This article was originally produced in conjunction with Boring Money for their Insights.
Some information contained herein has been obtained from third party sources and has not been independently verified by Polar Capital. Neither Polar Capital nor any other party involved makes any express or implied warranties or representations.
The potential, and the hype, around AI is no longer a novelty. The press has lauded the rapid advancements in the sector for several years already, and this has only been magnified by the success of generative AI software such as Microsoft’s ChatGPT and Google’s Gemini (formerly Bard). Now it hardly seems a day goes by without either using AI, or at the very least discussing it with someone at work - or even at home. We may be only halfway through the 2020s, but it would be difficult to argue that AI isn’t already the story of the decade.
However, the tech sector’s meteoric rise has not been without its wobbles. The recent volatility in tech stocks at the start of 2025, triggered by the launch of controversial Chinese AI DeepSeek, has served as a stark reminder that investor sentiment can shift quickly. Technology cycles are often vulnerable during their early years, and the start of this year has seen some sharp pullbacks, particularly in AI-heavy stocks.
Major AI milestones over the last 15 years
 |
Demis Hassabis, the CEO of Google’s AI research laboratory DeepMind, says: “If we get it right, it should be an incredible new era of discovery and a new golden age, maybe even a kind of new renaissance”.1
However, it’s a narrative we’ve been subjected to somewhat relentlessly in recent years, and concerns are well and truly seeping in that the skyrocket momentum AI (and the wider technology sector) has enjoyed may have finally run out of steam.
The launch of Chinese firm DeepSeek sent shockwaves through markets around the world in January, as bigwig US tech companies grappled with the uncomfortable reality that they may not be able to sustain their preeminence in this rapidly advancing sector. Nvidia, the foremost producer of AI hardware components globally, saw an unprecedented –17% fall in its share price, wiping $600 billion off its titanic market value.2
So clearly, tech stocks and the sector at large are not invincible, despite what their performance over the last few years may have led some to believe.
Are tech stocks currently overvalued and overconcentrated?
The question on the tip of many investors’ tongues is, invariably, how long can the tech sector’s hot streak go on for? And is it all about to come crashing down?
There is much to be said about the argument that AI stocks in particular are overvalued, and many of the largest stock market indices in the world have a “perilous” overconcentration in some of the big names – Google, Microsoft, Tesla and the like.3
Take Tesla, for instance, one of the foremost names in the AI scene. The American automotive firm shot to a market value of over $1 trillion in 2024 and, thanks to the bombastic reputation of its founder and CEO Elon Musk, was almost impossible to ignore.4
Despite its share price soaring over +90% in 20245, analysts are apprehensive about the firm’s near-term prospects. Morningstar strategist Seth Goldstein assigned Tesla stock a fair value estimate of $210 per share in December 20246, but at the time of writing, it’s trading at almost twice that price.7
It has a Morningstar Rating of just 1 star, meaning it’s considered significantly overvalued. Yahoo Finance’s roundup of analyst ratings also shows the average price target is considerably below Tesla’s current value.8
The table below illustrates the huge range of price targets for Tesla stock, with a spectrum spanning from $135 to $550. However, with the average sitting at a little under $400, it’s evident that many experts think Tesla is nevertheless significantly overpriced.
Average analyst price target vs current share price for Tesla, Yahoo Finance
Current Price* | Average Analyst Price Target | Lowest Analyst Price Target | Highest Analyst Price Target |
$383.68 | $324.99 | $135.00 | $550.00 |
| Source: Yahoo Finance, correct as at 4 February 2025. | |||
Goldstein explains Tesla’s overvaluation is in part due to the seemingly unchallenged assumption that Tesla won’t put a foot wrong over the next year, and that this puts investors in a precarious position if it isn’t able to deliver on its promises:
“We don’t think investors should buy Tesla at the current price. The current stock price implies nearly everything goes right for Tesla in terms of new product launches, revenue growth, and margin expansion”.9
Meanwhile, concerns continue to grow about the stock market’s overconcentration in megacap tech stocks, including Tesla and the rest of the so-called “Magnificent Seven”. The 10 largest US stocks account for over 37% of the S&P 500’s market value, well above the 27% share at the peak of the tech bubble back in 2000, and all but one of these firms belong to the technology sector.10
| S&P 500 sector breakdown |
 |
| Source: S&P Global, correct as at 4 February 2025. |
According to Howard Silverblatt, Senior Index Analyst at S&P Dow Jones Indices, tech stocks were responsible for 56.5% of the S&P 500’s total +25% return over 2024. Excluding tech companies would have seen the index return a comparatively meagre +11% instead.11
But in the shadow of the sector’s meteoric rise, Christopher R. Jackson, Senior Vice President at UBS Wealth Management, says investors should be wary of whether many of these firms can sustain their momentum:
“I don’t think anyone would tell you that AI is not a generational investment theme, but I think the concern, especially over the near term, is how quickly things have gone”.12
Jackson’s thoughts are echoed by Jim Covello, Head of Global Equity Research at Goldman Sachs:
“AI technology is exceptionally expensive, and to justify those costs, the technology must be able to solve complex problems, which it isn’t designed to do”.13
“It’s certainly a risk to the thesis,” concedes Jackson. “Can companies continue putting out this kind of money without any near-term return?”14
Will the costs start catching up to tech giants?
