What tech stocks' volatility tells us about the future of AI

How the recent tech volatility has reignited debate over the future of AI and the possibilities for smaller caps

Artificial Intelligence technology has been an important driver of equities over the past 18 months.

For example, at Microsoft, capital expenditures are up 310 per cent since late 2019, with similar figures seen at Google.

The magnitude of the moves seen in AI-related companies, particularly within the mega-cap space, have propelled their market capitalisation and therefore their weightings in stock market indices. 

They have, as a result, become bigger drivers of returns for investors, necessitating a larger degree of scrutiny. 

AI bubble comparisons

Recent volatility among big tech stocks has reignited the debate about the viability and direction of travel for AI.

James Burns, head of MPS at Evelyn Partners, says: “So far the excitement has been in the infrastructure of AI, such as chips. I think the debate regarding it being a bubble or not comes down to whether AI can help generate revenues for companies, eg Copilot for Microsoft.”

Chris Ainscough, director of asset management at Charles Stanley, says that while the term bubble may be slightly extreme for current conditions, there has certainly been a significant level of optimism priced into large-cap AI names that assumes high levels of growth of revenues and/or profitability.

“Despite this, there remains a lot of uncertainty surrounding the integration of AI within corporate ecosystems. This creates risks, especially against a backdrop of significant positivity," Ainscough adds.

Investor uncertainty

The majority of investments in AI hardware is driven by large cloud players and internet companies. 

Martin Hermann, portfolio manager, equities, at Berenberg, notes that recently investors have been questioning the return on these investments, since payoffs on software and efficiency improvements are multi-year in nature. 

Hermann says: “During recent earnings seasons, tech companies have alleviated some of these fears since they elaborated on the importance of infrastructure upgrades and committed to increase AI investments in 2025.

"Historical build-out phases of IT infrastructure have almost always led to overinvestments due to long investment cycles and the difficulty in forecasting demand. 

"We believe this AI cycle will play out in a similar fashion, but so far the indications for investment spending in 2025 have been encouraging.”

Additionally, Hermann notes that any comparison with the dotcom bubble is unfair.

He adds: “With regards to an AI bubble, it is worth noting that the valuation of some major AI beneficiaries – Microsoft, Amazon, Google, Meta, Nvidia – currently stands at 27-times expected next 12-month earnings, which is below the 10-year average of 31-times. 

"The increase in stock prices has gone hand-in-hand with rising earnings estimates, for example, +70 per cent for Amazon, +49 per cent for Meta and +21 per cent for Microsoft over the last 12 months, which is fundamentally different from the situation during the dotcom bubble.”

In the run-up to the dotcom bubble bursting, the growth of the internet created a buzz among investors who were quick to pour money into startup companies.

These companies were able to raise enough money to go public without a business plan, product or track record of profits. They quickly ran through their cash, which caused them to go under.

But investors are beginning to ask themselves whether future revenue streams are coming fast enough to support current tech investments – recently highlighted by the volatility experienced by tech stocks.

Peter Garnry, chief investment strategist at Saxo Bank, says: “All the new AI investments are fundamentally changing the fundamentals of the world’s leading software companies from being asset light to more capital intensive and with higher reinvestment rates. 

“That [means potential headwinds] for technology companies down the road. We are positive on the long-term outlook of AI, but in the short-term volatility will increase as expectations will be reset as the market begins to doubt the speed of AI application and subsequent revenue streams to sustain the pace of AI investments.”

Tech volatility outlook

The recent volatility seen in tech names has stemmed from a series of events that led to a reduction in risk appetite across investment markets.

These included poor US economic data, geopolitical tensions and sharp changes in the value of the yen. These events created an environment where investors were reducing risk exposure. 

Ainscough says: “Given the extended valuation levels of technology companies as a whole, as well as their recent strong outperformance, they were more susceptible to volatility in this environment than other parts of the market. 

“Despite having higher levels of technology exposure, US equity markets fell in a similar magnitude to Europe, whereas the UK fared better with its more defensive characteristics. 

“Japanese equity markets saw the largest falls due to concerns regarding the pace of interest rate rises from the Bank of Japan at a time when the rest of the world are looking towards rate cuts.”

Evelyn Partners’ Burns says valuation is the principal reason for the volatility seen in tech stocks: “A lot of hope has been built into valuations, so any factors that might knock the story are having a disproportionate effect.”

“More specifically software has struggled YTD due to lower IT budgets, but China and politics (regarding the chip companies) have also played a part. 

“This has had a more pronounced effect in the US and European marketsl, where tech is a much greater part of the indices, but this has obviously not been relevant in the UK due to its lack of global tech leaders.”

If AI is unable to generate the expected revenues for companies like Microsoft, Burns adds, “then the huge amounts spent on capex make no sense and spending on chips and infrastructure will just disappear overnight. However, so far, the evidence suggests there will be real-world uses for it”.

Valuations have come down a bit after the recent sell-off. However, depending on the company, earnings estimates have also decreased somewhat compared with early summer.

Impact of tech volatility

The impact of volatility in tech stocks on investors has meant that, due to equity market concentration, any changes to expectations around AI are causing increased volatility, and many individual investors are overweight on technology, so they experience a higher degree of volatility.

