Is there an AI bubble?

What happened in May?

In May 2024, tech giant Open AI announced its latest model, GPT 4o (“four-oh”). Videos of the large language model (LLM) tutoring mathematics, translating conversations between multiple languages, and orchestrating games of rock, paper, scissors quickly spread online, and the internet lit up with chatter as people were blown away and creeped out in equal measure. The announcement began a series of high-profile Artificial Intelligence (AI)-related announcements from Silicon Valley giants that are expected to continue over the summer. Away from the buzz of excitement amongst tech enthusiasts, financial commentators continue to raise the potential of a bubble existing in the share prices of AI-related companies. Whilst some seem to believe this is the case, others believe that it’s likely that there are small bubbles for specific aspects of the AI revolution, but that the underlying technology itself is revolutionary, and therefore foundational (i.e. not a bubble) to the financial markets. As a discretionary fund manager with a passive buy-and-hold philosophy, ebi does not make short-term predictions of stock market movements and therefore does not have an opinion on the extent of a potential  bubble in the AI space. Nevertheless, here we look at why bubbles matter to investors in general, and what our view is on how investors can best weather bubble bursts when they inevitably do happen. 

Why might there be a bubble in 2024, and why does this matter?

When the dot-com bubble burst in early 2000s the Nasdaq (a US index weighted to technology and healthcare companies) eventually fell 77% from its peak value, causing significant harm to investors. Despite this being the most well-known technology related financial bubble of recent times, there are multiple instances throughout history of new technologies being overvalued by markets, only to underdeliver on expectations and disappoint investors. Even going back to the 1840s, when ‘railway mania’ took over the British investing landscape – the bubble burst when it became clear that many railways would not become profitable in the long term. Therefore, the question is now being asked; as AI mania sweeps the developed world, is this the latest bubble set to burst? 

Data from 1995 to 2010 | Source:

What causes bubbles to emerge?

The market value of a stock, that is, how much market participants are willing to pay for it, can diverge from the actual, or ‘intrinsic’ value that it provides. Under the widely used discounted cash flow modelling approach, company stocks should have an intrinsic value closely related to the value of the potential future cash flows (dividends or interest) that they are expected to provide. Sometimes however, excitement, excessive media exposure and herd mentality can lead investors to act irrationally. If the excitement around the technology has over-exaggerated the potential for the future cash flows that the underlying stocks can reasonably be expected to provide, then their value may naturally (and perhaps suddenly) correct downwards to a more accurate representation of their value. 

Although bubbles are a natural feature of financial markets, and their corresponding corrections are healthy in the long term, they can be disastrous in the short term for investors who didn’t intend to hold the stocks for the long term or, bought at the ‘top’ of the market. Similarly, investors who have ‘leaned in’ to the tech boom over the last couple of years by overweighting the technology sector, will suffer proportionally more than those who maintained a more thoroughly diversified portfolio if the sector does take a downturn.

How have markets responded in previous years? 

Every time that market bubbles have burst as a consequence of new technology, the overall market has gone on to eventually recover and surpass its ‘peak bubble’ value, but importantly, that doesn’t mean that individual companies are guaranteed to survive the crash or thrive after the market downturn. When considering the market as a whole, it isn’t so much a question of whether it will recover as how long the recovery will take. Even the market reaction to the COVID pandemic took only a few months to completely recover, rather than multiple years. If there were to be an AI-crash, it would likely bring the market ‘back down to earth’, but it would be widely accepted that the overall trajectory over the following decade or more would be in an upwards direction. 

ebi’s view

ebi has long employed a passive buy-and-hold approach to investment management and has stuck to this philosophy through periods of market boom and bust. Should the current market activity turn out to be a bubble fuelled by overly excessive AI ‘hype’, given the increasing weight of AI stocks in market indices, index-focused investors around the world would naturally experience declining values in their portfolios as the bubble corrects. However over the longer term, financial markets has grown consistently, as ‘deadwood’ stocks have fallen away as part of the correction, leaving only the companies adding real underlying value remaining, which drive growth into the future.

It is also worth noting that because the progress in AI is fuelled primarily by just a handful of technology companies, investors might be tempted to invest directly in the stocks of individual companies like Apple, Alphabet (which owns Google) or Meta. It’s vital to remember the principle of diversification and understand that overly weighting these individual stocks opens the investor up to the unsystematic (company specific) risk associated with the turbulence of internal politics. Open AI itself suffered conflict on its board of directors earlier this year, leading to CEO Sam Altman being forced to leave the company before quickly returning as a result of public pressure. Diversifying investments across a range of countries, sectors and asset classes can help avoid company specific risk and instead ‘ride the wave’ of upwards market momentum.

As ebi solely invests in the markets via funds, we do not invest directly in AI companies, however our portfolios still contain noticeable exposure to the main technology companies involved via our developed market equity funds by virtue of their size. Whilst over 8% of the equity exposure in our flagship Vantage Earth suite is allocated to Microsoft, Apple, Alphabet, Amazon, Nvidia or Meta, the Earth equity portfolio still contains over 8,600 holdings which include equities from around the world. ebi’s portfolios are also diversified across market sectors, including sectors like Health Care and Consumer Staples, that are less ‘boom and bust’ than the technology sector, meaning that investors aren’t ‘all in’ on the AI revolution. The Small factor, which tilts Earth (and our other factor-based suite, World) away from mega and large market capitalisation companies and towards smaller market caps, further means that their exposure to any AI bubble that may exist is reduced compared to portfolios that solely aim to mirror market performance. 

We continue to watch all market activity with interest but remain committed to our principle of buying and holding over long-time horizons, diversifying across a wide range of investments and using market factors like momentum, minimum volatility, small-cap, quality and value in seeking to deliver a premium for our investors over market returns.

What else have we been talking about?