As demand for AI infrastructure snowballs, tech firms are shelling out billions of dollars trying to get hold of semiconductor chips and graphics processing units (GPUs) - both crucial pieces of AI hardware.
Recent research indicates 43% of firms plan to increase their AI hardware budgets by 10-25%15, while David Cahn, Partner at Sequoia Capital, believes that AI companies will soon have to earn about $600 billion per year to offset the cost of their AI infrastructure.16
Nvidia is one of the tech stocks thought to be poised to benefit from this spending splurge. The American firm manufactures semiconductors, GPUs and a range of other hardware integral to AI and computing systems. Off the back of the soaring demand for its services, Nvidia’s share price famously rocketed +171.2% in 2024 alone17, and has booked eye-watering returns of over +785% to date since the start of 2023.18
| NVIDIA performance |
 |
| Source: FE FundInfo, correct as at 4 February 2025. |
With the need for AI hardware gathering pace, tech firms are preparing to pump up their spending budgets. This year, analysts expect Microsoft to spend $84.24 billion into improving its AI infrastructure19, and research from Goldman Sachs anticipates mega tech firms, corporations, and utilities will spend around $1 trillion in the coming years to support AI development.20
The concern, therefore, is that this focus on hardware acquisition could start to drag on profits. But the advent of DeepSeek has called into question whether this need be the case. The Hangzhou-based company claims to have been able to develop its R1 model for a mere fraction of the cost that many of the major tech firms have been paying to date, at just $6 million.21
Some have suggested this has pulled the rug out from the feet of those who have been forecasting massive AI spending over the coming years. John Naughton, Professor of Public Understanding of Technology, says that DeepSeek’s arrival has the potential to completely undermine the preeminence of the US tech giants:
“The low training and inference costs of R1 will turbocharge American anxiety that the emergence of powerful – and cheap – Chinese AI could upend the economics of the industry, much as the advent of the PC transformed the computing marketplace in the 1980s and 90s”.22
Marc Andreessen, a former software engineer who has advised US President Trump on tech policy, has already heralded DeepSeek as a watershed moment for the AI industry:
"Deepseek R1 is AI's Sputnik moment," he said in a recent post on X - a reference to the 1957 Russian satellite launch that triggered the Cold War space exploration race between the Soviet Union and the USA.23
Consensus remains tech is a long-term play
Despite this shake-up for the industry, and although the jury is still out on the near-term, the general consensus remains that the long-term prospects for AI and tech stocks on the whole are compelling enough to justify any wobbles along the way.
“While we understand fears that AI excitement will fade, we believe the opposite is true and that the longer-term potential of this highly disruptive technology will become much more visible as 2025 progresses,” says Ben Rogoff, Lead Manager at the Polar Capital Technology Trust.24
So strong is this conviction that Rogoff told The Times in November 2024 that he expects AI to “have a transformative effect on the world in the same way that steel, electricity and the internet once did”.25
Looking ahead at 2025, a recent BMP report states the technology sector is “primed for continued change and reinvention”26 and McKinsey & Company has already forecast that AI will add trillions to the global economy in the coming years.27
This evolution is set to spread across all corners of the tech sector. The metaverse, for example, which allows individuals to interact in shared virtual reality spaces, is expected to generate up to $5 trillion in value by 2030 – about the size of Japan’s entire economy.28 In addition, there are forecasts estimating the global cybersecurity market could reach $425bn by 2030 – up from $154bn in 2022.29
 |
Adam Benjamin, Portfolio Manager at Fidelity, is one of the chorus of experts who are optimistic about tech’s prospects, but also suggests investors should retain a degree of caution in the near-term:
“There is already evidence of AI-driven gains in productivity, customer service, coding, graphic design, and translation services, and I believe AI could eventually be disruptive to the way most work is performed. Progress may not be linear, though, and investors must be mindful of stock valuations and the timing and potential impact of further technological advances in the field, as well as the broader macroeconomic environment”.30
At Polar Capital, Rogoff also acknowledges there may still be some bumps in the road ahead, but emphasises these are likely to be short-lived and should not put off investors who are looking for a long-term growth angle for their portfolio:
“As AI adoption accelerates, supply constraints ease and the benefits of adopting generative AI likely become more obvious during the year, we expect to see a broadening in the technology market as a wider range of companies provide both the enabling technologies and benefit from the application of AI.”
Alastair Unwin, Deputy Manager at the Polar Capital Technology Trust, added:
“Looking ahead to 2025, we expect strong demand for NVIDIA’s Blackwell processors, coupled with high-profile generative AI model upgrades and product releases, will drive further investor interest in AI-exposed stocks. Pro-business policies from a Trump/Republican/Musk government should ensure the US remains one of the stronger global economies and at the forefront of generative AI development, while regulation does not hinder AI progress”.31
Rogoff is adamant that, while fears around overvaluation and overconcentration are valid and near-term volatility may continue, AI is still one of the most compelling investment opportunities out there:
“We are all-in. We are AI maximalists. This is the worst AI will ever be, which if you are worrying about it is a concerning observation, but if you are excited and investing in it, is hugely important. I get a bit goosebumpy thinking about it”.32