Ainscough says not only are technology stocks a larger portion of some regional equity indices due to their recent outperformance, but surveys have shown that asset managers are generally overweight the sector as well. 

With investment portfolios generally having greater exposure to technology stocks as a result, the recent volatility will have had a larger impact on portfolio performance. 

However, the drawdowns have proven short lived, with a rapid recovery witnessed over the past few weeks. 

He adds: “Those that kept a calmer head and maintained positions rather than selling down during the initial market panic will have recovered the majority of the losses seen. 

“One associated impact that we’ve seen during recent volatility is the return of the diversification benefits of government bonds within multi asset portfolios.

"With inflation at lower levels and the rate cutting cycle beginning/already begun in western economies, investors are now more comfortable relying on these securities.”

Russ Mould, investment director at AJ Bell, says if investors panicked and sold off their stocks in a rush they would have felt an impact from the volatility, “as they may have missed the subsequent rebound or even locked in losses on positions”. 

Mould adds: “Volatility is really the friend of the investor, not the enemy, as it gives them chance to buy cheaply and sell expensively when others panic, but it requires discipline and fortitude to do that and go against the crowd. 

“It also requires a portfolio that is designed to see the investor through such squalls, and that is prepared for several possible market outcomes, not just one.”

But probably the most obvious impact from the recent volatility has been the debate about market-cap versus equal weight. 

“It all boils down to diversification, but it has obviously been a tough period for active management, particularly in the US,” Burns adds.

Focusing on underlying innovations

According to Rahul Bhushan, managing director at Ark Invest Europe, short-term volatility is to be expected in high-growth areas.

As a result, he stresses the need for investors to stay focused on the underlying innovations that are setting the stage for the next era of economic growth. 

He views these fluctuations as “opportunities to invest in the leaders of tomorrow at potentially attractive valuations”.

Bhushan adds: “Volatility in tech stocks has certainly impacted funds, especially those heavily weighted towards disruptive innovation. However, we view this as part of the natural evolution of high-growth sectors. 

“Our conviction in the long-term potential of these technologies remains unwavering. In the short term, our funds might experience fluctuations, but over the long term we believe that investors who stay the course will be rewarded as these innovations mature and deliver exponential growth.

"As we’ve seen in the past, periods of volatility often precede significant breakthroughs and market leadership."

Small-cap opportunities

While AI is currently dominated by the large-cap companies that are building out the infrastructure with massive capex, in time, once the plumbing is in place, the implementation of AI could also open up possibilities for small and mid-caps.

Rupert Rucker, investment director, US small cap, at Schroders, says: “We are currently in the build-out phase of data centers with AI (back-end) which will eventually lead to the implementation of that AI. This is currently dominated by large and mega-cap companies in terms of capex and the necessary components like GPUs. 

“The next phase will be implementation and efficiencies, which is when demand will spill over to service providers, which are found much more in the mid to small-cap universe. 

“Smaller companies tend to be more of the picks-and-shovels type that will enable the AI revolution, but more over the long term as we all try and work out how to make the new technology productive. We saw a similar cycle with the internet in the 2000s.”

Charles Stanley’s Ainscough says the roll-out of AI technology is more nuanced than simply investment by mega-cap companies accruing down the size spectrum.

He adds: “The AI theme is certainly dominated by mega-cap companies currently. They have the capital to invest in the infrastructure but also to acquire smaller companies that have made, or look key to making, significant technological progress. 

“Given this significant investment, however, these mega-cap companies will require significant returns in the future.

"This is particularly pertinent given the preference of these companies to overinvest today in order to avoid conceding competitive advantage to other businesses. 

“The monetisation of AI will therefore become the key next stage in the proliferation of the technology, meaning that the productivity benefits for smaller companies will come at a cost.

"The cost benefit trade-off for these businesses remains uncertain, but some business models lend themselves more to the benefits of the technology than others.”

Valuations of AI-focused companies can sometimes seem stretched when viewed through a traditional lens, Ark Invest’s Bhushan says.

However, he says traditional valuation metrics often fail to capture the full scope of opportunities presented by AI and other disruptive technologies.”

According to Bhushan, these companies are not just playing in the margins; they are redefining entire industries and creating new markets. 

For instance, Meta is a unique case with its open-source large language models, powered by its own data centers. 

He adds: “This gives Meta a strategic advantage that is often undervalued by conventional models. Similarly, companies like Palantir and Databricks are not only likely undervalued but also misunderstood by most investors. They are integral to the AI ecosystem, providing the infrastructure and tools necessary for AI to thrive across various sectors.

"Moreover, the demand for AI training and inference is creating new layers of value within the AI stack.

"Nvidia’s dominance in AI training underscores the importance of specialised hardware, but as we transition to inference, a broader array of companies, including those focusing on custom silicon and AI-specific infrastructure, will likely see their valuations rise. 

“This is where valuation discrepancies can arise, and we are using the current volatility to double-down on our core bets."

Ima Jackson-Obot is deputy features editor at FT Adviser

ima.jacksonobot@ft.com