stock subsequently fell more than 90 per cent. Elon Musk appeared in 2021; the following year, his net worth plunged by US$200 billion as Tesla shares tanked.

A bubble is also mathematically observable in the valuations produced. Benjamin Graham, the ‘intellectual dean of Wall Street’, noted "the future value of every investment is a function of its present price. The higher the price you pay, the lower your return will be". When Price Earnings (PE) ratios become extremely high—reflecting extreme popularity—you run the risk of the market becoming less popular in the future, which eats into your returns.

2. Are We in an AI Bubble?

It would be unusual, given the great enthusiasm, high asset prices, and overbuilding, if this boom doesn’t conform to the pattern of past bubbles and hype cycles. 

That AI and its architects also made TIME’s 2025 cover is, in my view, an ominous sign for investors. But beyond magazine covers, the fundamental signs of a bubble are flashing a warning.

The current AI buildout mirrors the railroad boom of the 1800s. Railroads were then the technology “central” to advancing civilisation, yet fortunes were lost when the debt-fuelled bubbles burst in 1873 and 1893.

We’re seeing a similar dynamic today. Sure, most of the AI Hyperscalers are using cash, but "GPU-rich" clouds are raising billions in debt with GPUs as collateral. CoreWeave, for instance, has been called "AI’s ticking time bomb," saddled with massive debt and no obvious path to profitability. Reports indicate CoreWeave’s GPU-backed debt has exceeded $US10 billion.

Doug O’Laughlin of Fabricated Knowledge and SemiAnalysis fame has noted that what was once a "disciplined, cashflow-funded race may now turn into a debt-fuelled arms race.' Even major players like Meta are teaming up with private capital firms to form Special Purpose Vehicles (SPVs) to fund data centres, while keeping the debt off their books.

I reckon today’s AI investors need to recognise a fundamental difference between the tech darlings of the last decade and the AI giants of today. Since 2010, we, like many investors, have loved Software as a Service (SaaS) businesses because their code is written once and then sold infinitely.  Consequently, the Saas business enjoyed gross margins of up to 90%.

AI is not SaaS. Delivering AI-generated outputs from prompts introduces a tangible, physical cost to every interaction. A standard Google search costs roughly $US0.00003, whereas a generative AI response costs upwards of $US0.01 to $0.02—a 300 to 400-fold increase in marginal operating costs.

And let’s not forget how many trillions are being invested to build the infrastructure. If every one of the world’s 1.6bn iPhone users pays $US35 per month on AI tools, maybe these hyperscalers can generate enough revenue to generate a return that covers their cost of capital.  Maybe. But history says, probably not.

You see, the AI theme is perceived to be structural – a smooth north-easterly line out into the future.  The reality is the AI customers aren’t structural.  They’re cyclical.  When a structural wish meets a cyclical reality, hope gives way to selling.

The hype has also created a disconnect between price and reality. Nvidia reached a market capitalisation of $US5 trillion despite 2026 revenue forecasts of just $US65bn—a valuation 77 times revenue. Meanwhile, OpenAI is valued at half a trillion dollars despite projecting current losses of $US9bn to grow to an annual loss of $US74bn by 2028.

And then there are also physical constraints. OpenAI’s goal of 250 gigawatts of capacity by 2033 is reportedly more than India's entire peak power demand. Plans for Sydney’s datacentres are estimated to require 20 per cent of the city’s water supply.  Limits are assured. 

Where to Diversify

If we accept that valuations are elevated and a "shadow" is being cast over the market, it’s rational to take some profits and diversify. The goal is not to exit the market entirely—that is an immature approach—but to rebalance toward assets with nil or negative correlations to the S&P 500.

Alternative funds may sound risky, but wholesale and sophisticated investors have been adopting them to diversify for decades.  Long/Short Funds, Market Neutral Funds and High Frequency Arbitrage Funds – strategies I have written about elsewhere - are strategies that take no directional risk, and profit from volatility and market inefficiencies rather than the price movement. Listed versions are also available.

One example of a fund I know of that isn’t listed, however, arbitrages futures on digital currencies, locking in a profitable spread at the time each trade is established.  Trading thousands of times a day using computer-driven algorithms, the fund has generated a return of just over 22% p.a. From a diversification perspective, since its inception, the fund has produced a positive return in every month the S&P 500 has fallen.

For those seeking to diversify some of their AI profits and prioritise income, carefully chosen private credit funds could be the solution. Funds that eschew property development exposure, focus on secured lending, and are externally-rated BBB or higher are worth considering. Some listed options exist, but keep in mind that in any sell-off, they will trade—albeit temporarily—like equities.

Critically, the alternatives must come with robust safety profiles; otherwise, you’re jumping from the frying pan into the fire. 

The higher the price you pay, the lower your return will be.  With the S&P500’s PE ratio at historic highs, I suspect the big gains have been made and the next few years will produce single-digit average annual returns.  Indeed, at today’s PE ratio of 22.75, US-based Apollo Global Management predicts the next decade will see the S&P 500 produce negative average annual returns. 

If, at best, the S&P500 returns even low positive single-digit returns, and remembering stock market declines of 30-50% are possible, if infrequent, then rebalancing your portfolio ahead of 2026 seems a wise move.  You might miss out on being fully invested in another great year, but you’ll be diversified and in funds that still produce attractive returns.  It’s just that those returns won’t be correlated if the equity market tumbles.

PE ratios stretching valuations and the classic signs of a bubble on display, 2026 will likely look very different from the last three years. If you’d like to discuss how to rebalance your portfolio with our team, give David Buckland or Rhodri Taylor a call on (02) 8046 5000.

All prices and analysis at 17 December 2025.  This information has been prepared by Montgomery Investment Management Pty Ltd ABN 73 139 161 701 AFSL 354 564. The content is distributed by WealthHub Securities Limited (WSL) (ABN 83 089 718 249)(AFSL No. 230704). WSL is a Market Participant under the ASIC Market Integrity Rules and a wholly owned subsidiary of National Australia Bank Limited (ABN 12 004 044 937)(AFSL No. 230686) (NAB). NAB doesn’t guarantee its subsidiaries’ obligations or performance, or the products or services its subsidiaries offer.  This material is intended to provide general advice only. It has been prepared without having regard to or taking into account any particular investor’s objectives, financial situation and/or needs. All investors should therefore consider the appropriateness of the advice, in light of their own objectives, financial situation and/or needs, before acting on the advice.  Past performance is not a reliable indicator of future performance.  Any comments, suggestions or views presented do not reflect the views of WSL and/or NAB.  Subject to any terms implied by law and which cannot be excluded, neither WSL nor NAB shall be liable for any errors, omissions, defects or misrepresentations in the information or general advice including any third party sourced data (including by reasons of negligence, negligent misstatement or otherwise) or for any loss or damage (whether direct or indirect) suffered by persons who use or rely on the general advice or information. If any law prohibits the exclusion of such liability, WSL and NAB limit its liability to the re-supply of the information, provided that such limitation is permitted by law and is fair and reasonable. For more information, please click here.